TECHNICKÁ UNIVERZITA V LIBERCI

Transkript

1

2 TECHNICKÁ UNIVERZITA V LIBERCI KATEDRA VOZIDEL A MOTORŮ Sborník přednášek XLVIII. mezinárodní vědecká konference českých a slovenských univerzit a institucí zabývajících se výzkumem motorových vozidel a spalovacích motorů KOKA 2017 Liberec září 2017

3 Recenzent: prof. Ing.Celestýn Scholz, Ph.D. Technická univerzita v Liberci 2017 ISBN

4 OBSAH SIMULATION OF CYCLE-BY-CYCLE VARIABILITY OF SI ENGINE... 4 EVALUATION OF IGNITION PRE-CHAMBER GEOMETRY BY MULTI-ZONE MODEL OF COMBUSTION EFFECT OF METHANE CARBON MONOXIDE MIXTURE ON THE PARAMETERS OF COMBUSTION ENGINE USING OF WASTE HEAT OF INTERNAL COMBUSTION ENGINES AND DRAFT OF EXHAUST GAS EXCHANGER COMPARISION OF EMISSIONS FROM CONTINUOUS COMBUSTION OF CONVENTIONAL AND ALTERNATIVE FUELS FUEL INJECTION IN ENGINE SUCTION DUCT IN NON-STATIONARY MODES INCREASING THE MECHANICAL EFFICIENCY OF TURBOCHARGERS ELECTRICAL TURBOCHARGING THERMODYNAMIC POTENTIAL UNDER STEADY STATE OPERATION PHASE CHANGE MATERIALS FOR ENGINE WASTE HEAT ACCUMULATION AND STORAGE THE EFFECT OF EGR COOLER DESIGN ON PARTICULATE FOULING SNIŽOVÁNÍ EMISÍ OSOBNÍHO AUTOMOBILU A LEGISLATIVA IMPROVED IDENTIFICATION OF ENGINE PREDICTIVE MODELS OPTIMUM LIMITS OF MOTOR VEHICLE DRIVING MEASUREMENT WITH IMPROVED TUMBLE METER DYNAMIC CYLINDER DEACTIVATION FOR A SPARK IGNITION ENGINE THE VARIABLE VALVE TIMING MEASUREMENT AREA OF UTILIZATION OF ALTERNATIVE COMPRESSED AIR DRIVE PRACTICAL APPLICATION INFLUENCING OF ENGINE BLOCK LOAD BY MEANS OF CRANKSHAFT DESIGN ADDRESSING THE LAST FRONTIER OF CLEAN DIESEL ENGINES: REAL DRIVING EMISSIONS OF NITROGEN COMPOUNDS EMISSIONS AND PERFORMANCE OF A PASSENGER CAR SIZE DIESEL ENGINE FUELLED WITH HVO -DIESEL FUEL MIXTURES MĚŘENÍ PASIVNÍCH ODPORŮ PÍSTOVÉHO MOTORU PROTÁČENÍM SE ZVYŠOVÁNÍM TLAKU VE VÁLCI WAYS OF RECHARGING ELECTRIC BUSES AT STOPS ZEMNÍ PLYN JAKO AUTOMOBILOVÉ PALIVO EVALUATION OF THE IMPACT OF DIESEL FUEL ADDITIVE ON SELECTED COMBUSTION ENGINE PARAMETERS DIESEL ENGINE EMISSIONS IN REAL WORLD DRIVING: LABORATORY LIMITS ACHIEVED ON THE ROAD? SIMULACE JÍZDY HYBRIDNÍHO VOZIDLA PŘENOS VÝKONU A ŘÍZENÍ PARAMETRŮ VLIV VYBRANÝCH BIOPALIV NA VÝKONOVÉ PARAMETRY MOTORU ŠKODA ROOMSTER 1.4 TDI

5 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES SIMULATION OF CYCLE-BY-CYCLE VARIABILITY OF SI ENGINE Karel Páv1 Abstract This paper deals with utilization of an adaptive combustion model in order to simulate cycle-by-cycle variability of SI engine. The empirical adaptive combustion model consists of two parts: the first part for ignition delay prediction and the second part for in-cylinder combustion process description. There is proved mutual independence of these two phases that are generally given by normal distribution in terms of cycle-bycycle variability. The practical utilization of the cycle-by-cycle variability simulation is demonstrated by computational analysis of different variability levels at chosen engine part load. There is shown particular benefit in fuel consumption through the reducing cycle-by-cycle variability of the combustion process. 1. INTRODUCTION The cycle-by-cycle variability is among engineers dealing with engine development well known phenomenon. This variability is partly caused by continuously changing incylinder turbulence during intake period [1]. The resulting cycle-by-cycle air flow variations then lead to non-uniform cylinder charge structures movements. The mass of injected fuel is influenced, besides closed-loop feedback lambda-control, by the pressure fluctuation in the fuel rail. It causes the changes in mixture richness and consequently the cycle-by-cycle differences as well. However, the strongest impact on resulting cycle-by-cycle variability has the onset of the combustion process, which is particularly significant in case of standard SI engines. The flame formation is negatively affected due to non-homogeneous mixture in vicinity of the spark plug and intensive charge movement [2]. The magnitude of the cycle-by-cycle variability is being usually expressed by coefficient of variation COV of mean indicated pressure or maximum pressure. However, the variability of mean indicated pressure strongly depends on the piezoelectric pressure transducer quality, especially on the thermo shock drift [3]. Therefore, the variability of maximum pressure is more convenient parameter. Its limit value is COVpmax < 10%. 1 Doc. Ing. Karel Páv, Ph.D., Technická univerzita v Liberci, Katedra vozidel a motorů, Studentská 2, Liberec , Czech Republic, karel.pav@tul.cz 4

6 The main task of this paper is to investigate cycle-by-cycle variability in terms of combustion process differences, i.e. without turbulence effects on changes in mass of charge trapped in the cylinder. For the simulation of different conditions during combustion process the empirical adaptive combustion model [4] has been chosen. This model offers very simple implementation of variability control. 2. ADAPTIVE COMBUSTION MODEL The empirical adaptive combustion model consists of two parts. The first part predicts ignition delay ign which is defined as a crank angle difference between spark ignition and very first sign of the heat release, as graphically shown in Figure rb [1/ CA] Ignition delay ign ign TDC [ CA] Figure 1: Normalized burn rate rb, definition of ignition delay [5] Ignition delay is mathematically given by improved empirical formula originally coming from [4], [5]. The current form of this formula ign Aign Aign,v n 0.7 pign Tign 0.2 mix x r,st [ CA,1, 1, 1/min, bar, K, 1, 1] 5 (1) was achieved by investigation of more SI engines. The improvement consists in change of multiplier and single exponents based on regression analysis of extended experiments. Besides calibration factor Aign there is also multiplier Aign,v which is used for generation of the cycle-by-cycle variability. The ignition delay depends on the engine speed n, in-cylinder pressure pign and mean in-cylinder temperature Tign at the moment of spark ignition. Relative air/fuel ratio in unburned mixture mix is given by relative air/fuel ratio and by residual mass fraction xr in the cylinder mix xr 1 1 xr mix 1 1. (2) Mass fraction of stoichiometric residual gases xr,st is given by formula xr,st xr 1 Lst 1 Lst xr,st xr (3)

7 where Lst denotes stoichiometric wet air/fuel ratio. The main combustion process is described by change of mass fraction of the exhaust gas xexh,b along burning. Improved empirical formula covering wider range of SI engines has form 1.5 dxexh,b V Ab Ab,v n 0.5 p0.3 T 0.4 c d V 0.8 mix x 4.4 r,st xm1ix (4) [1/ CA,1, 1, 1/min, bar, K, m3, 1, 1, 1, CA] where p, T and V denotes instantaneous in-cylinder pressure, mean in-cylinder temperature and cylinder volume respectively. Vc is the compression volume at piston top dead centre and xmix is the instantaneous unburned mixture mass fraction in the cylinder. The multiplier Ab,v which extends the calibration factor Ab is used for the cycleby-cycle variability formation. The meaning of the angles and 0 follows from Figure 1. The mass of burned fuel mfuel,b is then dmfuel,b mc dxexh,b d 1 Lst d (5) where mc is the total cylinder mass. 2.1 Determination of calibration and variability factors The investigation of measured in-cylinder pressure data in terms of burn rate variability analysis had been carried out before ultimate cycle-by-cycle variability simulation. Determination of the single cycle based calibration factor for ignition delay comes from rearrangement of equation (1) Aign Aign,v ign,meas 0.2 ign 5 10 n p ign T 0.2 mix x r,st (6) 5 where ign,meas is the measured ignition delay for each single cycle. The product Aign Aign,v consists of the main calibration factor Aign which is constant for all cycles and the variability factor Aign,v which deviates from value 1 based on current cycle. Similar procedure has been done for the main combustion phase. For the determination of the single cycle based calibration factor in equation (4), the most important region at the rate of burning curve (see Figure 1) is the vicinity of its maximum. Therefore, this calibration factor can be determined based on in-cylinder conditions at 50% of burned fuel. Thus, can be derived Ab Ab,v rb,50% 1.5 V n 0.5 p500.%3 T500%.4 c V50% 0.8 mix x 1 x 4.4 r,st r 0.1 0, % (7) Similarly to previous case, the calibration factor Ab is constant over all cycles and variability factor Ab,v deviates from value 1 based on current cycle. Typical distribution of both mentioned variability factors is shown in Figure 2. There is obvious no mutual dependency between Aign,v and Ab,v. While the frequency distribution 6

8 of Aign,v is quite symmetrical and corresponds with normal distribution, the distribution of Ab,v is asymmetrically outspread to higher values. Frequency of Ab,v [%] Frequency of Aign,v [%] Ab,v [-] Aign,v [-] Figure 2: Typical distribution of variability factors for ignition delay A ign,v and main combustion phase Ab,v over 870 successive cycles As the distribution of the variability factor for ignition delay Aign,v is symmetrical, one can write it in symbolic form Aign,v 1 (8) where is a random positive number which deviates from zero and follows normal distribution. The combustion variability factor Ab,v was analysed in more detail by the help of calculation tool [6]. Evaluation of the frequency distribution of measured normalized burn rate maxima rb max shows its symmetrical pattern which approximately corresponds with asymmetrical distribution of the variability factor Ab,v in such a manner as depicted in Figure Ab,v [-] Frequency of rb max [%] aprox. 20 Ab,v 1 - Ab,v 1.0 Ab,v rb max rb max rb max [-] rb max [-] Figure 3: Typical distribution of normalized burn rate maxima rb max, relationship between rb max and combustion variability factor A b,v 7

9 In order to achieve the same positive and negative deviation for rb max the variation of combustion variability factor has to follow the relation Ab,v 1 1. (9) Although this relation is labeled as an approximate in Figure 3, possible differences from the real engine behavior are negligible. 2.2 Magnitude of variability factors As a measure of the cycle-by-cycle variability can be assumed standard deviation of particular variability factors Aign,v and Ab,v whose centric values correspond with medians of sequences. This approach allows separate assessment of preflame and main combustion phase. The situation in case of ignition delay is clear but the standard deviation of combustion variability factor Ab,v has to be evaluated for values lower than 1 only because of its asymmetrical distribution A n ign i 1 ign,v i Aign,v median A n 2, n 1 b b,v i i 1 Ab,v median 2 n 1 Ab,v i 1. (10) Typical standard deviations of variability factors as a function of ignition angle which is strong influencing parameter are shown in Figure 4. Other parameters like an engine speed, mean indicated pressure, combustion duration etc. do not show so close dependency. 0.5 Standard Deviation of Ab,v [-] Aign,v [-] Ignition Angle [ CA] Figure 4: Typical standard deviations of variability factors Aign,v and Ab,v as a function of ignition angle, bands of possible occurrences 8

10 3. SIMULATION OF CYCLE-BY-CYCLE VARIABILITY The Simulation of the cycle-by-cycle variability is here demonstrated at one engine operational point. The swept volume of investigated four cylinder naturally aspirated SI engine was 1.6 dm3. As the minimizing of the fuel consumption is the main goal, the engine part load at n = 2000 min-1 and IMEP = 2.5 bar has been selected. For purposes of this simulation, it was necessary to generate several sequences of variability factors of different variability levels as an input for calculation. The artificial generation of variability factors with respect of relations (8) and (9) has been done in LabVIEW 2010 environment by using function Discrete Random. There are shown patterns of variability factors Aign,v and Ab,v for five different variability levels given by standard deviation in Figure Ab,v [-] =0 = 0.1 = 0.2 = 0.3 = Aign,v [-] Cycle [-] Figure 5: Patterns of artificially generated variability factors Aign,v and Ab,v used for calculation The calculation was performed in modified calculation software [6] for 200 successive cycles. Since the calculation input was formed by the grid of 5x5 patterns of variability factors, it was possible to evaluate the contribution of single factors separately. Additionally, in order to consider the influence of different rates of burning the calculation was carried out for three calibration factors Ab = 0.5, 1 and 2 which represent slow, intermediate and fast combustion respectively. 3.1 Discussion of results The influence of both variability factors on fuel consumption at investigated engine operational point is shown in Figure 6. This calculation has been performed for intermediate combustion speed. The relative fuel consumption 100% was assigned to uniform combustion without any cycle-by-cycle variability effect. The base point in the graph marks the usual real standard deviations of Aign,v and Ab,v which can be theoretically improved. Even if the variability is reduced in such huge degree as marked 9

11 Fuel Consumption [%] by target point in Figure 6, the fuel consumption reduction can be expected less than 0.4%. The importance of both variability factors is similar as indicated by horizontal and vertical cut in Figure Base Point Standard Deviation of Ab,v [-] Target Point Standard Deviation of Aign,v [-] Fuel Consumption [%] Figure 6: Influence of variability factors Aign,v and Ab,v on fuel consumption at engine part load, n = 2000 min-1, IMEP = 2.5 bar, Ab = 1 The general potential for fuel consumption improvement in terms of cycle-by-cycle variability reduction is summarized in Figure 7. There is shown the relative fuel consumption as a function of coefficient of variance of maximal in-cylinder pressure which is simply detectable parameter. This value is normally COVpmax < 10%. In case of slow combustion, the potential for fuel consumption improvement is higher. However, the real achievable fuel savings are not higher than mentioned 0.4% Slow Combustion Intermediate Combustion Fuel Consumption [%] Fast Combustion Coefficient of Variance of PMAX [%] Figure 7: Fuel consumption as a function of coefficient of variance of maximal in-cylinder pressure at engine part load, n = 2000 min-1, IMEP = 2.5 bar 10

12 4. CONCLUSION The cycle-by-cycle variability is often discussed topic related to working cycle efficiency and thereby the subject for optimization. The simulation of the cycle-by-cycle variability is possible by using empirical adaptive combustion model extended by variability factor which can be artificially generated according to required variability level. This approach leads to very good agreement with real engine behavior. Although, it is theoretically possible to reduce cycle-by-cycle variability the carried out sensitivity analysis shows that expectable fuel consumption reduction at engine part load is up to 0.4% only. On the other hand, with increasing variability the engine efficiency deterioration is more progressive. Further step in simulation of the close to real engine behavior at full load is the incorporation of the knock prediction in used physical model. This simulation could make possible to find the most convenient strategy of the knock control in terms of maximum engine torque achievement. REFERENCES [1] [2] [3] [4] [5] [6] HEYWOOD, J. B.: Internal Combustion Engines Fundamentals, McGraw-Hill 1988, ISBN X BEROUN, S., PÁV, K.: Vybrané statě z vozidlových spalovacích motorů, Liberec 2013, ISBN PISCHINGER, R.: Engine Indicating, User Handbook, AVL Graz 2002 PÁV, K.: Adaptivní model hoření homogenní směsi ve válci zážehového spalovacího motoru, Habilitační práce, Liberec 2016 PÁV, K.: Adaptive Combustion Model for SI Engines, XLVII. International Scientific Conference of Czech and Slovak University Departments and Institutions Dealing with the Research of Combustion Engines, pages , Brno 2016, ISBN PÁV K.: Simulace pracovního oběhu 4T, Calculation software Microsoft Excel,

13 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES EVALUATION OF IGNITION PRE-CHAMBER GEOMETRY BY MULTI-ZONE MODEL OF COMBUSTION Jiří Hvězda1, Zbyněk Syrovátka2, Michal Takáts3, Jiří Vávra4 Abstract The paper deals with evaluation of ignition pre-chamber geometry using multi-zone model of combustion and heat transfer. Predictive model approach analyses mainly the losses of fresh methane air filling forced out of the system by expansion of earlier combusted part. Geometrical characteristics of given geometries have been obtained by means of CAD tool. The model described by differential-algebraic equation system is solved in numerical way. Selected results for two main designs of ignition pre-chamber including influence of spark plug position movement in output holes direction are presented. 1. INTRODUCTION A development of scavenged pre-chamber on a gas engine for light duty truck introduced in [1] has been followed by one negative phenomenon relatively small effect of inner combustion notable by means of the slight deviation of indicated in-prechamber pressure trace compared to the indicated in-cylinder pressure trace for just cranked engine. Jiří Hvězda, Josef Bozek Vehicle Centre for Sustainable Mobility Czech Technical University in Prague, Technická 4, Praha 6, Czech republic Jiri.Hvezda@fs.cvut.cz 2 Zbyněk Syrovátka, Josef Bozek Vehicle Centre for Sustainable Mobility Czech Technical University in Prague, Technická 4, Praha 6, Czech republic Zbynek.Syrovatka@fs.cvut.cz 3 Michal Takáts, Josef Bozek Vehicle Centre for Sustainable Mobility Czech Technical University in Prague, Technická 4, Praha 6, Czech republic Michal.Takats@fs.cvut.cz 4 Jiří Vávra, Josef Bozek Vehicle Centre for Sustainable Mobility Czech Technical University in Prague, Technická 4, Praha 6, Czech republic Jiri.Vavra@fs.cvut.cz 1 12

14 One of the possible reason for behavior like this, except low chemical efficiency or expressive cycle-to-cycle variability, is the abnormal loss of fresh filling caused by its forced-out to the main combustion chamber by expansion of already combusted part at spark plug area. To analyze this effect, multi-zone model of combustion and heat transfer processes was selected, as it is able to simulate this force-out phenomenon. Moreover, prechamber geometry including its spark plug position can be taken into account in simple way here. Two engineering designs of pre-chamber marked V1 and V2 are available. Those models are shown in Figure 1. Variant V1 has almost cylindrical shape, variant V2 is extended at spark plug region in volumetric way. So, its volume is considerably bigger compared to variant V1. At the bottom part, the both variants are equipped by 12 output holes with diameter of 1.2 mm. These holes are arranged in two cones, upper one having apex angle of 120 and lower one with 60. Figure 1: Pre-chambers V1 and V2 models The basic pre-chamber dimensions and spark plug coordinates are presented in Figure 2. Positions P1 are used for real design locations, positions P2 to P5 represent theoretical spark plug locations to evaluate an effect of significant spark plug movement in output holes direction, no matter if it is possible in design way. 13

15 Figure 2: Pre-chambers V1 and V2 dimensions and coordinates 2. MULTI-ZONE MODEL DESCRIPTION Multi-zone model of combustion and heat transfer processes was presented in [2] as an interesting alternative to perform the simulation of high-pressure part of working cycle of four-stroke spark ignition engine. Its differential-algebraic equation system is derived by means of the mass conservation law, energy conservation law, state equation, total volume condition and empirical equations for calculation of heat transfer coefficient and turbulent flame velocity. In general, each zone can be considered as thermodynamic non-isolated open system. Onward, there are the presumptions of common value of pressure, individual values of zone temperatures and general homogenous chemical composition in each zone. For the case of pre-chamber simulation and 3-zone model, the total volume is divided by means of spherical surfaces with center located in the gap of spark plug. Thus, fresh mixture and residual gases area (zone 1), flame front (zone 2) and the combustion products region (zone 3) are created. As it is shown in Figure 3 and explained in the chapter dealing with geometrical characteristics, multi-zone model, heaving information about available output holes flow areas, is able to recognize from which zone (and what kind of chemical component) the flows out of the system will occur. 14

16 Figure 3: Pre-chamber divided into 3 zones Change of general chemical composition in each zone is given by transfer and transformation of species shown in Figure 4. There are sequential transfer of mass using zone-to-zone mass flows, combustion and post-combustion processes in zones 2 and 3 and possibility of outflow from any zone. Figure 4: Transfer and transformation of species 3. GEOMETRICAL CHARACTERISTICS The differential-algebraic equation system described above contains a large number of geometrical quantities, which have to be supplied at every step of the numerical calculation. That is why, data files containing tabulated values of these quantities, named geometrical characteristics, have to be created. Geometrical characteristics are defined by the relationship between a quantity with a geometrical character and independent variables. Here, where pre-chamber events 15

17 are not dependent on piston kinematics, all geometrical characteristics for each zone are the functions of radius of forward spherical border surface and radius of backward spherical border surface only. Moreover, subtraction of volumes and areas can be used here again to decrease the number of independent variables for tabulation of geometrical characteristics, because the general geometrical characteristics always contain the entire data from the center of the spherical border surface to this border surface. At the end, pre-chamber general geometrical characteristics are defined just by the radius of spherical border surface, as only independent variable. Total number of geometrical characteristics is 16: Total volume of the pre-chamber Volume of zone i Area of the border surface between zones i and i 1 Area of heat transfer surface between zone i and pre-chamber wall Flow area for 12 pre-chamber output holes Graphical representation of these particular geometrical characteristics is shown in Figure 5. And again, AutoLisp programming language was used for automated volumes and areas modeling and data storage. Figure 5: Geometrical characteristics for pre-chamber V1 4. RESULTS Most of the following results belongs to the standard outputs of multi-zone model of combustion and heat transfer processes. The charts presented from Figure 6 to Figure 12 contain the results of combustion simulation in pre-chamber variant V1. For the case of this variant with real pre-chamber having spark plug position P1, indicated pressure traces for pre-chamber and main combustion cylinder are also available. 16

18 The values of turbulent coefficient for a front and a back flame front border and output holes discharge coefficient from this scope of simulation parameters have been used as crucial tuning constants to achieve the best possible accordance with experimental data: Methane-air mixture Engine speed 1800 rpm Ignition angle 20 before TDC Pre-chamber inner wall surface temperature K Filling condition at ignition point: air-excess coefficient 0.88 pressure 1.605e6 Pa temperature K Residual gases coefficient 0.25 Output holes discharge coefficient 0.65 Turbulent coefficient 0.22 / 0.07 (front / back flame front border) Combustion - chemical equilibrium approach using data from [3] The comparison of indicated and simulated pressures including its angular derivative are shown in Figure 6. As it can be seen, the effect of pre-chamber inner combustion is really very small. Figure 6: Pressures and pressure derivative Figure 7 introduces the all zones temperature trends as well as the mean one. Normally the highest temperature of zone 2 (flame front) is decreased here by significant outlet of firing components enthalpy through the output holes out of system. 17

19 Figure 7: Zone and mean temperatures The plots of corresponding zone volumes are presented in Figure 8. Here for the case of three-zone model, a fresh mixture area (zone 1) is transformed sequentially through a flame front (zone 2) to the combustion products region (zone 3) during combustion. Pre-chamber total volume remains on constant value, independent on piston kinematics. Figure 8: Total and zone volumes Progress of the front (Radius 12) and the back (Radius 23) border radius among zones is shown in Figure 9. Time derivatives of these radii represent the final velocities of their movements. Calculated values of the pure turbulent flame front velocity according to [2] are also mentioned for comparison. 18

20 Figure 9: Border radii, their derivatives and turbulent flame velocity The plots of corresponding ROB and ROHR curves and their cumulated values are shown in Figure 10. Once entire fuel is consumed, negative ROHR and its decreasing cumulative curve occur due to significant enthalpy outflow. Figure 10: ROB and ROHR and their cumulative curves Mass concentration trends for particular species during the combustion are shown in Figure 11. The equilibrium amount of fuel CH4 corresponding to the given ratio of fueloxidizer mixture is reached here. 19

21 Figure 11: Mass fractions The most interesting output for this project are the trends of instantaneous and cumulated (normalized by initial filling mass) mass flows through the output holes from particular zones presented in Figure 12. In this case, pre-chamber variant V1 with spark plug position P1, there is a loss of 44.6 % of fresh filling by its outflow through the output holes before affection of its remaining part by flame front. Figure 12: Mass output flows and their cumulative curves And the comparison of these loses for the both pre-chamber variants V1 and V2 with their five spark plug position options P1 to P5 follows in Figure 13. Primary movement of spark plug to the positions P2 and P3 does not lead to required drop of the escaped fresh filling amount. The losses could be even higher. There is a reason for it, that 20

22 shifted spark plug is surrounded by fresh filling in better way causing all directions flame front propagation followed by higher combustion pressures and stronger forceout effect. This phenomenon seems to be stronger than shorter flame distance and a time, which is available for creation of this kind of loss. Figure 13: Percentage of forced out fresh filling 5. CONCLUSION Above presented results proclaim a successful utilization of modified multi-zone model of combustion and heat transfer processes for evaluation of two geometrical variants of scavenged pre-chamber, mainly with reference to the spark plug position. In a model, these geometries are taken into account by means of the in-advance obtained geometrical characteristics for particular engineering design. Unfortunately, the both of pre-chamber mechanical variants V1 and V2 embody with their standard spark plug positions marked P1 a significant loss of fresh filling, which is forced out through the output holes before its affection by flame front. By the prechamber variant V2, even the volumetric extension of spark plug region does not bring any benefit in this respect. Obviously, bigger volume pre-chamber is profitable for ignition in general. On the other hand, the same percentage of fresh filling loss stand for bigger fuel waste in absolute numbers. At least the spark plug displacement to the positions P4 or P5 leads to the significantly smaller level of these losses. But, so radical movement is not feasible by the both variants in design way. On the basis of these results, it is recommended to overwork the pre-chamber cylindrical shape to more compact one, spherical one in optimal case. Moreover, spark plug position should be located closer to the output holes. In this way, the combustion process inside pre-chamber becomes more intensive due to the rapid flame front propagation in all directions. Thanks to the relatively short flame front track from spark 21

23 plug to output holes, outflow of combustion products will appear sooner, and the loses of forced out fresh mixture will be significantly lower. REFERENCES [1] [2] [3] Vávra J., Syrovátka Z., Takáts M., Barrientos E., Scavenged Pre-Chamber on a Gas Engine for Light Duty Truck, ASME, Internal Combustion Engine Division Fall Technical Conference, ASME 2016 Internal Combustion Engine Division Fall Technical Conference: V001T03A014, doi: /ICEF , Hvězda J., Multi-Zone Models of Combustion and Heat Transfer Processes in SI Engine, SAE 2014 World Congress, SAE Paper , Kee R. J., Rupley F. M., Miller J. A., Coltrin M. E., Grcar J. F., Meeks E., Moffat H. K., Lutz A. E., Dixon-Lewis G., Smooke M. D., Warnatz J., Evans G. H., Larson R. S., Mitchell R. E., Petzold L. R., Reynolds W. C., Caracotsios M., Stewart W. E., Glarborg P., Wang C., Adigun O., CHEMKIN Collection, Release 3.6, Reaction Design, Inc., San Diego, CA ACKNOWLEDGEMENT This research has been realized using the support of The Ministry of Education, Youth and Sports program NPU I (LO), project # LO1311 Development of Vehicle Centre of Sustainable Mobility. This research has been realized using the support of Technological Agency, Czech Republic, programme Centres of Competence, project # TE Josef Božek Competence Centre for Automotive Industry. These supports are gratefully acknowledged. 22

24 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES EFFECT OF METHANE CARBON MONOXIDE MIXTURE ON THE PARAMETERS OF COMBUSTION ENGINE Andrej Chríbik1, Marián Polóni2, Ján Lach3 Abstract Analysis of influence of individual components of synthesis gases helps to better understand the mechanism of combustion of synthesis gas fuel and consequently helps to achieve optimum parameters of the engine or cogeneration unit. The present paper deals with the influence of carbon monoxide in the mixture with methane on the power and economic parameters of the Lombardini LGW 702 combustion engine. It also evaluates and analyses the process of combustion of stoichiometric mixture flowing in the engine cylinder. It can be concluded that the hourly fuel consumption increases and the power parameters decrease with increasing carbon monoxide content in the fuel. The maximum cylinder pressure and maximum pressure rise rate values increase up to 90% vol. of CO content in the mixture and, as a follow-up, by increasing the CO proportion, these parameters significant decrease. 1. INTRODUCTION At present, more and more emphasis is placed on the most effective utilisation of waste as a source of energy. An important source of energy is represented by efficient recovery of energy from municipal waste through gasification leading to synthesis gases, which are subsequently effectively combusted in the combustion engine of cogeneration units. The synthesis gas produced generally consists of methane, hydrogen, carbon monoxide and inert gases (carbon dioxide and nitrogen). The proportions of the individual components depend on the production method and input raw materials entering the process of the synthesis gas production [1]. This article is a partial study of the influence of carbon monoxide in mixture with methane on the integral and internal parameters of the combustion engine with the aim to understand Andrej Chríbik, Slovak University of Technology in Bratislava, Námestie slobody 17, , Bratislava, andrej.chribik@stuba.sk 2 Marián Polóni, Slovak University of Technology in Bratislava, Námestie slobody 17, , Bratislava, marian.poloni@stuba.sk 3 Ján Lach, Slovak University of Technology in Bratislava, Námestie slobody 17, , Bratislava, jan.lach@stuba.sk 1 23

25 the influence of carbon monoxide alone on the overall parameters of combustion engine operated on synthesis gases. The following table (Table 1) lists the basic physical-chemical properties of the mixture of methane (CH4) and carbon monoxide (CO). Parameter Unit CH4 CH4 / CO [% vol.] 100/0 75/ Molar Mass Density (NTP) Lower Heating Value Air to Fuel Ratio Heating Value of Mixture Fuel in Air [g.mol ] [kg.m-3] [kj.kg-1] CH4CO25 CH4CO50 CH4CO75 CO 50/50 25/75 0/ [kg.kg-1] [kj.m-3] [% vol.] Table 1: Physical and chemical properties of methane (CH4) and carbon monoxide (CO) stoichiometric mixtures, NTP 20 C and Pa Increasing proportion of carbon monoxide in the mixture with methane causes that the lower calorific value decreases from the initial value 50.0 MJ.kg -1 for pure methane to 10.1 MJ.kg-1 for pure carbon monoxide. As can be seen from the table, increasing proportion of CO in the mixture brings about increase in the heating value of the fuelair mixture. Nevertheless, lower overall engine performance can be expected. The reason for this situation is a lower volume of the combusted gases while the CO proportion increases, which is due to the lower pressure in the cylinder of the combustion engine when combusting pure CO, compared to combusting CH4. The fuel consumption with increasing CO in the mixture with air rises up to 29.5% vol. and, on the opposite, the required amount of air to form stoichiometric mixture by gradually adding CO to the mixture decreases (Table 1) to the value 2.46 kg of air per 1 kg of carbon monoxide. 2. EXPERIMENTAL RESULTS Experimental measurements were performed on a naturally aspirated Lombardini LGW 702 spark ignition engine designed to drive a cogeneration unit [2]. In case of synthesis gases utilisation, the small swept volume (686 cm3) combustion engine with 12.5:1 compression ratio was a good choice for the experiments, because of reduced operating costs (fuel consumption). Various combinations of gases had been mixed from pressure bottles. The preparation of air-fuel mixture was performed by a diffuser mixer and the control of stoichiometric mixture was done by means of a feedback control unit. The optimal angle of ignition advance was set manually. All measurements were performed at full load. In order to determine the optimal ignition advance angles at the engine speed min-1, measurements of regulation characteristics were performed during operation on just methane alone, or just carbon monoxide. Subsequently, under the same number of revolutions, methane was gradually mixed 24

26 to carbon monoxide, approximately 5 l/min of methane. The performance and economic results of these measurements are shown in the following figures. For the analysis of the combustion process, also the pressure courses in the combustion chamber of the engine were recorded. To serve this purpose, a measuring system consisting of an integrated piezoelectric sensor in the KISTLER 6117 BDC15 spark plug, a charge amplifier, and an A / D converter were used. The pressure sensing in the pipe system was measured by a piezoresistive pressure sensor, also from KISTLER. The crankshaft position was monitored by the sensor Kistler 2614A. In the Matlab program environment, a program was developed to process data from the pressure sensors, the crankshaft position, and the moment of ignition timing. When measuring the regulation characteristics, the optimum angle of ignition advance (26 CA BTDC) for methane was found that allowed to achieve the best performance parameters (44.7 N.m), or for carbon monoxide (29 CA BTDC N.m). Each measurement allowed for an analysis of 195 consecutive cycles of the pressure in the engine combustion chamber. The following figure (Figure 1) shows the torque pattern depending on the amount of CH4 in the stoichiometric mixture with CO in constant operating conditions. The ignition advance angle was set to 27 CA BTDC. The maximum torque (44.6 N.m) was reached during the operation on methane. As the proportion of carbon monoxide in the mixture with methane was increased, the torque gradually decreased to 39.4 N.m, which represents approximately 11.7 %. As the CH4 was gradually added to the mixture with CO, at certain amounts of CH4 (about 5 to 10% vol.), the torque slightly decreased. This phenomenon will be explained in more detail in the next section. Hourly fuel consumption with increasing CH4 fell from the original value of 8.55 kg.h -1 for pure CO to 1.51 kg.h-1 for pure CH4. Figure 1: Course of momentum, hourly consumption of fuel and temperature of exhaust gases for various compositions of mixtures of methane with carbon monoxide. Conditions: start of ignition 27 CA BTDC, full load, speed of engine n = min-1, stoichiometric mixture Exhaust gas temperature was the highest (717 C) when carbon monoxide was burned and after the methane was gradually mixed, this temperature was reduced to 595 C 25

27 for pure methane. Exhaust gas temperature is related to the ignition angle, but mainly to the temperature during combustion. The adiabatic combustion temperature for CO is C and for CH C. With gradual addition of CH4 to CO the actual brake specific fuel consumption decreased from the original value g.kw -1.h-1 for CO to 223 g.kw -1.h-1 for CH4. The overall efficiency of the engine combusting CO was 25.8 %. After successive gradual addition of CH4 to the mixture with CO the efficiency rose up to 33.4 % of the value characteristic for pure methane. A slight decline of efficiency occurred around 10 % vol. of CH4, at which the efficiency reached 25.1%. Figure 2: Course of pressure in the combustion chamber as dependent on the crank angle, for various mixtures of carbon monoxide with methane (SOI start of ignition, TDC top dead centre) The above figure (Figure 2) introduces courses of pressure in the cylinder around of the top dead centre for the two-component fuel mixture (CH4CO) of CO and CH4. The lowest value of maximum pressure occurs while combusting carbon monoxide (4.7 MPa). Increasing CH4 by small amount in the mixture with CO (about 13.9 % vol. CH4), there occurs a sudden and considerable increase in pressure (7.1 MPa). The maximum pressure for combusted methane is 5.6 MPa. The position of maximum pressure after increasing CH4 up to 15% gradually approaches the top dead centre (TDC) and then, after further increase of CH4, it slightly diverges from the TDC. The coefficient of variation (COV) of maximum pressure for CO is 8.8 % and for CH4 this value is 8.1 %. By gradual increase of CH4 in the mixture with CO the coefficient drops sharply to 3.4 % at approximately 15% of present CH4 and then, under gradual increase of CH4, the coefficient rises up to the value for pure methane. A similar pattern as maximum pressure can be noticed with pressure rise rate. When combusting pure carbon monoxide, this value is MPa/1 CA and increasing CH4 to about 8 % vol. increases the pressure rise rate sharply to MPa/1 CA. Subsequently, by further increasing CH4, the pressure rise rate progressively decreases to MPa/1 CA for pure methane. 26

28 The methane in the mixture with carbon monoxide will also influence the course of combustion. Figure (Figure 3) shows the mass fraction burned (MFB) of the fuel and the fuel combustion rate (dmfb/dα) as dependence on the angle of crankshaft rotation for various compositions of the CH4CO mixture. The period between the start of ignition (SOI) and burning of 5 % of carbon monoxide is approximately 20.6 CA, compared to burning methane, in which case that time is 19.9 CA. A greater difference can be observed with the mixture containing 13.9 % CH4, in which case the period decreases to 14.6 CA. A similar pattern has been recorded with 50 % of fuel burned, in which case heat (363 CA) is fastest released from fuel at 13.9 % proportion of CH4 in the CH4CO mixture. For CO combustion, the angle at which the 50 % mass fraction has been burned is approximately 374 CA, or for methane it is 371 CA. The duration of the main combustion (10 % mass fraction burned 90 % mass fraction burned) is 30.8 CA for CO, or 24.7 CA for methane. A considerably short burning period (23.2 CA) has been measured for the proportion of 13.9 % vol. of CH4. The coefficient of variation (COV) for 5 % of mass fraction burned is 0.36 % for both CO and for CH4. The lowest COV value is for 13.9 % proportion of CH4 and is 0.24 %. Greater differences in scattering have been recorded at the position in which 50 % of the fuel has been burned. Specifically, for CO it is 0.68 % and for CH4 it is 0.58 %. Once again, the lowest COV value (0.41%) is at 13.9 % proportion of CH4 in the mixture with CO. The greatest differences in COV are recorded for fuel reduction (90 % MFB) when this value is for CO 1.10% and for CH %. The slightest difference between the individual cycles in which 90 % fuel has been burned is found for the position 13.9 % vol. of CH4. Its value is 0.29 %. Figure 3: Course of burning of fuel as dependent on the crankshaft angle for various mixtures of CH4 and CO (SOI Start of Ignition, SOC Start of Combustion, EOC End of Combustion) In general, the presence of a small amount (from 5 to 15%) of CH4 in CH4CO mixture significantly increases the rate of oxidation reactions. This is due to the Hcontaining species enhancing the oxidation reactions of CO with O2 by providing Hbearing radicals such as HO2, H, and especially OH during their own oxidation reactions with oxygen, which accelerate the oxidation reaction rates of CO in air [3, 4]. 27

29 3. CONCLUSION For burning methane mixed with carbon monoxide, the following findings regarding the power and economic parameters, as well as parameters related to combustion of the mixture of CH4 and CO in the engine cylinder at min-1 can be summed up: - by increasing the CH4 content in the mixture with CO, the volumetric heating value of stoichiometric fuel-air mixture is reduced from kj.m-3 for CO to kj.m-3 for CH4, - the performance parameters with a growing share of CH4 increase by about 11.7 %, from 39.4 N.m for CO to 44.6 N.m for CH4, - with increasing CH4 there occurs a decrease in the hourly fuel consumption, the difference between pure CO and pure CH4 is approximately 82 %, - increasing percentage of CH4 also leads to higher maximum pressure in the cylinder, from 4.7 MPa for carbon monoxide to 5.6 MPa for methane, - increase in the pressure rise rate has been recorded after gradual addition of CH4 to the mixture, from the original value MPa/1 CA for CO to MPa/1 CA for CH4, - with a small amount of methane (around 5 to 15% vol.) in a mixture with carbon monoxide, the mixture is significantly faster burned, which is also demonstrated by maximum pressure in the cylinder (7.1 MPa) or pressure rise rate (0.326 MPa/1 CA). This proportion of CH4 also leads to a more uniform running of the engine in terms of the coefficient of variation for the monitored internal engine parameters, - conclusive results show that a greater proportion of CH4 than CO in synthesis gases suits better to achieving higher power and less fuel consumption. REFERENCES [1] [2] [3] [4] AHMED I. I., GUPTA A. K.: Pyrolysis and gasification of food waste: Syngas characteristics and char gasification kinetics, Applied Energy, 2010, pp ISSN CHRÍBIK A., POLÓNI M., LACH J., RAGAN B.: Utilization of synthesis gases in combustion engine, KOKA 2015, Bratislava 2015, pp ISBN LI H., KARIM G. A.: Experimental investigation of the knock and combustion characteristics of CH4, H2, CO, and some of their mixtures, Proceedings of the Institution of Mechanical Engineers, Part A: Journal of Power and Energy, 2006, Vol. 220, pp ISSN LI H., KARIM G. A., SOHRABI A.: An Experimental and Numerical Investigation of Spark Ignition Engine Operation on H2, CO, CH4, and Their Mixtures, Journal of Engineering for Gas Turbines and Power, 2009, Vol. 132, pp. 8. ISSN ACKNOWLEDGEMENT This work was supported by the Slovak Research and Development Agency under Contracts-No. APVV and was also supported by the Scientific Grant Agency under the Contract No. VEGA 1/0301/17. 28

30 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES USING OF WASTE HEAT OF INTERNAL COMBUSTION ENGINES AND DRAFT OF EXHAUST GAS EXCHANGER Miloš Brezáni1, Peter Baran2, Róbert Labuda3, Dalibor Barta4 Abstract Article discusses about the use of heat exchangers for stationary combustion engines and cogeneration units. The paper is dedicated to the problem of unused thermal energy in stationary engines. It analyses possibilities of accumulation of heat energy and its possible application in various fields. The paper deals with the classification of heat exchangers and with the subsequent description of design solutions of heat exchangers types used in given field. Resolves draft of exhaust gas heat exchanger according to the required parameters of internal combustion engine and subsequent simulation in the simulation program Comsol multiphysics to verify the correctness of the design and the construction solution of the exhaust gas heat exchanger for stationary combustion engines. 1. INTRODUCTION Nowadays if we omit alternating economic crisis we can talk about ecological time. Political thinking towards just environmental but also economical, gives new insight into the lifestyle and comfort of man. A great impact just on these aspects has energetic. It is due to the increasing energy demands of human society, on which depends in no small measure the environmental burden and efficiency of energy use. Possibility how to reduce energy consumption, is the way of savings. Reduction in fuel consumption can be utilized in a direction, which deals with the production of several types of energy, and possibly also of the products from the primary source at the same time. To this category can include cogeneration, trigeneration and polygeneration. Find Miloš Brezáni, Ing., PhD., University of Žilina, Univerzitná 8215/1, Žilina, , milos.berezani@fstroj.uniza.sk 2 Peter Baran, Ing. PhD., University of Žilina, Univerzitná 8215/1, Žilina, , peter.baran@fstroj.uniza.sk 3 Róbert Labuda, doc. Ing. PhD., University of Žilina, Univerzitná 8215/1, Žilina, , rober.labuda@fstroj.uniza.sk 4 Dalibor Barta, doc. Ing. PhD., University of Žilina, Univerzitná 8215/1, Žilina, , dalibor.barta@fstroj.uniza.sk 1 29

31 a use for the heat is not as easy as in the case of electrical energy. But nevertheless is being offered several options, such as use of heat for hot water or direct water heating and its subsequent use for houses or large objects, depending on the performance of the cogeneration unit itself. Another option would be to use the absorption unit to transform heat to cold, making it possible to extend services to the production of cold water, for example for supply of air conditioners. Figure 1: Cogeneration principle For all these systems the energy transformation is decisive the method how to submit it. For this intention in case of heat is used inseparable part of most of the systems which is called the thermal coupling node. Thereby may be various types of heat exchangers, coolers, condensers etc. The most common devices nowadays belong heat exchangers. In this case, for the generation of thermal energy from the exhaust gas and its subsequent use in other applications. 2. USE OF HEAT EXCHANGERS IN COGENERATION UNITS AND STATIONARY INTERNAL COMBUSTION ENGINES. For use of stationary internal combustion engine to generate electricity, or in other applications, arises a waste heat [1]. In most cases, this heat is not used in any way, but today s time more and more forcing producers and consumers to invest in technology that can leverage the potential of unused energy and contribute to cost saving. To this end has started to use exhaust gas heat exchangers. An exhaust heat exchanger is positioned on the exhaust pipe, removing heat flue gases, which could then be used for various applications. 30

32 Figure 2: Position of exhaust gas exchanger in stationary combustion engine Exhaust gas temperature at the start of the exhaust pipe is in the range C. This means that the exhaust gases offer a great potential for utilization of waste heat. The exhaust gases in the most of cases heat up liquid, which can subsequently be used in several ways. 3. HEAT EXCHANGER Device used for targeted transfer heat energy from the one heat medium to another one, according to the second law of thermodynamics, is called a heat exchanger [2]. These facilities include a large group and can be found in many sorts of systems without us realizing it. According to the purpose and primarily according to the action, which takes place in the heat exchanger can be divided into the condensers, evaporators, coolers, regenerative heat exchangers etc. Another division is quite normal according to the method of heat transfer, i.e. whether there is contact between the media etc. Heat exchangers are divided into: Recuperative - media are separated by a solid impermeable wall and not coming into contact Regenerative - occurs periodically substituted flow heating and cooling media in the defined area [3]. Contact - media come together for some time in contact without chemical reaction, and then are separated. Mixing - media are in a certain place mixed and continuing as a mixture. Design with shaped tubes represent different tube axis arranged in the shape of a helix, spiral etc., located in the shell. 31

33 Figure 3: Spiral tube heat exchanger Exchangers in design tube in tube are among the simplest device in the above category. It s occurred like dismountable and non-dismountable which are exclusively for pure thermal media. Figure 4: Shell & tube heat exchanger The most commonly used type of heat exchanger is recuperative. This group primarily include tubular and plate heat exchanger. From the point of view flow is the most common counter-flow design, which results in better heat distribution than in parallel flow design. 3.1 TUBE HEAT EXCHANGER In this type of exchanger, heat exchange takes place between the tube and the tubularspace. Tubular space normally consists of pipes or tubes of circular cross section, but we can meet with cross-sections of other shapes such as oval, square etc. To reduce the dimensional parameters of tubular exchangers can use all sorts of ways to increase the area of the pipe from the side of the pipe as well as the tubular space. For this purpose are used, for example ribs. 32

34 Tube heat exchangers can be divided according to the construction on: With shaped tubes With straight pipes o Tube in tube o Tube in the shell In general the tube exchanger with jacket is the most commonly used heat exchanger, where the main structure consists of a tube bundle placed in the shell of a cylindrical shape. These exchangers are manufactured at many different versions, depending on the configuration of inlet and outlet orifices, pipes, construction attachment of different thermal dilatation of tubes and plastics etc. This type of heat exchanger typically includes partitions that perform two basic functions. Arrestment of tubes resulting in a reduction of bending and vibration and also primarily direct the flow of media that is purposely altered to the cross-flow that increase the intensity of heat transfer. The system also has the disadvantage that, with the inclusion of partitions create higher pressure drop. Figure 5: Tube heat exchanger with baffles Tube heat exchangers are characterized by good heat resistance and affordable price. However, their disadvantages are small compactness and high weight. The case of the pipes with small diameter, in which is the aqueous medium dirty it s expected gradual decrease of the cross-section pipe up to its complete clogging. Figure 6: Real construction of tube heat exchanger with baffles 33

35 3.2 Plate heat exchanger Plate heat exchangers are based on a patent that has already been registered in 1878 by German inventor Albrecht Dracke. This principle, when one liquid cooling another liquid and liquids are flowing on both sides of group thin metal plates, became the basis for the construction of the heat exchanger - commercial plate pasteurizer Alfa Laval. For more than 130 years was plate heat exchangers developed and structurally modified to devices that are used in thousands of different applications in all industries. Plate heat exchanger was previously designed for heating and cooling of the milk, but now is commonly used for heating and cooling in industrial processes and it is the basis of air-conditioning in buildings or it provides heating of hot water for hundreds of millions of people. This type of heat exchanger is characterized with a row lying plates which bear shaped reinforcements create turbulence of heat transfer medium and enlarge the heatconveying surface. The heat transfer medium, as shown in the figure flows between the slabs of small thickness, whereby the heat is transmitted between substances mainly convective. Plate heat exchangers can be sorted into dismountable and nondismountable. Non-dismountable exchangers are usually occur in the brazing or welding design, which can also be used in case of the aggressive heat transfer medium. For plate heat exchangers is a clear advantage of their higher performance per unit area, therefore low weight and small size which is for the same performance about 5 times smaller than in tubular heat exchangers. However, the benefits are offset by higher prices and demanding production technology. Figure. 7: Plate heat exchanger 34

36 Figure 8: Fluid flow in plate heat exchanger 4. DRAFT OF EXHAUST GAS HEAT EXCHANGER Draft of exhaust gas heat exchanger was realized according to these input parameters Exhaust gas side Required power 4,4 kw Cooling liquid side Flow 82,5 m3/h Flow 0,69 m3/h Inlet temp. Outlet temp. 584 C Inlet temp. Outlet temp. 89,5 C 115 C 95,6 C Table 1: Input parameters for draft of heat exchanger Based on the input values I calculated the parameters of the flue gas heat exchanger which are as follows: Structural design Calculated power 4,8 kw Number of pipes Length of heat exchanger Volume of heat exchange area mm 2,03 m2 Table 2: Calculated parameters of the exhaust gas heat exchanger Based on the calculated parameters I suggested and modelled flue gas heat exchanger. 35

37 Figure 9: 3D model of designed exhaust gas exchanger Figure 10: Stream direction of working substances cross the heat exchanger Stream direction of working substance cross the heat exchanger was chosen like a counter-flow how you can see in fig 10. On the proposed exhaust gas heat exchanger was made simulations of flow of exhaust gases and heat transfer in a 3D simulation program COMSOL Multiphysics. Figure 11: Layering of the mesh in the model 36

38 Figure 12: Simulation of velocity magnitude of exhaust gas heat exchanger in m/s Figure 13: Temperature of exhaust gas heat exchanger in degc Figure 14: Comparison of temperatures between inlet and outlet cooling liquid 37

39 Figure 15: Comparison of temperatures between inlet and outlet exhausts In figure 14, 15 can be seen the comparison between inlet a) and outlet b) temperature of cooling liquid and between inlet temperature c) and outlet temperature d) of exhaust gases. Inlet temperature of cooling liquid is 363 K (89.85 C) and mean outlet temperature of cooling liquid is 370 K (97 C). Inlet temperature of exhaust gases is 857 K (584 C) and mean outlet temperature of exhaust gases is K (98.2 C). The simulation shows that the draft is correct, because outlet temperatures of working substances are very similar to the required parameters. Cooling liquid side required outlet temperature 96.5 C vs outlet temperature from simulation = 97 C. Exhaust gas side required outlet temperature 115 C vs outlet temperature from simulation = 98.2 C. Figure 16: Streamline in exhaust gas heat exchanger 38

40 5. CONCLUSION The aim of my work was based on the input parameters to calculate and design a 3D model of the exhaust gas heat exchanger and create a simulation of the proposed model in COMSOL Multiphysics. The simulations were performed under conditions of the internal combustion engine warmed up to operating temperature. In the next plan will be made simulation under the condition of a cold combustion engine. After verifying the correctness of the draft by created simulations will be made the real model of exhaust gas heat exchanger. Subsequently will be performed the measurement of heat exchanger connected to the real combustion engine for the purpose of validate the correctness of the draft. REFERENCES [1] Holubčík, M.- Hužvár, J.: Jandačka, J.: Combined production of heat and electricity with use of micro cogeneration, IN-TECH 2011 International Conference on Innovative Technologies, year 2011, s , ISBN [2] Nemec, P.: Hužvár, J.: Proposal of heat exchanger in micro-cogeneration unit, configuration with biomass combustion in: Development of materials science in research and education: the nineteenth joint seminar. - [Bratislava: Slovak Society for Industrial Chemistry, 2009]. - ISBN S [3] Blatnický, M., Dižo, J.: Konštrukčný návrh regenerátora Stirlingovho motora. In: Technológ. Roč. 9, Č. 1 (jún), S ISSN ACKNOWLEDGEMENT This contribution is the result of the project implementation: Modern methods of teaching of control and diagnostic systems of engine vehicles, ITMS code , supported by the Operational Programme Educational. The work was supported by the Cultural and Educational Grant Agency of the Ministry of Education of the Slovak Republic in project No. KEGA 077ŽU-4/2017: Modernization of the Vehicles and engines study program. This article was supported by the project VEGA 1/0927/15 Research of the use of alternative fuels and hybrid drives on traction vehicles with aim to reduce fuel consumption and air pollutants production. 39

41 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES COMPARISION OF EMISSIONS FROM CONTINUOUS COMBUSTION OF CONVENTIONAL AND ALTERNATIVE FUELS Ľubomír Miklánek1, Ondřej Gotfrýd2 Abstract This article deals with an experimental research with conventional fuel (diesel) and alternative fuels (Biodiesel and Bioethanol) in framework of sustainable mobility research. Goal of investigation is to compare emissions from continuous combustion of mentioned fuels. As shown measured data, there is an increase of CO and, at the same time, a decrease of NOx in case of Bioethanol fuel compared to diesel fuel, as expected. On the other hand, emissions at Biodiesel fuel are similar to conventional diesel fuel, as expected as well. 1. INTRODUCTION An experimental research in order to investigate the influence of substitute fossil fuels for renewable fuels on exhaust emissions has been carried out in the context of sustainable mobility research work. Specifically, it was a substitute for conventional fuel - diesel fuel (with the addition of FAME (Fatty Acid Methyl Ester from Rapeseed oil) up to 7% by volume, according to Act No. 201/2012 Coll., about air protection, [1]). Biodiesel and Bioethanol has been chosen as fuels from renewable sources. Fuels characteristics are listed in Table 1. In addition, this research was focused on so-called continuous combustion, which is used, for instance, in independent vehicle heaters or in combustion engines with socalled external combustion, unlike the classical I.C.E. The objective of replacing fossil fuel with biofuels is not only to reduce CO 2 emissions, but to contribute to energy self-sufficiency and also to use of biomass [1]. Ľubomír Miklánek, CTU in Prague, Faculty of Mechanical Engineering, Vehicle Centre of Sustainable Mobility, Přílepská 1920, Roztoky, Lubomir.Miklanek@fs.cvut.cz 2 Ondřej Gotfrýd, CTU in Prague, Faculty of Mechanical Engineering, Vehicle Centre of Sustainable Mobility, Přílepská 1920, Roztoky, ondrej.gotfryd@fs.cvut.cz 1 40

42 Task to be solved: To analyze the characteristic of applied fuels and to build assumption for trends in NOx, CO, HC and CO2 emissions when replacing conventional fuels with alternative fuels, additionally by considering a continuous combustion process. Respect the specific emission limits. Chosen approach: To investigate experimentally the above-mentioned assumption for trends using a service analyzer. The continuous combustion ensured by an independent vehicle heater. 2. PROPERTIES OF APPLIED FUELS The specific characteristics of applied fuels are listed in Table 1, see [2] and [3]: Approx. CxH1,9x 14,0 86,0 0 až 0,6 0 0,001 Biodiesel (FAME) Approx. C19H35O2 12,0 77,0 11, ,5 3,5 4,8 1,1 1,2 > ,6 34,3 21,3 42,7 38,5 26,8 kj.kg kg/kg m3/dm3 fuel 14,57 13,2 9,00 9,83 9,44 5,81 C C above 55 above 100 app. 20 Item Chemical formula H C O S Density Viscosity (40 C) Cetane number Hu,v Hu Evaporative heat Lt Lt,V Ignition temperature Flash-point Units %mass %mass %mass %mass kg.m-3 (15 C) mm2.s-1 MJ.dm-3 (25 C) MJ.kg-1 Diesel fuel Bioethanol C2H5OH 13,0 52,2 34,8 - Table 1: Specific characteristics of applied fuels in this research, [2], [3]. Hu- Net Calorific Value (NCV), Lt- Mixing ratio, For a better overview, Table 2 shows the coefficients of carbon c, hydrogen h, oxygen o and sulfur s - for following Diesel fuel without and also with bio-content and other fuels. In addition, the Gasoline and Methane fuels are also listed for comparison. 41

43 Diesel Diesel+6.5% of FAME FAME Bioethanol Gasoline Methane c h o s Table 2: Constitutional coefficients of all considered fuels, [2] and [4] The above values of coefficients of considered fuels show clearly that they differ from each other, so it is meaningful to calculate the theoretical value of the molar fraction of CO2 in the so-called dry exhaust gases when the fuel is burned completely. The resulting values are shown in the graph in Figure 1. Figure 1: Comparison of the max. theoretical values of the molar fractions of CO 2 of specific fuels The CO2MAX theoretical value was calculated according to the formula shown in [5]: cco 2,MAX c (1) h o h o 4.76 c A minimal difference can be observed between Diesel fuel without and with 6.5% of FAME based on comparison of CO2MAX theoret. values. The theoretical max. value of CO2 for Bioethanol is also very close to that of Diesel fuel. Conversely, FAME shows larger difference in max. theoret. value CO2 compared to Diesel fuel. It is well known that the CO2 value has a dominant effect on the result in the computation of air-excess coefficient ( ) based on components in dry exhaust gas, e.g. using Brettschneider method, [5]. The calculation formula for is usually stored in memory of the service gas analyzer. 42

44 Based on this fact it is not possible to use just one formula with coefficients for one fuel to calculate lambda for all three types of fuels (Diesel, Biodiesel and Bioethanol) if we want to have precise results. For demonstration, Figure 1 shows the theoretical values of CO 2 for methane which is the dominant component of nature gas. The low value of CO 2 is the reason why nature gas is considered as a suitable alternative fuel to reduce CO 2 emissions in automotive. Moreover, the Table 1 shows also the heat values of considered fuels per volume (Hu,v), and the lowest heat value has the fuel: Bioethanol. Obviously, to achieve the same power (energy), more Bioethanol has to be burned (up to 70% more than Diesel fuel). The Bioethanol-Air stoichiometric mixture has the lowest heat value of the above mentioned fuels as well, see Figure 2. Figure 2: Heat values comparison of stoichiometric mixtures of specific fuels with air In addition, the mixing ratio (Lt, v) given in Table 1 shows that the stoichiometric mixture of Bioethanol has the largest ratio of fuel in 1m3 of the mixture, see Figure 3. Figure 3: Comparison of the ratio of specific fuels in the 1m3 of stoichiometric mixture 43

45 3. CONTINUOUS COMBUSTION EQUIPMENT A fuel-operated air-heater (FOAH), producer co. Brano Group, [6], was applied to reach the continuous combustion. To understand clearly the operation principle of the diesel FOAH, the Figure 4 shows a view inside a conventional diesel heater, [7]. Figure 4: View inside the conventional diesel FOAH, [7] In principle, the fuel is burned continuously in the combustion chamber. The exhaust gas flows through the heat exchanger and heats it up. Air flows on the other side of the exchanger and is heated up. Heated air comes out from the heater and flows into the target area (e.g. crew compartment). More details about the operating principle of the diesel FOAH is presented in work [8]. A sample of the diesel FOAH was at disposal thank to the partner [6]. Therefore, authors research was carried out on this type of the air-heater. To understand better the process of continuous combustion in this type of air heater, the combustion chamber is described in more details. The design of the combustion chamber allows continuous burning of hardly vaporizable fuel which cannot be ignited by electric discharge at room temperature and atmospheric pressure. A heat source is a glow plug that delivers the needed heat during starting the heater. The plug heats a screen (refractory material) placed around the plug. The fuel is not injected but flows under low pressure on the screen surface. The fuel evaporates on the hot surface of the screen. Then fuel vapors are mixed with air and mixture is ignited by heat emitting from the hot screen. Both the screen and the combustion chamber are heated by burning the fuel. Once the chamber is warm enough, the plug doesn t need to supply heat anymore and shuts off. The detail of the combustion chamber is in Figure 5, [9]. As it can be seen on the Figure 5, heat is not intensively conducted from the walls of the chamber in to the cooling system as in I.C.E. This is the reason for different exhaust emissions, mainly the components HC and CO. The assumption is for lower content of these components in comparison to the I.C.E. 44

46 Fuel inlet Air inlet Exhaust Figure 5: Detail of combustion chamber of the diesel FOAH. 8-Glow plug screen, 9-Glow plug, 10- Glow plug housing, 11- Flame sensor, 17- Combustion chamber, 18- Combustion chamber seals, [9] 4. EXPERIMENTAL RESEARCH 4.1 Test bench setup Heater Control PC Emission analyzer Sample of exhaust gas Tested fuel Fuel metering pump Figure 6: Test bench setup of diesel FOAH, control PC and emission analyzer A sample of exhaust pipe of the heating system was analyzed using the Combined emission station Actigas 605 (AT ). This station includes an AT 505 analyzer to measure following emission components [10]: - CO, CO2, HC(hexane) using nondispersive infrared analyzers, - O2 using an oxygen sensor; - NOx using an electrochemical sensor. 45

47 The measured data of the molar fractions of the above components in the dry exhaust gases were continuously recorded and stored into a file with a frequency of approx. 0.6 Hz. 4.2 Measured data Emission measurements were carried out with FOAH in four different power stages, marked as: P0, P3, P6 and P9 (corresponding to the heating power: 25%, 50%, 75% and 100%) for all three fuels (diesel, biodiesel and bioethanol). The measured emission values are shown in Figures 7, 8 and 9. Emission station is also used to calculate the air-excess coefficient ( ). The calculation is performed using the Brettschneider method with the coefficients for gasoline, see Table 2. The air-excess values displayed by the emission station during the measurements were therefore considered as indicative only. The air-excess values for specific fuel are additionally calculated using the data in Table 2. The results are shown in Figure 10. Differences between measured and calculated values are shown in Figure 11. Figure 7: Measured molar fractions of CO for all three fuels, (emission limit: 0,1%vol.) Figure 8: Measured molar fractions of NOx for all three fuels, (emission limit: 200 ppm) 46

48 Figure 9: Measured molar fractions of CO2 for all three fuels Figure 10: Calculated air-excess values ( ) for all three fuels Figure 11: Differences between measured and calculated air-excess values for specific fuel 47

49 It is necessary to point out that the production of HC (hexane) emissions was very low in all measurements and only zero values were measured (emission limit: 100 ppm). Therefore molar fractions HC are not shown in any figure. Fuel consumption was not measured during these experiments, as the objective was to determine the effect of the fuel type on the exhaust gas emissions. 4.3 Comparison of assumed and measured trends in emission A comparison of assumed and measured emissions is summarized in Table 3. Replacement by Bioethanol Replacement by Biodiesel (FAME) Assumption Measured CO decrease (due to content CO according to assumption of O2), NOx roughly the same or NOx according to assumption small decrease (due to lower Hu compared to diesel), HC according to assumption HC very small, similar to diesel fuel. CO increase (due to smaller Hu mixture need to be enriched, more fuel enters the chamber) CO according to assumption ONLY in higher power output (P6, P9). Otherwise decrease. NOx decrease (due to smaller Hu compared to diesel), NOx according to assumption HC very small, similar to diesel fuel. HC according to assumption Table 3: Comparison of assumed and measured trends in emissions by use of Biodiesel(FAME) and Bioethanol instead of Diesel fuel 3. CONCLUSION Experimental investigation of emissions behaviour has been performed for the case that alternative fuels - Biodiesel and Bioethanol are applied instead of the conventional one - diesel. Properties of both alternative fuels have been studied. Based on this, constitutional coefficients have been determined to calculate the air-excess for each type of applied fuel. A continual combustion has been accomplished using diesel fuel-operated air-heater (FOAH). The assumption of emission production for Biodiesel has been confirmed in all power modes (CO decrease, NOx very small decrease in comparison to the Diesel fuel). On the other hand the assumption of emission production for Bioethanol was not confirmed in low power modes in which it is not necessary to enrich the mixture so much. Conversely, the assumption was confirmed for high power modes in which the mixture needs to be enriched CO increase, NOx decrease. The results show that the reduction of increased CO emissions could be possible by optimisation of both the combustion chamber and the combustion process. 48

50 Last but not least the experiment showed difference in air-excess values measured by service analyser (dedicated for gasoline) and the calculated values for each fuel. The largest difference is in case of Bioethanol: calculated value of air-excess is smaller (-) than measured value about 1,5 to 2%. The smallest difference is for FAME: about + 0,5%. REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] Ministerstvo průmyslu a obchodu. Národní akční plán České republiky pro energii z obnovitelných zdrojů, Srpen, 2012, pp MATĚJOVSKÝ V. Automobilová paliva, Grada Publishing, a.s., Praha, 2005, pp ISBN MACEK J. Spalovací Motory, Vydavatelství ČVUT v Praze, Praha, 2012, pp ISBN BLAŽEK J., MAXA D., ŠIMÁČEK P., et al. Emissions of Organic Compounds from the Skoda 1.4 Petrol Engine, Sborník abstraktů XXXVII. KOKA 2006, Praha, 2006, pp. 37, ISBN TAKÁTS M. Měření emisí spalovacích motorů, Vydavatelství ČVUT v Praze, Praha, 1997, pp ISBN BRANO GROUP, a.s., Eberspächer Climate Control Systems GmbH MIKLÁNEK, Ľ., GOTFRÝD, O. Study of Exhaust Emissions Reduction of a Diesel Fuel Operated Heater During Transient Mode of Operation. MECCA, 2014, Vol. 12, No. 1, p ISSN MV Heating Ltd ACTIA CZ s.r.o ACKNOWLEDGEMENT This research has been realized using the support of The Ministry of Education, Youth and Sports program NPU I (LO), project # LO1311 Development of Vehicle Centre of Sustainable Mobility. This support is gratefully acknowledged. This research has been realized using the support of Technological Agency, Czech Republic, programme Centres of Competence, project # TE Josef Božek Competence Centre for Automotive Industry. This support is gratefully acknowledged. This research has been realized using the support of EU Regional Development Fund in OP R&D for Innovations (OP VaVpI) and Ministry for Education, Czech Republic, project # CZ.1.05/2.1.00/ Acquisition of Technology for Vehicle Center of Sustainable Mobility. This support is gratefully acknowledged. This research has been realized also using the support of BRANO GROUP, a.s., thanks for providing their independent air-heating system. 49

51 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES FUEL INJECTION IN ENGINE SUCTION DUCT IN NON-STATIONARY MODES Celestýn Scholz1, Aleš Dittrich 2 Abstract Optical methods are an ideal tool in the field of combustion engine research. The main advantage is to provide information about activities that are hidden to the human eye, thus enabling you to understand the context of ongoing events. The paper deals with the research of a methodology suitable for assessing the course of fuel injection into a suction duct of a piston engine in non-stationary transition modes. 1 ÚVOD Optické metody jsou ideálním nástrojem v oblastech výzkumu spalovacích motorů. Hlavní výhodou je podávání informací o činnostech, které jsou lidskému oku skryté a umožní tak pochopit souvislosti probíhajících dějů. Zkušenosti získané při používání těchto metod slouží k optimalizačním činnostem sledovaných součástí za účelem zlepšení jejich využití. Snahou konstruktérů při vývoji pístových spalovacích motorů je například optimalizace proudových polí v sacích kanálech a ve válci motoru pro získání nejvyšší možné účinnosti. Vizualizace dějů uvnitř spalovacího motoru je velmi důležitý a významný pomocník při optimalizaci tvorby směsi a hoření. Cílem tohoto příspěvku je výzkum metodiky vhodné pro posouzení průběhu výstřiku paliva do sacího kanálu pístového motoru při nestacionárních přechodových režimech. Metodika by využívala stávající vizualizační, indikační a emisní techniku. Při rychlých změnách otáček a zatížení, které nastávají zejména při akceleraci (nebo deceleraci), se provádí u moderních motorů kompenzace okamžitého nedostatku (nebo přebytku) vstříknutého množství paliva. 2 EXPERIMENT K měření byl použit vozidlový zážehový motor EA211 se standartní EŘJ, umístěný v Laboratoři pohonných jednotek Ústavu pro nanomateriály, pokročilé technologie 1 2 prof. Ing. Celestýn Scholz, Ph.D., Technická univerzita v Liberci, celestyn.scholz@tul.cz Ing. Aleš Dittrich, Technická univerzita v Liberci, ales.dittrich@tul.cz 50

52 a inovace Technické univerzity v Liberci na zkušebním stanovišti s vířivým dynamometrem Schenck WT 190. Zážehový, 16 ventilů, DOHC 76,5 X 86,9 mm cm3 80 kw 152 Nm 10,5 : 1 Kapalinou Typ Vrtání X Zdvih Počet válců Zdvihový objem Maximální výkon Maximální točivý moment Kompresní poměr Chlazení Tabulka 1: Parametry použitého motoru. 2.1 Konstrukční úprava K zavedení endoskopu pro vizualizaci vstřikování paliva do sacího potrubí bylo nutné konstrukčně upravit standartní sání motoru. Požadavky na umístění nechlazeného endoskopu jsou zejména umístění co nejblíže sacím ventilům a zároveň co nejblíže ose sacího kanálu. Z tohoto důvodu bylo použito sání pro dvoupalivové provedení daného motoru, kdy do otvoru, určeného pro umístění vefukovače CNG, byla vložena navržená vložka endoskopu. Příslušný motor byl uzpůsoben pro vizualizaci Obr. 1: Příprava úprav pro umístění nechlazeného endoskopu, zobrazení zorných kuželů 3D model. 2.2 Použitá měřicí technika Brzdové stanoviště Zkušební stanoviště s vířivým dynamometrem Schenck WT 190 je vybaveno řídicím a měřícím systémem pracujícím pod OS Windows s plně automatizovaným sběrem dat (ve frekvenci 1Hz) na pravidelně kalibrovaných 32 kanálech. Součástí zkušebního stanoviště je teplotní stabilizace provozních médií motoru (voda, olej, palivo a vzduch), pro zážehový motor je na stanovišti přívod 2 druhů benzinů (BN95 a BN98). 51

53 Maximální výkon Maximální točivý moment 190 kw 600 Nm Maximální otáčky min-1 Tabulka 2: Parametry vířivého dynamometru Schenck WT Měření a vyhodnocování tlaku ve válci Měření průběhu tlaku ve válci bylo prováděno pomocí indikační aparatury AVL Indimeter 619 (umožňující sledovat průběh měřených parametrů on-line) sestávající ze zesilovače náboje AVL3066A03, snímače otáček a polohy klikového hřídele 364C a snímačů tlaků GU 21D (umístěných na každé válcové jednotce). Pro určení pozice přeskoku jiskry na zapalovací svíčce a doby a polohy vstřikování paliva (oboje umístěno na příslušenství první válcové jednotky) bylo použito proudových kleští Kistler 2105A30 včetně příslušenství. Vyhodnocování bylo provedeno v SW AVL Concerto 4.6. TRIGGER CDM AVL VISIOSCOPE Kistler 2105A30 Snímače tlaků GU 21D Obr. 2: Schéma zapojení zařízení pro analýzu spalovacích tlaků Indimeter Vizualizační technika Vizualizační systém Visioscope AVL pracuje s elektronickým snímáním, přenosem a zpracováním obrazu tak, že uživatelem zvolenou frekvencí se postupně zaznamenává obraz ve válci motoru v jednotlivých polohách klikového hřídele. Výsledný průběh sledovaného děje je tedy složen z obrazů zaznamenaných pro zvolenou oblast z různých pracovních cyklů motoru. Pro všechna měření byl použit nechlazený endoskop (který byl optickým vedením připojen na jednotku zdroje světla) a barevná VGA kamera PCO PixelFly. Pro záznam a vyhodnocení obrazového záznamu bylo využito nejnovější verze SW AVL Visioscope

54 CDM AVL INDIMETER TRIGGER 1. CCD kamera 2. monitor 3. jednotka zdroje světla 4. propojovací kabel zdroje světla 5. kabel monitoru 6. kabel kamery 7. počítač 8. propojovací kabely k Indimeteru 9. optické vlákno pro osvětlení Obr. 3: Schéma zapojení zařízení pro vizualizaci Visioscope Emisní aparatura Měřicí, plně automatická, aparatura Horiba Mexa-One je uzpůsobena pro odběr vzorku horkých výfukových plynů a následně umožňuje analýzu odebíraného vzorku on-line. Sběr dat probíhá ve frekvenci 1Hz a 10Hz. Vzorek výfukových plynů byl odebírán z výfukového potrubí před katalyzátorem a veden vyhřívaným potrubím do sestavy analyzátorů k on-line určení molárních zlomků (objemových koncentrací) výfukových škodlivin a dalších složek ve výfukových plynech. CO(L) CO(H) CO2 O2 MOTOR NO NOx THC CH4 MEXA-ONE KATALYZÁTOR OVN Obr. 4: Schéma odběru vzorku z výfukových plynů. 2.3 Program měření Provozní režim motoru při měření byl zvolen tak, aby došlo k vyvolání nestacionárních vlastností vyvolaných prudkou změnou otáček a zatížení. Měřicí cyklus spočíval v provozu ve volnoběžných otáčkách, následně došlo k prudké změně otáček a zatížení (s krátkou dobou setrvání) a poté nastal opět přechod do volnoběžných otáček. Ve všech provozních režimech motoru byla prováděna vysokotlaká indikace 53

55 pomocí měřicí techniky AVL Indicom, záznam pomocí AVL Visioscope a rovněž byl odebírán vzorek výfukových plynů do zařízení Horiba Mexa-One. Údaje o seřízení motoru v jednotlivých provozních režimech byly sledovány pomocí lambda-měřiče ETAS (širokopásmová λ sonda) a výpisem dat z řídící jednotky pomocí diagnostického SW DiagRA-D s připojením diagnostického Interface Basic+AC. Obr. 5: Zobrazení měřicího cyklu (otáčky, klapka) v závislosti na čase. Nechlazený endoskop Optický světlovod Stativ PCO Camera Obr. 6: Pohled na měřicí stanoviště. 2.4 Výsledky a diskuse Pro vyhodnocení výsledků bylo provedeno nezbytné časové sladění 4 souborů dat zaznamenaných rozdílnou technikou s odlišnou frekvencí: Brzda WT 190 Schenck (CMS, Diagra) 1 Hz IndiCom AVL (Concerto) 0,5 CA (10 50 khz) VisioScope AVL (snímky) Hz MEXA-ONE Horiba (plynné emise). 10 Hz 54

56 2.4.1 Průběh akcelerace Akcelerace byla vyvolána rychlou změnou polohy plynového pedálu a zátěže z 0% na 100% v čase 1 sec, z volnoběhu 800 rpm na 100% zátěže při 4500 rpm. 1a 1b 1c 1961 rpm, IMEP 10,29 bar 2a 2b 2c 2785 rpm, IMEP 10,25 bar 3a 3b 800 rpm, IMEP 1,6 bar Obr. 7: Graf průběhů parametrů začátku a konce vstřikování paliva, otáček motoru, poloh plynového pedálu, škrticí klapky při akceleraci. Graf je doplněn obrázky výstřiku paliva při 3 úrovních (1,2,3) otáček motoru a okamžité polohy v cyklu (a, b, c). 55

57 Obr. 8: Graf průběhů indikovaných parametrů spalovacího procesu, předstihu zážehu a zatížení motoru, při akceleraci. Na obrázcích 7 a 8 je patrný pomalý nárůst otáček a středního efektivního tlaku motoru a naopak rychlý nárůst indikovaných parametrů tlaků ve válci motoru. Lze pozorovat fáze výstřiků paliva do společného sacího kanálu dvou ventilů. Na začátku akcelerace (úroveň 3) je posílen nedostatek paliva dvojitým výstřikem paliva Průběh decelerace Decelerace byla vyvolána rychlou změnou polohy plynového pedálu a zátěže ze 100% na 0% v čase 1 sec, z 100% zátěže při 4500 rpm na volnoběh 800 rpm. 56

58 rpm, IMEP 12,25 bar rpm, IMEP 7,05 bar rpm, IMEP 2,2 bar rpm, IMEP 0 bar 57

59 rpm, IMEP 0 bar rpm, IMEP -0,8 bar Obr. 9: Graf průběhů parametrů začátku a konce vstřikování paliva, otáček motoru, poloh plynového pedálu, škrticí klapky při deceleraci. Graf je doplněn obrázky výstřiku paliva při otáčkách motoru a okamžité poloze v cyklu. Obr. 10: Graf průběhů indikovaných parametrů spalovacího procesu, předstihu zážehu a zatížení motoru při deceleraci. Na obrázcích 9 a 10 je patrný pomalý pokles otáček a středního efektivního tlaku motoru a naopak rychlý pokles indikovaných parametrů tlaků ve válci motoru. Lze pozorovat fáze výstřiků paliva do společného sacího kanálu dvou ventilů. Od začátku decelerace se snižuje dávka paliva délkou trvání výstřiku. Nakonec v rozmezí otáček od 3000 do 1600 rpm dojde 58

60 k úplné absenci výstřiků, čímž se předchází k výskytu přebytku paliva a pravděpodobně k vzniku smáčení stěn kanálu (tzv. wall wettig) Koncentrace plynných emisí Vývin plynných emisí je v obrázku 11 znázorněn v průběhu jejich koncentrací během celého měřícího cyklu, tj akceleraci a následné deceleraci. Obr. 11: Graf průběhů koncentrací plynných emisí v průběhu měřícího cyklu. 3 ZÁVĚR Práce popisuje experimenty na nepřeplňovaném benzinovém motoru s palivovým systémem MPI s masivním použitím měřící a výpočetní techniky. Byla vyvinuta metodika vhodná k posouzení průběhu výstřiku paliva do sacího kanálu pístového motoru při nestacionárních přechodových režimech. Použitý motor disponoval řídící jednotkou s vhodným SW, který vhodně reaguje na nestacionární režimy rychlé akcelerace a decelerace. Metodiku založenou na vizualizaci výstřiku paliva a souběžné indikaci časování výstřiku, počátku otvírání sacích ventilů a průběhu tlaků ve válci a měření emisí pracoviště TUL využívá k dalším optimalizačním pracím. Ty již využívají systém řízením parametrů vstřikování pomocí univerzální řídící jednotky EURO 4. LITERATURA [1] [2] [3] SCHOLZ, C., BLAŽEK, J., DROZDA, H.: Vizualizace procesu spalování a vstřikování paliva, KOKA 2002, XXXIII. Mezinárodní konference kateder a pracovišť spalovacích motorů českých a slovenských škol, Slovenská pol n ohospodárska univerzita v Nitre, Ráčkova dolina, 2002, ISBN BEROUN, S., BLAŽEK, J., SCHOLZ, C.: Výzkum metod vhodného optického přístupu do vnitřních prostor pístového spalovacího motoru (etapa E , řešení 2005, 2006). Projekt 1M0568, výzkumná zpráva SM 552/2006, KVM FS TU v Liberci, 19 str., BLAŽEK, J.: Vizualizace dějů uvnitř spalovacího motoru. Studijní materiál, dostupné online: 59

61 [4] ČECH, K.: Vizualizace vstřikování paliva do sacího potrubí pro výzkum zážehových motorů, Bakalářská práce, TUL PODĚKOVÁNÍ The results of this project LO1201 were obtained through the financial support of the Ministry of Education, Youth and Sports in the framework of the targeted support of the National Programme for Sustainability I and the OPR&DI project Centre for Nanomaterials, Advanced Technologies and Innovation CZ.1.05/2.1.00/ and the project Technological Agency, Czech Republic, programme Centres of Competence, project # TE Josef Božek Competence Centre for Automotive Industry 60

62 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES INCREASING THE MECHANICAL EFFICIENCY OF TURBOCHARGERS Pavel Novotný1, Petr Škara2 Abstract The paper describes the modern trends in a development of turbochargers with a focus on mechanical efficiency of bearing system and a reduction of gas blow-by thru the sealing system. There are also briefly presented descriptions of the principles and a subsequent applications of computational modelling approaches. The capabilities of the innovative computational models and sample results are demonstrated on a middle size turbocharger. The developed simulation tool uses multibody software extended by user-written subroutines and specialized programs developed by authors. Main results are verified by technical experiments. 1. INTRODUCTION Today s trends as downsizing, downspeeding, exhaust gas turbocharging or engines with a low number of cylinders are modern approaches for the powertrain design. Utilization of turbocharging technology presents effective way how to meet the requested demands. On the other hand, the wear or noise and vibration issues become more important for turbocharger producers. One of the greatest challenges of the turbocharging technology are its very demanding operating conditions including high revolution speeds, high temperatures and unsteady-state mass flow rates of the intake air and exhaust gas. Modern turbochargers operate in speeds up to the min -1 and temperatures of rotor components in hundreds of degrees of Celsius. The turbocharger issues can be effectively solved by computational approaches on different system level. Turbocharger rotor dynamics and its modelling are historically based on simple linear models, but this solution is not sufficient. A breakthrough was introduced by Schweizer [1], who used multibody dynamics software (MBS) based tool for turbocharger rotor dynamics prediction. This type of a model is the most commonly used model for turbocharger rotor dynamics simulation, similar models present Knoll [2] for example. The computational models can be also used for analyses of noise and 1 2 Pavel Novotný, Brno University of Technology, Brno, Technická 2896/2, novotny.pa@fme.vutbr.cz Petr Škara, Honeywell, spol. s r. o., Brno, Tuřanka 96/1236, petr.skara@honeywell.com 61

63 vibration as presents Wolf [3]. For example, Tian [4] used this type of a model to create a model to study influence of the rotor unbalance and engine vibration on turbocharger rotor dynamics. 2. AIMS OF THE SIMULATION MODELLING A computational model of the turbocharger rotor including a bearing system being presented in this paper is solved in time domain considering positives and negatives of transient nonlinear dynamics. This approach enables to incorporate different physical problems, including various nonlinearities, but sometimes leads to long solution times. The solution speed is always critical for turbocharger developers and thus it is defined as main condition which has to be considered in computational model development. An efficient and robust computational model capable to solve the turbocharger rotordynamics is the main aim of this research work. The computational model being presented has to consider: Non-linear behaviour of rotor dynamics. Non-linear and temperature dependent behaviour of hydrodynamic bearings. Model robustness and solution effectiveness. Direct calculation of tribology results. Direct calculation of NVH results. Capabilities for assembling into large powertrain models. 3. SIMULATION MODELLING BACKGROUND The model is assembled, as well as, numerically solved in multibody system ADAMS. ADAMS is a general code and enables an integration of user-defined models directly using ADAMS commands or using user written FORTRAN or C++ subroutines. The computational model includes all significant components necessary for rotordynamics solution. Rotor shaft, compressor and turbine wheels and hydrodynamics floating bearings are the main parts of the model. The rotor shaft is modelled using flexible body based on discretised FEM (Finite Element Method) principles. Figure 1 Floating ring bearing model arrangement used 62

64 The turbocharger rotor incorporates semi floating ring bearing to increase damping effects of common hydrodynamics bearings. Since the journal bearing is represented in the MBS by a set of pre-calculated databases, several presumptions [5] have to be considered to enable a creation of the databases in advance. The scheme and general dimensions of fully floating ring bearing is shown in Figure 1. The bearing consists of three parts housing bore (sleeve), shaft and floating ring. The floating ring separates shaft from housing bore, and therefore, two oil films are created inner and outer oil film. For precise simulation of rotor dynamics all bearing parameters have to be known bearing diameter, width, bore dimensions and clearances for both inner and outer oil film, oil parameters dynamic viscosity, oil input pressure and oil temperature. Both oil films are treated separately as a plain journal bearing. The hydrodynamic pressure and consequential forces and moments are solved numerically by Reynolds equation. The equation is based on the modification of the Navier-Stokes equation and continuity equation transformed for cylindrical shapes of the bearing oil gap. The full form of the equation has been simplified and modified based on work of Novotny [6]. The Reynolds differential equation is discretised by finite difference method and iteratively solved by Gauss-Seidel method employing successive over-relaxation strategy (SOR), principles can be found in [7]. The cavitation condition ensures that the pressure in the oil film is always higher or equal to cavitation pressure as presents Novotny [6] or Marsalek [8]. The dimensionless forces and moments calculated from hydrodynamics pressure distribution are stored in hydrodynamics databases and they are read during dynamics solution. The transition to the dimensionless form enables to solve the equation effectively and also to solve the equation for groups of bearing (depending only on individual bearing dimensions). A geometric visualisation of the turbocharger rotor in Multibody dynamic software is presented in Figure 2. Figure 2 Multibody dynamics model of turbocharger rotor 63

65 4. RESULT EXAMPLES 4.1 Verification of Simulation Modelling Approach In general, a verification of computational methods in the case of turbocharger rotordynamics is highly difficult. High rotational speeds, small dimensions or high temperatures are the reasons why only a few experimental approaches can be used. It is very important to mention, that there are a lot of input parameters influencing the rotordynamics and these parameters have to be very carefully considered. Rotor unbalance, bearing clearances, temperature deformations and angular positions of unbalances are very important to correctly set-up. Table 1 Selected operating data of turbocharger rotor during revolution sweep simulation Maximal rotor speed Time of revolution sweep Lubricating oil (SAE specification) Input oil temperature Input oil pressure rpm 5s 15W C 0.5 MPa Revolution sweep from minimal to maximal rotor speed during time period is the operating regime being simulated and measured. Table 1 presents selected operating data of the turbocharger rotor during the revolution sweep simulation. The computational model is verified by a measurement of compressor nose movement in two directions. Figure 3 presents the peak-to-peak values of displacement magnitude of a compressor nose vs. rotor speed defined as ep 2 p 2 ex2 ex2, (1) Amplitude [mm] where ex and ey are eccentricities of the compressor nose in x direction and y direction, respectively, considering the fact, that z is a coordinate in rotor axis Measurement Calculation Place of measurement Rotor speed [min ] Figure 3 Peak-to-peak values of displacement magnitude of compressor nose v. rotor speed The rotor shows stable behaviour only for speeds below approx. n = min-1. Instabilities in outer oil film of radial bearing start above this revolution speed, the 64

66 amplitudes of vibrations steeply increase and only high nonlinearity of bearing stiffness restricts the turbine or compressor wheels from touching the housing and causing serious damage. The model shows two peaks corresponding to rotor bending natural frequencies around speeds n = min-1 and n = min-1. These peaks are not clearly seen in the measurements. The processes in turbine and compressor chamber during rotation are the reasons why the measured vibrations are more damped. Bearing tribology, noise and vibrations are the results being rigorously reviewed during development process of turbocharger and they will be mentioned more in detail. 4.2 Analysis of Radial Bearing Tribology Tribology becomes more important in automotive turbochargers in terms of synthetic lubricating oils, friction reduction, adhesion and abrasion friction and wear reduction in the oil-film bearings including journal and thrust bearings. Typical analyses of radial bearing tribology start reviewing basic results like relative eccentricities (a ratio of eccentricity and bearing clearance), ratios of minimal oil film thickness and combined surface roughness, maximal oil film temperatures and maximal pressures in oil film. Relative eccentricity of pin in shell is a basic quantity for every bearing in development process. The optimal value of relative pin eccentricity is roughly ~ 0.7 [9] then the hydrodynamic theory presumes an optimal value of load capacity to friction moment ratio. Figure 4 on the left shows relative eccentricities for inner and outer oil films for bearings on turbine and compressor sides for the operating speed range of the rotor. A ratio of minimal oil film thickness and combined surface roughness can be used as an indicator of increased wear of pin or shell surface. The ratio is defined as h hmin s2 p2, (2) where hmin is minimal oil film height, s and p is arithmetic mean roughness of shell and pin, respectively. Figure 4 Relative eccentricities (left) and ratio of minimal oil film thickness and combined surface roughness (right) of inner and outer oil films for bearings on turbine and compressor sides 65

67 The values of hmin and combined surface roughness ratios for inner and outer oil films for bearings on turbine and compressor sides are presented in Figure 4 on the right. Oil film temperature presents a significant impact on rotor behaviour via modification of oil dynamic viscosity and thermal deformations of components. The influence of temperature is particularly higher for the inner oil film due to the lower oil flow. At the same time, the turbine side is more thermally loaded, hence the heat transfer to the oil is highest in the inner oil film of the turbine side bearing. At the same time, the increased temperature of turbocharger components influences the thermal deformations of the bearing parts and thus produce the changes in the bearing clearance. As an example, the maximal oil film temperatures of inner and outer oil films for bearing on compressor side are shown in Figure 5. Figure 5 Maximal oil film temperatures of inner and outer oil films for bearing on compressor side 4.3 Analysis of Thrust Bearing Design The axial bearing represents greater impact on mechanical losses and oil flow requirements compared to radial bearings. The basic design parameters of the bearing require a detailed analysis of their effect on the measured quantities. The shape, position, or number of bearing surfaces is important in terms of bearing capacity to reduce frictional losses. A typical example is the use of low viscosity synthetic oils. Synthetic oils used in the automotive engines result from combining PAOs (poly-alpha-olefins) with about 15 % of a synthetic ester. The oils are commonly describer by HTHS viscosity (high temperature high shear) defined as the effective oil dynamic viscosity in the operating condition at the high oil temperature of 150 C and large share rate of 106 s 1. The expected reduction in frictional losses, however, results in a reduction in bearing load capacity. Therefore, to maintain the expected load capacity, it is necessary to modify the design parameters of the bearing. This can be achieved by, for example, changing the diameter or changing the shape of running surface. Figure 6 shows bearing integral results for two lube oils for prescribed minimal oil film thickness. The bearing design parameter with highest sensitivity is the bearing width B (or difference between inner and outer bearing diameters) and this parameter introduces an effective way to increase load capacity. 66

68 Figure 6 Comparison bearing integral results of base bearing design with different lubricating oils Figure 7 presents the comparison of relative power losses of new bearing design using lube oil 5W20 with standard bearing design using lube oil 15W40. Figure 7 Relative power loss comparison of base bearing design with different lubricating oils 67

69 5. CONCLUSION Modern trends in the development of turbochargers lead to the use of low-viscosity highly additive oils. There is also significant pressure on changes of turbocharger operating conditions, for example by increasing the oil temperature when entering the turbocharger or reducing the input oil pressure and thus also the necessary oil flow. An increase in the mechanical efficiency of the turbochargers together with reducing the energy demands of the accessories is the results, but often at the cost of increasing wear or noise. These contradictory requirements increase the design requirements as well as a detailed understanding of the dynamic process. Proposed computational models using strategies of multibody dynamics with embedded user functions allow a deeper understanding of the principles of dynamics and tribology of turbocharger rotors and bearing systems. These models are primarily designed for computational efficiency leading to relative short calculation times and enabling parametric studies. An important advantage is also the incorporation of analytical equations and thus the comprehensible presentation of the influences of individual parameters for the turbocharger designers. ACKNOWLEDGEMENTS The research leading to these results has received funding from Technology Agency of the Czech Republic, programme Competence Centres, project Josef Bozek Competence Centre for Automotive Industry, project No. TE and with help of the project FSI-S granted by specific university research of Brno University of Technology. The authors gratefully acknowledge this support. REFERENCES [1] SCHWEIZER B., RIEGER N. F., BLACK H. F., THOMAS CH. B. Dynamics and stability of turbocharger rotors. Archive of Applied Mechanics, 80(9), 2010, p , ISSN [2] KNOLL G., SEEMANN W., PROPPE C., KOCH R. Run-up of Turbocharger Rotors in Nonlinearly Modelled Floating Bush Bearings, MTZ worldwide, Issue No. 04, [3] WOLFF K., STEFFENS CH., AYMANNS R., STOHR R., PISCHINGER S. Turbo Charger Noise - Development of a Methodology for the Acoustic Turbo Charger Layout. FVV research report , [4] TIAN L., WANG W. J., PENG Z. J. Dynamic behaviours of a full floating ring bearing supported turbocharger rotor with engine excitation. Journal of Sound and Vibration, 330(20), 2011, p , ISSN X [5] BUTENSCHÖN H., J. Das hydrodynamische, zylindrische Gleitlager endlicher Breite unter instationärer Belastung, Ph.D. Thesis. Universität Karlsruhe, [6] NOVOTNY P. Virtual Engine A Tool for Powertrain Development. Inaugural Dissertation, Brno University of Technology,

70 [7] NOVOTNÝ P., MARŠÁLEK O., RAFFAI P., DLUGOŠ J., KNOTEK J. Mixed Lubrication Solution with Consideration of Elastic Deformations and Real Surface Roughness Structures. Journal of the Balkan Tribological Association. Book 4, Vol. 22. ISSN , [8] MARŠÁLEK O. Advanced Methods for the Solution of Journal Bearing Dynamics. Dissertation, Brno University of Technology, [9] STACHOWIAK G. W., BATCHELOR, A. W. Engineering Tribology. Fourth edition. Boston: Elsevier Butterworth-Heinemann, ISBN

71 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES ELECTRICAL TURBOCHARGING THERMODYNAMIC POTENTIAL UNDER STEADY STATE OPERATION Oldřich Vítek 1, Jan Macek 2 Abstract The paper deals with comparison of classical turbocharging with electrical one under steady state operation for the case of automotive 3-cylinder engine. The concept of electrical turbocharger is represented by either a compressor/turbine when each device has its own electric machine or a compressor connected to a turbine by means of a shaft while there is an electrical machine connected to the turbocharger shaft. The results were obtained by means of simulation (0-D/1-D CFD approach) while all considered cases were fully optimized using genetic algorithm considered engine load range covers medium, high and full load. 1. INTRODUCTION Due to recent development of hybrid powertrains, the availability of high power electrical energy in a vehicle has increased significantly. This allows applying boosting systems which are driven/assisted by high-speed electrical motors such systems are usually called e-boosting systems. Application of such systems allows for increased hybridisation level due to the fact that exhaust gas energy can be harvested/recuperated and then stored in on-board storage device(s) of electric energy. Simulation approach is a good way to evaluate theoretical performance of a new design. Hence, it makes sense to evaluate thermodynamic potential of e-boosting system, which is the main focus of the paper. There are many possible layouts of a boosting system while using e-boost devices. The considered layout is shown in Figure Chyba! Nenalezen zdroj odkazů.. The main goal of the paper was to estimate upper limits of achievable efficiency for the case of highly flexible ICE for passenger cars while considering electrical turbocharging concept. Additional goals are the following to compare electrical turbocharging concept(s) with classical turbocharging one, to considered different 1 2 Oldřich Vítek, CTU in Prague, Technická 4, Prague 6, oldrich.vitek@fs.cvut.cz Jan Macek, CTU in Prague, Technická 4, Prague 6, jan.macek@fs.cvut.cz 70

72 combustion modes (SI, CI, RCCI), to perform sensitivity studies of selected parameters (e.g., efficiency of e-machine, HP exhaust volume). 2. ENGINE MODEL The target engine parameters are summarized in Table Chyba! Nenalezen zdroj odkazů.. The engine is based on existing design of unspecified manufacturer, however there were some design changes thus, the target engine model corresponds to a virtual engine. The engine model was built in GT-Power 0-D/1-D code [1] which enables to simulate the whole engine cycle including exhaust gas energy transfer between the cylinders and the turbocharger. The engine layout is shown in Figure Chyba! Nenalezen zdroj odkazů. in terms of boosting system configuration, it can be considered as either e-turbocharger (e-booster + ETC; left subfigure in Figure Chyba! Nenalezen zdroj odkazů.) or the electrical turbo-compound (ETC; right subfigure in Figure Chyba! Nenalezen zdroj odkazů.). e-machine: hel e-machine: hel throttle WG valve throttle WG valve e-machine: hel e-machine: hel e-machine: hel EGR valve EGR valve Figure 1: Basic engine layout including BMEP control actuators, low EGR circuit and location of all electrical machines left: e-turbocharger, right: electrical turbo-compound (ETC). All results presented in the paper were obtained by means of simulation. Sensitivity studies were performed and relative comparison of tested variants was done. As the paper is focused on thermodynamic potential, very high engine flexibility was assumed the engine is supposed to be equipped with fully variable VVA system at intake/exhaust valves and variable compression ratio device. Moreover, different combustion concepts were considered the engine is supposed to be able to run in standard SI mode (knocking is taken into account), classical CI mode and ideal RCCI mode. 71

73 Engine Parameter Unit Value Bore [mm] 75.1 STROKE [mm] [1] [rpm] 75.1 Variable (9-25) Turbocharged RCCI/SI/CI Compression Ratio Engine Speed Charging Combustion Mode Fuel Fuel Injection Configuration Number of Intake Valves Number of Exhaust Valves Indolene/Diesel Direct R3 2 2 Table 1: Main engine parameters. Reasonably calibrated engine model was needed as many operating conditions, which were tested by simulation, cannot be verified on an existing engine(s). Hence, submodels with high predictive ability were applied. The following predictive sub-models were applied. Simplified FE (Finite Element) model was used to calculate combustion chamber temperatures while Woschni formula (c.f. [19]) was applied to estimate the heat transfer coefficient between the in-cylinder gas and the walls (boundary condition for FE model). The engine model was built while using as much experimental information as possible. However, as the engine model is a virtual one, unavailable information was estimated by the authors using experience with similar engines. Additional information concerning engine model can be found in [3, 4, 5]. Concerning turbocharger maps, the standard approach was adopted. This means an application of lumped compressor/turbine model(s) using standard maps provided by turbocharger manufacturer. Dealing with modelling of electric motors/generators (labelled as e-machines in the text below), the following can be stated. There are 3 emachines in the model c.f. Figure 1. The first one is attached to the engine, the second one propels the compressor while the third one is connected to the turbine. All these e-machines are modelled in the most simplified way each e-machine can operate as a motor or a generator depending on requirements from control system. Moreover, constant electric efficiency was assumed the default value of that is 90% (based on experience from EU FP/H2020 projects and data from [6]). 3. COMPUTED CASES Engine performance was tested at the following BMEP levels: 13, 18, 25 bars and full load. Engine setting was optimized for all considered combustion modes (RCCI, SI, CI) while considering ETC, e-turbocharging and classical turbocharging. A sensitivity study of e-machine efficiency was performed at engine speed of 2000 rpm. Finally, constant pressure configuration of exhaust manifold was considered for both eturbocharger and ETC this was achieved by means of increasing significantly the diameter of exhaust piping between an engine and a turbine. 72

74 3.1. Optimization Procedure The optimization procedure is based on application of genetic algorithm [2]. Due to high engine variability/flexibility, time-demanding multi-variable multi-constraint singletarget optimizations were performed to find thermodynamic optimum (either maximum BMEP at full load operation or minimum BSFC at prescribed engine load). The variables to be optimized were the following: compression ratio, intake/exhaust VVA setting (4 variables in total), combustion phasing, combustion duration (RCCI combustion mode only), air excess, EGR rate, turbine BSR (e-turbocharger only) and WG flow (e-turbocharger and ETC only). The optimizations were constrained by many limits (e.g., maximum in-cylinder pressure, compressor surge, minimum air excess, knocking) depending on considered combustion mode. 4. DISCUSSION OF RESULTS This section presents the most important results from all simulated cases. The main focus is on a comparison between classical turbocharging approach (no special label in figures; dashed lines), e-turbocharging one (labelled as e-turbo in figures; dashed lines) and ETC (labelled as e-assisted_turbo in figures) Full Load Performance The results concerning full load operation is plotted in Figure Chyba! Nenalezen zdroj odkazů.. The main conclusion is that there is little difference among all considered variants with the exception of very low engine speed range where performance of classical turbocharger is strongly limited low boost pressure is a consequence of low turbine power at this engine speed range. The issue of low boost pressure can be compensated if e-turbocharger/etc is applied additional compressor power is provided by electric generator connected to engine crank train (c.f. Figure Chyba! Nenalezen zdroj odkazů.). Qualitative trends of important engine design variables and operating conditions are very similar for all considered turbocharging concepts (ETC, e-turbocharger, classical turbo). Low engine speeds are negatively dominated by in-cylinder heat transfer, hence low compression ratio and delayed combustion are applied. Compressor surge can also be an issue high valve overlap is an optimal solution. At middle and high engine speeds, maximum in-cylinder pressure is the main constraint this is a consequence of high boost pressure, lower importance of incylinder heat transfer and increased compression ratio. Maximum BMEP is achieved between 1750 and 3000 rpm at higher engine speeds, BMEP drops significantly due to decreasing volumetric efficiency, higher pumping and friction losses. 73

75 Figure 2: Comparison of maximum BMEP for the case of all combustion concepts (RCCI, SI and CI) while considering e-turbocharger (label e-turbo), ETC (label e-assisted_turbo) and classical turbocharger left subfigure: ETC vs classical turbo, right subfigure: ETC vs e-turbocharger Influence of Engine Load Different engine BMEP levels were tested 25, 18 and 13 bar. These clearly require different boost pressure levels, hence boost group performance is expected to be less important when approaching lower BMEPs. Engine setting was optimized to achieve the lowest possible BSFC. Stoichiometric operation was prescribed for SI combustion mode (due to three-way catalyst performance). Concerning the case of BMEP of 25 bar, it is presented in Figure Chyba! Nenalezen zdroj odkazů. and Figure Chyba! Nenalezen zdroj odkazů.. Based on full load performance (c.f. Figure Chyba! Nenalezen zdroj odkazů.), the engine speed range had to be limited to the interval between 1750 and 4500 rpm. In terms of BSFC, there is a significant difference between e-turbocharger/etc and classical one at the lowest engine speeds for combustion modes SI and CI. Classical turbocharger is operated just on its limit at 1750 rpm, therefore exhaust gas temperature needs to be increased in order to provide sufficient boost pressure (which is relative high). On the other hand, e-turbocharger and ETC are more efficient due to its higher flexibility. When engine speed is increased, performance of classical turbocharger is improved, which leads to similar BSFC levels as for the case of e-turbo/etc. Overall lower efficiency of eturbocharger layout (due to additional losses in e-machines c.f. Figure Chyba! Nenalezen zdroj odkazů.) is compensated by higher flexibility of the system it is possible to operate all turbomachinery under slightly better conditions, which allows for better overall gas exchange process (c.f. Figure Chyba! Nenalezen zdroj odkazů.). Regarding ETC system, it is also more flexible when compared with classical turbocharger it can control boost pressure by means of applying electrical power to e-machine (attached to turbocharger shaft), hence it can avoid waste-gaiting, hence improving gas exchange phase (c.f. Figure Chyba! Nenalezen zdroj odkazů.). Moreover, ETC system is more efficient when compared with e-turbo (c.f. Figure Chyba! Nenalezen zdroj odkazů.) due to the fact that most of the turbocharger required power is transformed via turbocharger shaft, hence losses in e-machines are minimized. This is a general trend observed in almost all considered cases gas exchange process is usually slightly better for e-turbo and ETC while indicated efficiency of HP phase (of engine cycle) is usually better for classical turbocharger (mainly due to ability to apply stronger Miller cycle and higher compression ratio). Qualitative trends of important engine design variables and operating conditions are 74

76 similar when comparing classical turbocharger case with e-turbo and ETC. Although heat transfer is the most dominant factor at low engine speeds, relatively high compression ratio is used for RCCI and CI combustion modes. This is not possible for SI mode because of knocking. Due to its higher flexibility, RCCI mode allows for using the highest compression ratio when comparing different combustion modes. The maximum in-cylinder pressure constraint is typically the main limiting factor for all considered engine speeds. Compressor surge is less of an issue and maximum compressor speed is reached at medium and high engine speeds. Optimal combustion phasing is relatively late regardless of engine speed or combustion mode. This is related to heat transfer (low engine speeds) and maximum in-cylinder pressure constraint (middle and high engine speeds). The ETC variant seems to be the best one. Figure 3: Comparison of BSFC at BMEP of 25 bar for the case of all combustion concepts (RCCI, SI and CI) while considering e-turbocharger (label e-turbo), ETC (label eassisted_turbo) and classical turbocharger left subfigure: ETC vs classical turbo, right subfigure: ETC vs e-turbocharger. Figure 4: Comparison of gas exchange process (pumping work and attachment work) at BMEP of 25 bar for the case of all combustion concepts (RCCI, SI and CI) while considering e-turbocharger (label e-turbo), ETC (label e-assisted_turbo) and classical turbocharger left subfigure: ETC vs classical turbo, right subfigure: ETC vs eturbocharger. The results concerning BMEP of 18 bar are shown in Figure Chyba! Nenalezen zdroj odkazů.. The relative comparison amongst variants is very similar to the abovedescribed case of BMEP of 25 bar (Figure Chyba! Nenalezen zdroj odkazů. and Figure Chyba! Nenalezen zdroj odkazů.), however the differences are smaller due to 75

77 the fact that the boost groups is less important as lower boost pressure is needed. The general trend is that (again) the ETC variant seems to be better than e-turbo or classical turbocharger. Figure 5: Comparison of BSFC at BMEP of 18 bar for the case of all combustion concepts (RCCI, SI and CI) while considering e-turbocharger (label e-turbo), ETC (label eassisted_turbo) and classical turbocharger left subfigure: ETC vs classical turbo, right subfigure: ETC vs e-turbocharger. The results concerning BMEP of 13 bar are presented in Figure Chyba! Nenalezen zdroj odkazů.. The qualitative trends are (again) the same as for the case of BMEP of 18 bar (Figure Chyba! Nenalezen zdroj odkazů.) while the differences among the variants are even smaller. The reasons behind that are the same required boost pressure is low, hence the importance of boost group performance is relatively low. Figure 6: Comparison of BSFC at BMEP of 13 bar for the case of all combustion concepts (RCCI, SI and CI) while considering e-turbocharger (label e-turbo), ETC (label eassisted_turbo) and classical turbocharger left subfigure: ETC vs classical turbo, right subfigure: ETC vs e-turbocharger Influence of HP Exhaust Manifold Volume The original exhaust manifold of the considered engine is the pulsating design as it is very compact. This leads to significant pressure pulsations which lead to significant decrease of average turbine efficiency more details can be found in [5, 7]. As both eturbocharger concept and ETC one can use additional electric energy (stored in a battery or supercapacitor) to provide fast transient response, there is no need to keep the pulsating design of exhaust manifold. Based on these facts, it was decided to test 76

78 constant pressure design. The HP exhaust manifold volume was increased by factor of 15 this was achieved by means of increasing the exhaust manifold diameter while preserving the original length. The constant pressure design is labelled as dexhhp=117mm in figures and it was possible to significantly reduce pressure pulsations (c.f. [5] for more details). The comparison between pulsating design and constant pressure one is presented in Figure Chyba! Nenalezen zdroj odkazů.. Almost no improvement of BSFC was achieved small improvement can be observed at very low engine speeds. It seems that higher heat losses in HP exhaust manifold (due to larger surface), possibly nonoptimal size of turbine (due to smaller pressure pulses, turbine size might be a bit too large) and slightly worse gas exchange process counteract improved turbine thermodynamic efficiency. Moreover, it was observed earlier that high engine flexibility can compensate certain negative effects by means of changing the output power distribution between pistons and e-turbine. Finally, pulsations are significantly stronger at lower engine speeds while turbine power is low there. Hence, improved turbine efficiency has small effect on engine performance. As engine speed is increased, turbine power increases as well while pulsations become much smaller. Figure 7: Influence of exhaust manifold volume (dashed lines: large volume => low pressure pulsations) comparison of BSFC at BMEP of 25 bar for the case of all combustion concepts (RCCI, SI and CI) while considering e-turbocharger (label e-turbo) and ETC (label e-assisted_turbo) left subfigure: ETC, right subfigure: e-turbocharger. 5. CONCLUSION The paper deals with thermodynamic potential of electrical turbocharging under steady operation for the case of small turbocharged ICE for automotive application, namely eturbocharger and electrical turbo-compound (ETC) these two approaches are compared with classical turbocharging one. High flexibility of ICE was assumed while 3 different combustion modes were considered (classical SI, classical CI and ideal RCCI). Concerning engine operating conditions, maximum load and high/medium load were considered while taking into account the whole engine speed range. The results were achieved by means of simulation using 0-D/1-D system approach the engine model was calibrated and predictive sub-models were applied. The general observation (under steady operation) is that there is little potential of electrical turbocharging to improve important engine output parameters when compared with classical turbocharging approach. Obviously, electrical systems can provide significantly faster transient performance, which is important advantage. 77

79 One area of potential improvement is very low engine speed range and high load operation, which is limited by turbocharger performance (typically insufficient boost pressure) electrical systems (e-turbocharger, ETC) can improve that. Small BSFC improvement can be obtained at low engine speeds and high load when applying constant pressure HP exhaust manifold design this improves turbine efficiency while missing turbine power is provided by e-machine. Regarding SI engine, which is typically limited by knocking at high load, certain BSFC improvement can be achieved at lower engine speeds the main reason is that there is higher flexibility as electrical turbocharging can provide missing (turbine) power to avoid shifting combustion to expansion stroke (to increase turbine inlet temperature) and/or using low-intensity Miller cycle due to insufficient boost pressure. As this operation is usually constrained by knocking, which is very non-linear, any small improvement (in terms of knock) leads to relatively significant BSFC improvement. Presented simulation results allow for comparison of e-turbocharging concept with ETC one. The former system is clearly more flexible as it enables to separately control compressor and turbine. However, due to its layout (c.f. Figure Chyba! Nenalezen zdroj odkazů., left subfigure) it is less efficient. The latter one (c.f. Figure Chyba! Nenalezen zdroj odkazů., fight subfigure) is simpler from design/control point of view while it has significant advantage it is more efficient than e-turbocharger as there is mechanical connection between compressor and turbine while it is significantly more flexible when compared with classical turbocharger. Hence it seems to be the best compromise as this variant is usually the best one in all considered cases. REFERENCES [1] [2] [3] [4] [5] [6] [7] GT-Power User s Manual, GT-Suite version 7.3. Gamma Technologies Inc., modefrontier Multi-Objective Design Environment, version [CDROM], Vítek, O. Theoretical Potential of Future Automotive Internal Combustion Engine in Terms of Efficiency and Performance. In: KoKa 2013, Brno, CR [CD-ROM], ISBN Vítek, O., Macek, J., Doleček, V., Bogomolov, S., Mikulec, A., and Barák, A. Realistic Limits of ICE Efficiency. In: Proceedings of FISITA 2014 [CD-ROM], June Paper Code: F2014-CET Vítek, O. and Macek, J. Thermodynamic Potential of Electrical Turbocharging for the Case of Small Passenger Car ICE under Steady Operation. SAE Technical Paper , 2017, doi: / Terdich, N. and Martinez-Botas, R. Experimental Efficiency Characterization of an Electrically Assisted Turbocharger. SAE Technical Paper Series, Paper , doi: / Zinner, K. Aufladung von Verbrennungsmotoren. Springer, ACKNOWLEDGEMENT This research has been realized using the support of Technological Agency, Czech Republic, programme Centres of Competence, project TE : Josef Božek Competence Centre for Automotive Industry. 78

80 This research has been realized using the support of The Ministry of Education, Youth and Sports program NPU I (LO), project LO1311: Development of Vehicle Centre of Sustainable Mobility. All the help has been gratefully appreciated. ABBREVIATIONS CI ETC HP RCCI SI Compression Ignition Electrical Turbo Compound High Pressure Reaction-Controlled Compression Ignition Spark Ignition 79

81 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES PHASE CHANGE MATERIALS FOR ENGINE WASTE HEAT ACCUMULATION AND STORAGE Luboslav Kollar1 Abstract New phase change materials (PCM) mixtures were analyzed for waste heat accumulation in cooling loop of internal combustion engines (ICE). Sugar alcohols (SA) have been selected as basic materials with high latent heat of fusion providing a good volumetric energy density to heat accumulation. The mixtures of Erythritol and Mannitol with Urea were analyzed with focus on melting temperature change. The most SA with high latent heat have the melting point exceeds temperature level of the ICE coolant fluid. Adding Urea in SA has a significant affect to decrease a melting temperature of the PCM. 1. INTRODUCTION One of the priorities in development of the ICEs is increase the fuel economy and decreases the exhaust gas emissions. The significant impact on the efficiency of the ICE has good thermal management. The waste heat recovery represents way how to use the waste heat to improve fuel economy and emissions decreasing. The significant demand for the heat is at the engine start when all the fluids and engine block are cold. At that time is requirement for fast heat up the engine on nominal conditions. Exhaust heat recovery can significantly help to improve it by transferring of exhaust gas heat to cold coolant. This is limited by low exhaust gas temperature when the engine starts and the heat transfer is low. To improve this process can be used the heat accumulated by the PCM. The accumulation occurs when the engine runs on nominal conditions and needs to be cooled. PCM is located in hermetically closed channels of heat exchanger in engine cooling loop. PCM melting point needs to be under the nominal temperature of the coolant in engine cooling loop to be able absorbs the heat in form of latent heat. The modification of melting temperature to adjust it at temperature level of engine cooling loop is required. 1 Luboslav Kollar, Hanon Systems, Zavodni 1007, , Hluk, Czech Republic, lkollar@hanonsystems.com 80

82 There were analysed properties of mixtures SA with urea (Table 1) and its impact on melting temperature in this study. Urea Erythritol Mannitol 54,9/45,1 50/50 -/40/60 Table 1: Basic materials to mixture 2. WASTE HEAT ACCUMULATION BENEFITS The engine waste heat accumulation is primary proposed for engine thermal efficiency improvement. It brings lower fuel consumption with significant impact on emission reduction. The heat is accumulated in engine cooling loop and stored in form of latent heat in PCM. The amount of heat absorbed in PCM decreases the load of front radiator and it affects size and weight of radiator and cooling loop. Smaller radiator surface affects the final air drag with positive impact on fuel consumption. Storing the surplus of heat in engine cooling loop enables its use in the time when the heat is required. The availability of heat during the cold start supports faster engine oil heat up and decreases its friction loses. Heat from the coolant loop is transferred to engine block to help the engine reach the nominal working and combustion conditions. It takes less time to engine get at run conditions with optimal fuel efficiency. On the other side the exhaust gas increases the temperature more quickly which enables faster heat up of depollution system with positive effect on emission reduction and eliminates the post combustion process for heat up the system. Picture. 1: Erythritol / Urea thermal properties 81

83 3. PHASE CHANGE MATERIALS PROPERTIES There are required specific thermal properties of the PCM for the heat accumulation in the engine cooling loop. The important characteristics are melting temperature and latent heat. The fluid temperature of the engine cooling loop is limited up to 110 C. The thermal properties of the sugar alcohols are suitable for thermal storage applications. They are characterized with high latent heat of fusion and good thermal stability. The table 2 indicates the properties of three basic sugar alcohols and urea. Erythritol and Mannitol have the melting point above the temperature limit of the cooling fluid. Urea was as the additive compound to affect the melting point. Erythritol Mannitol1 Xylitol Urea1 Melting temperature 118,8 167,0 93,4 133,3 Latent heat 317,5 330,0 241,0 243,7 Table 2: Basic materials thermal properties 3.1 Modification and analysis of new substances For analysis of materials was used differential scanning calorimetry method. It was analysed melting temperature and enthalpy of the material samples weight 3-5 mg. Thermal stability was measured by thermogravimetry (TGA) method. Sample was heated up to 150 C or 250 C at the ambient conditions. To analyse the thermal properties of Erythritol Urea mixture was prepared a sample with eutectic compound 54,9 / 45,1 (w/w) [2]. The sample was heated up to 110 C with rate 2 C/min. Than it was cooled down to 20 C with rate 15 C/min. Table 3 shows the melting temperature and enthalpy of this new compound. Erythritol/Urea Ratio %(w/w) Melting temp ( C) 54,9/45,1 80,9 Melting enthalpy (kj/kg) 210,4 Table 3: Erythritol / Urea thermal properties 82

84 The DSC chart on picture 3 shows the melting process of the Erythritol/Urea between the 70 C to 90 C. dsc_0187_urea_erythritol_4_chlazeni 5 Onset x: C Enthalpy (normalized): J/g Exo Up Picture 2: Erythritol / Urea DSC thermal analysis The reverse process of the solidification (crystallization) of the mixture was not observed properly because of significant subcooling temperature and it needs to be analysed particularly. The thermal stability of the mixture depending on heating rate is evaluated on Table 4. It was measured at different heating rate and compared in the chart on Picture 3. Erythritol/Urea ratio 54,9/45,1 (% w/w) Heating rate Degradation temp. ( C/min) ( C) Table 4: Erythritol / Urea thermal properties 83

85 It shows that with increasing heating rate increases the degradation temperature of the material mixture. tg_0114_urea_erythritol_10 k_min Onset x: C Onset x: C Onset x: C Onset x: C Picture 3: Erythritol / Urea TGA degradation temperature depending on heating rate The second analysed mixture was Mannitol Urea with ratios 50/50 (w/w) [1] and 40/60 (w/w) [1]. The samples were heated up to 135 C with rate from 1 C/min to 7 C/min. Than they were cooled down to 20 C with rate 15 C/min. Table 5 shows the melting temperature and enthalpy with comparison of crystallisation properties of the mixture at different ratios. 50/50 Melting temp ( C) 102,1 40/50 102,0 Ratio %(w/w) Mannitol/Urea Melting enthalpy (kj/kg) 162,0 179,6 Crystallization Crystallization temp ( C) enthalpy (kj/kg) 44,8 138,2 68,5 150,7 Table 5: Mannitol / Urea thermal properties Picture 4 shows the DSC analysis of the mixture with ratio 40/60. We can assume according the Table 5 this mixture is closer to eutectic ratio. It has significantly higher crystallization temperature and also higher enthalpy. 84

86 DSC_227_MaMo46_5_melt Exo Up Picture 4: Mannitol / Urea, ratio 40/60 w/w, DSC thermal analysis 4. CONCLUSION Two mixtures of sugar alcohols with Urea were analysed. Adding of urea to mixture with sugar alcohols decreases the melting temperature. It also affects its latent heat of fusion and crystallization. Both mixtures have the melting point below the maximum temperature of the cooling fluid in the engine loop. Urea also significantly decreases melting enthalpy and affects the crystallization properties. The Erythritol/Urea was characterized with high value of subcooling for crystallization which limits its use in case of heat storage application. On the other side it has significantly higher melting enthalpy compared to Mannitol/Urea mixture. Both compounds shown high rate of crystallization what is important for thermal dynamics in the heat accumulation and storage applications. REFERENCES [1] Johan Göhl, Robert Paberit, Erik Rilby, Jan Swenson, Pär Johansson, Helén Jansson, Manipulation of phase transition temperatures and supercooling of sugar alcohol based Phase Change Materials (PCMs) by urea, INNOSTORAGE Conference, 2016 [2] G. Diarce, L.Quant, Á.Campos-Celador, J.M.Sala, A.García-Romero Determination of the phase diagram and main thermophysical properties of the erythritol urea eutectic mixture for its use as a phase change material, Publishing, 2016, pp ISBN

87 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES THE EFFECT OF EGR COOLER DESIGN ON PARTICULATE FOULING Jiří Bazala1, Guillaume Hébert2 and Oliver Fischer3 Abstract New trends in automotive industry for Internal Combustion Engines require EGR coolers with high efficiency which have to keep this efficiency during the lifetime of a vehicle. This article focuses on fouling test of two EGR coolers with different technology of heat transfer. After 80 hours of engine test calorimeter results showed us that fin type EGR coolers have relatively lower degradation of Heat efficiency compared to tube type ones. The contrary can be concluded in terms of pressure drop. 1. INTRODUCTION Particulate Fouling is one of the most important phenomena of EGR cooler which has direct effect on the first sizing of the cooler. Today's practice shows that EGR performance parameters are overdesigned, because of degradation of heat rejection and pressure drop during the lifetime of a vehicle. Emission legislation is the key driver for a car manufacturer to reduce emissions. It means that EGR-cooler politics by car manufacturers tries to achieve lower emission by EGR coolers which provide bigger efficiency. This is caused by the direct effect of cooled EGR efficiency on NO x emission reduction (for diesel engines). In other words, EGR coolers have to have the best ratio between size and heat rejection. This ratio parameter is suitable for comparison of EGR cooler technologies such as a tube type and a fin type. However, this parameter can be different for clean new coolers and after real usage in a vehicle. This paper describes and discusses an experiment dealing with establishing rules for cooler type selection. Jiří Bazala, Hanon Systems, jbazala@hanonsystems.com Guillaume Hébert, Hanon Systems, hebert4@hanonsystems.com 3 Oliver Fischer, Hanon Systems, ofische3@hanonsystems.com

88 2. BODY OF PAPER 2.1 Fouling engine test set-up: For better repeatability of the test as well as the reduction of testing time it was decided to test two coolers in parallel on diesel engine connected to the dynamometer (Fig.1). Exhaust gas from engine to fouling fixture was taken downstream of turbocharger and upstream of exhaust after treatment (DPF etc.) and the whole amount of the exhaust gas in fixture was split to the two branches. The first arm had Fin cooler and the second one was fitted with Tube cooler. In order to keep the same mass flow in every arm control mechanism consisting of the orifice, two pressure sensors and throttle body were used. If the pressure drop Δp1 and Δp2 are the same, mass flow has to be the same as well (Formula 1). 𝑄𝑣 = 𝑆2 𝑣2 = 𝑆2 2 1 (𝑆2 ) 2(𝑝1 𝑝2 ) (1) 𝜌 𝑆1 From this formula it can be observed that different variables include only pressure in front of and behind orifice if all dimensions are the same. Figure 1: Fouling engine test set-up with fouling fixture and EGR coolers. 2.2 Exhaust tested parameters: The literature says that unburned hydrocarbons emissions negatively affect EGR cooler particulate fouling. Hydrocarbon particulate with good adhesion can easily stick to the inner part of the cooler made of stainless steel and create small continuous layer on inner surface of the cooler. The layer has the ability to stick bigger dry soot particulates. The cumulative effect of mixing hydrocarbons and dry soot can completely clog the EGR cooler. This is the reason why the cycle test with mixed conditions (Table 1.) was selected. The first state aims to have the biggest amount of hydrocarbons. This state simulates driving with low load. In the second state there is an effort to supply as much as possible Particulate Matters (PM) to the measuring chain. The third state simulates higher load for example in highway cruising. The total length of the test was 80 hours. 87

89 Mass Duration flow cooler 1-10 h 7.5 g/s h 11.5 g/s h 7.5 g/s h 11.5 g/s h 7.5 g/s h g/s HC cooler Smoke cooler 1.1 mg/s 1.1 mg/s 1.1 mg/s Low 0.6 mg/s 0.6 mg/s Low Table 1: Exhaust parameters 2.3 EGR Coolers In terms of the test, two EGR coolers with different technology of inner core but similar heat rejection were selected. The fin type cooler has the measuring points on calorimeter bigger heat rejection by 5% in average. A tube type cooler was U-flow (double path) with round spiral tubes. The fin type cooler was also U-flow with wavy fin and return housing. 2.4 Results For the evaluation of the results, the EGR cooler was measured on calorimeter before and after the fouling test under the same conditions. Heat rejection and pressure drop were monitored during the test (for the test set-up control). Looking at the trend of heat rejection (Fig. 2), it is obvious that the tube type cooler has bigger tendency to be fouled over time. Figure 2: Live results of Heat rejection Calorimeter measurements seem to be more interesting. The main aim was to evaluate the relative degradation of Heat rejection and Pressure Drop between new and fouled states. In terms of the percentages, the initial difference between two coolers can be neglected considering pressure drop and heat rejection. 88

90 Figure 3: Change Heat rejection after particulate fouling Fig. 3 shows that the relative degradation of heat rejection was lower for the Fin type cooler. The results of the Tube type cooler show that the change of heat rejection was much worse with higher mass flow. For example, heat rejection of the Fin type cooler decreased by 3% between mass flows 10g/s and 15 g/s. On the other hand, heat rejection of the tube type cooler has tendency to decrease by 6%. Fig. 4 shows that change of pressure drop was significantly better for the Tube type cooler. The results of change of pressure drop show, that percentage change was generally constant over the measured range of mass flows: around 42% for the Fintype cooler while only around 7% for the tube type cooler. Figure 4: Change pressure drop after particulate fouling 2.5 Discussion The contradictory findings between pressure drop and heat rejection can be explained by different geometries of these heat exchangers; and the theoretical differences of their function. Fin type coolers have bigger cross section and therefore gas flow is more laminar; on the top, the heat transfer surface is much higher in comparison to Tube type coolers. That s the reason why heat in these tube type heat exchangers is mainly transferred by convection which is enhanced with turbulent flow depending on the Reynolds number. 89

91 From the perspective of heat rejection, the tube cooler has worse results. It can be explained by the combination of thermal insulation properties of soot and small cross section area of tube type cooler. This area is after fouling even smaller which results in higher Reynolds number. In this case, on the top of higher pressure drop we would expect also higher heat transfer. However, as the Fig.5 shows, this is not automatically the case. Figure 5: Heat Transfer coefficient in Turbulent flows From the perspective of pressure drop, the results can be explained by looking at the hydraulic diameter parameter and the heat transfer area. The Clean Fin type cooler has more than twice smaller total hydraulic diameter but also bigger heat exchange surface than the Tube type cooler. It means that losses caused by flow friction have bigger impact than fouling. In total, more soot is expected in large heat transfer area. It was also concluded that even the relative degradation of fin type cooler pressure drop is bigger, the total pressure drop is still lower than in case of fouled tube type coolers (Fig.6). Figure 6: Pressure drop new vs. fouled coolers 90

92 3. CONCLUSION 1. Speaking about the relative degradation of the pressure drop along life, the Tube type cooler looks more stable. 2. However, the low pressure drop of the fin type heat exchanger vs the tube type ones (at same heat rejection) results in the fact that, even after particulate fouling, the tested fin type Heat exchanger had better pressure drop than the tested tube one (absolute number). 3. In terms of heat rejection the fin type cooler is better, as new as well as all along the lifetime. 4. Some regeneration phenomenon, which is not described in the paper, has been identified during the study, the authors plan to further investigate it. REFERENCES [1] Yunus A. Cengel, Afshin J. Ghajar, 2011, Heat and Mass Transfer: Fundamentals&Aplications Fourth Edition, New York: McGraw-Hill [2] Incropera, F. P., & DeWitt, D. P., 2002, Fundamentals of heat and mass transfer, New York: J. Wiley. ACKNOWLEDGEMENT This work was supported by the internal BUT research project Reg. No. FSI-S

93 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES SNIŽOVÁNÍ EMISÍ OSOBNÍHO AUTOMOBILU A LEGISLATIVA Martin Hrdlička1 Abstract Příspěvek dává do souvislostí aktuální předpisy ke snižování emisí osobních vozů, jejich aktuální vývoj v různých regionech Evropy a světa, jejich vliv na vývojové cíle. Prezentuje praxi, zkušenosti a nutné investice do nových technologií a zkušeben včetně know-how provádění zkoušek. Porovnává zkušební cyklus NEDC a WLTP. Osvětluje praxi prováděných zkoušek RdE. Vysvětluje funkci EOBD a uvádí příklady této funkce včetně Readiness Code. Připomíná nutnost změny paliv v souvislosti s emisními předpisy a nový SHED test. Zmiňuje další, obecné souvislosti osobního automobilu a životního prostředí. 1. INTRODUCTION Ve svém příspěvku bych chtěl dát do souvislosti aktuální předpisy o snižování emisí osobních automobilů, jejich bouřlivý vývoj v různých regionech Evropy a světa a jejich vliv na vývojové cíle. Chtěl bych představit praktické zkušenosti a nutné investice do nových zkušeben, do zařízení a zkoušek. Porovnám zkušební cyklus NEDC a WLTP a osvětlím praxi prováděných zkoušek RDE. Vysvětlím funkci EOBD a uvedu i příklady funkce včetně readiness code. Zmíním vliv kvality paliv v souvislosti s emisními předpisy a i nový SHED test. Dotknu se i dalších obecných souvislostí ve vztahu osobního vozidla a životního prostředí. Vedle limitů na jednotlivé složky sledovaných emisí spalovacích motorů v osobním vozidle jsou v poslední době sledovány a diskutovány zejména hodnoty oxidu uhlíku (oxidu uhličitého), které vlastně vyjadřují jen jiným způsobem spotřebu benzinu, nafty či CNG vozu. CO2 není dosud pro vozidla limitován legislativou EHK a není obsaženo v legislativě emisních předpisů. Výrobci automobilů uvádí hodnoty CO 2 v g/km pro každý vůz a tyto hodnoty uvádí vždy na základě měření v předepsaném emisním cyklu, kde se měří i limitované složky emisí dle zákona. 1 Ing. Martin Hrdlička, PhD., MBA - vedoucí vývoje podvozku a agregátu ve Škoda Auto a.s. 92

94 2. EMISE AUTOMOBILU A LEGISLATIVA 2.1 CO2 Oxid uhličitý nepatří mezi přímo škodlivé látky, ale s ohledem na stále stoupající produkci CO2 vlivem lidské činnosti na naší planetě, je označováno množství CO 2 za jeden z podstatných vlivů, který stojí za globálním zvyšováním teploty na naší planetě Zemi. Bilance CO2 emisí na naší planetě je uvedena na obrázku 1. CO2 emise ve světě vliv lidské činnosti Celkový podíl osobních vozů na světové produkci CO2-emisí: 0,2 % 2 Obr. 1 Oxid uhličitý a hlavní zdroje Vlivem lidské činnosti vzniká cca 3,5% CO2 a z těchto 3,5% výrobou a provozem osobních automobilů vzniká cca 5,5%. Tedy celkový vliv výroby a provozu osobních automobilů na produkci CO2 je ve výsledku cca 0,2%. Výrobci a prodejci osobních automobilů budou od roku 2020 v EU zatíženi pokutou, která se bude počítat dle tzv. flotilové spotřeby prodaných vozidel dané značky v kalendářním roce. Pakliže flotilová spotřeba překročí stanovené limity CO 2 (stanovené v g/km) v letech (viz obrázek 2), zaplatí výrobce pokutu ve výši rovnající se součinu hodnoty překročení limitu, počtu prodaných všech vozů dané značky v EU a předepsané pokuty v daném období. Tedy při překročení flotilové spotřeby v EU o jeden gram v roce 2020 by výrobce, který prodá v daném roce např. 1 milion vozů, zaplatil pokutu ve výši 1 x x 95 = EUR. Výše pokuty se v letech zvyšuje a limitní hodnota CO2 pro flotily se v letech snižuje, takže výrobci jsou logicky v EU nuceni volit strategie ke splnění hodnot limitů. Výrobci samo sebou nechtějí platit pokuty, samozřejmě je pro ně efektivnější investovat do nových technologií a jejich zvládnutí a navíc zatím není zcela jasné, na co a jak budou v EU peníze z případných zaplacených pokut použity. Dnes je situace v EU o to složitější, že existují v různých zemích různé malusové či bonusové systémy, které jsou různě počítány a platí je někdy zákazník, někdy výrobce, 93

95 či oba. Není jednotný daňový systém pro řízení emisí CO 2 - nejčastější jsou systémy dovozní, registrační, roční daně, či jejich kombinace. CO2 Plán snižování emisní náročnosti v EU, Flotila CONFIDENTIAL 3 Obr. 2 Limity podle návrhu EU Některé limity CO2 pro výpočet daně přitom rostou skokově, jiné lineárním přepočtem. Obecně existují v různých zemích různé limity bonusů, malusů a nebo daní a to jak pro flotilové spotřeby, tak někdy i pro jednotlivá vozidla (obr. 3). Vedle podpor či sankcí států je jistě pro výrobce vozidel významný konkurenční boj, tedy porovnání se s konkurencí a pro spotřebitele reálná spotřeba vozidla. Je nutno říci, že výrobci automobilů v EU jsou povinni udávat v technické dokumentaci a katalozích hodnoty CO2, které jsou výsledkem měření v emisní laboratoři, v tzv. emisním cyklu NEDC za okrajových podmínek daných přesně zákonem. Není na rozhodnutí výrobce, jaké hodnoty udává, ale musí udávat hodnoty, které vyplývají z NEDC testu. V budoucnu hodnoty, které budou vyplývat z měření nového testu tzv. WLTP, budou podle zákonných norem přepočteny na ekvivalent hodnoty CO2 z testu NEDC. To aby zákazníci měli stále jakousi kontinuitu a hodnoty byly pro zákazníky a jejich vnímání pochopitelné. Technicky lze množství CO2 ovlivnit několika způsoby - přímo typem pohonné jednotky a její optimalizací, valivým odporem, aerodynamikou, hmotností (obr. 4). 94

96 CO2 - Limity pro výrobce i pro zákazníky Na emise CO2 se vztahuje legislativa pro výrobce: průměrné hodnoty CO2 všech prodaných vozů značky za rok (např. EU28 nebo Čína), ale i na jednotlivé zákazníky (např. Nizozemí, Francie) A B Flotilové limity Za splnění svých limitů na flotilu prodaných aut v daném roce je zodpovědná každá značka. Sankce za nesplnění jsou finanční pokuty či zákaz prodeje. Spotřeba (l/100 km) CO2 (g/km) 5,6 Nižší CO2-hodnoty jednotlivých vozidel, znamená daňové zvýhodnění pro zákazníka, nižší daň při registraci vozidla apod ,1 Limity pro jedn. vozidla , ,1 148 /g , , Spotřeba (l/100 km) 7, CO2 (g/km) /g /g ,3 103 /g 7, ,9 7,0 <95 6, ,4 6, ,0 6,0 126 <88 0 /g 0 /g 136 5,5 5, /g /g <82 0 /g 2023,490 CONFIDENTIAL 5,0 5,0 4, CONFIDENTIAL Obr. 3 Legislativní limity Cíle při vývoji vozu vlivy na CO2 nízká spotřeba motor převodovka požadavky pneumatiky hnací ústrojí dobrý jízdní výkon cwxa hmotnost VERTRAULICH Obr. 4 Ústrojí a parametry vozu významně ovlivňují produkci oxidu uhličitého 95

97 2.1.1 Technické možnosti ke snižování podílu CO2 Každou technickou možnost ke snížení CO2 detailně sledujeme a vyvíjíme k lepším hodnotám, každá z nich je ale také stále častěji předmětem přímo, či nepřímo nějakých zákonných zkoušek či homologací. Příkladem může být valivý odpor pneumatik ten však zdaleka není jediným parametrem, který pneumatika musí splňovat a který musí vylepšovat, parametrů je mnoho - a i například odvalovací hluky jsou sledovány legislativou. Brzdné dráhy na suchu, na mokru, životnost pneumatik, komfort jízdy, dynamika řízení, grip a mnoho dalších je ve hře a technik vývoje je musí zohlednit. 2.2 Škodlivé složky výfukových plynů Téma škodlivin ve výfukových plynech spalovacích motorů osobních vozů bych chtěl začít výčtem sledovaných a tedy limitovaných složek, kterými jsou oxid uhelnatý, oxid uhličitý (zákonem však není přímo limitován), oxidy dusíku, uhlovodíky a pevné částice. Pro představu chci upozornit, že limitované složky činí například u vznětového motoru celkem jen 0,3% objemu všech plynů, které z výfuku jsou vypuštěny. Byť se technici i politici snažili od počátku existence emisních norem (konec 60-tých let v USA) tyto normy celosvětově sjednotit, tak se to z mnohých důvodů bohužel nepodařilo. Existují stále trhy, kde dnes žádné limity stanoveny pro emise nejsou, nebo mají EU2, EU3, EU4, nebo v Indii BS4 v severní Americe LEV, Tier apod. a tedy kladou nároky na automobilky vyvíjet jiné pohonné jednotky a palivové systémy, aby těmto vždy vyhověly. Roli zde hraje mimo jiné kvalita paliva, kterou vlády musí pro dané emisní normy zajistit. Jde nejen o čistotu, kde podíl síry v naftě (v ppm), či ppm manganu v benzínu, nesmí být příliš vysoký, ale i o další kvalitativní parametry paliva dle norem, tak, aby správně mazalo, bylo vzníceno, či zapáleno a posléze dobře hořelo. Síra v některých zemích přesahuje v naftě i 500ppm a spolehlivě ucpe katalysátory a nebo filtry částic, byť daná země chce splňovat normy blízké EU4 a podobně vysoké podíly manganu v benzínu zalepí katalysátory benzinových motorů. Důležité tedy je sledovat kvalitu paliva a pro danou emisní normu mít zajištěnou dodávku paliva odpovídajícího v daném státu vyžadované emisní normě. Některé země stanoví emisní normu předpisem (např. EU3 nebo EU4), ovšem bez funkce EOBD, která je vyžadována v EU. Taková varianta agregátu musí být opět exkluzivně pro daný trh výrobcem automobilu vytvořena. Znamená to vyvinout, odzkoušet, homologovat, vyrobit a posléze servisovat tisíce variant agregátů, které se liší nejen softwarem řídící jednotky, ale i hardwarem motoru, výfukové soustavy, převodovky a palivového systému. V EU v současné době platí emisní norma EU6W. Emisní norma má indexy, jelikož nezůstává zákon a předpisy stabilní, ale v poslední době dochází ve stále kratších časových odstupech ke změnám v požadavcích a tyto se označují indexy. Tyto změny mají velmi rozsáhlé dopady do konstrukce pohonných jednotek automobilů a i do jejich ceny. Předpisy vstupují v platnost vždy ve dvou termínech, a to v termínu pro nově homologované typy vozidel (vozy co před uvedením předpisu nebyly v prodeji v EU) a v termínu většinou cca o rok posunutém dozadu předpisem platným pro všechny typy vozidel (tedy jak pro nové tak i pro vozy, co už na trhu jsou v prodeji). Tyto termíny je nutno znát včas, ovlivňují rozhodování o výrobě vozů i agregátů. Vzhledem k tomu, že automobil se vyvíjí cca 4 roky a nutno vystavět i nové továrny na nové motory a jejich komponenty a dokonce i zavést zcela nové technologie, či některé 96

98 technologie uvést do hromadné výroby je včasná znalost plánovaných limitů a termínů zavedení nových předpisů velmi významná. Změny v předpisech dokumentuje níže uvedený obrázek 5. Velmi významný je dlouho diskutovaný požadavek na změnu testovacího cyklu z platného NEFZ na nový tzv. WLTP cyklus má své termíny a každé písmenko za sebou skrývá již konkrétní opatření a termínové plány výrobců. Je nutno například zavést stoprocentně vstřikování močoviny u dieselových motorů, je nutno zavést filtry pevných částic u benzinových motorů, je nutno vyvinout nádrže z nových materiálů a zavést další opatření ke snížení emise CH vypařováním a difusí. Na schematickém obrázku 6 chci popsat měření a podmínky v laboratoři emisí. Zákonem je předepsáno jak zatížit válce, složka valivých odporů a aerodynamiky je z tzv. doběhové zkoušky, kterou dozoruje autorita certifikující homologace. EU Emisní předpisy, aktuální stav od 12/2015 NEFZ do 12/ NEFZ bez RDE nové typy všechny vozy EU6 W nové typy Monitoring RDE NOx NEFZ s RDE EU6 ZD všechny vozy WLTP bez RDE od 06/2016 WLTP EU6 AD RDE NOx: Monitoring RDE PN: CF=1+0,5* nové typy WLTP s RDE všechny vozy RDE Nox: NOx: Monitoring CF=2,1 EU6 AD RDE RDE RDE PN: PN: CF=1+,05* CF=1+0,5* nové typy všechny vozy RDE NOx: CF=1+0,5* EU6 AJ RDE PN: CF=1+0,5* Obr. 5 Změny v předpisech (přechod z EU6W na EU6ZD a i plánované EU6ZD a EU6ZG předpisy, které byly na bázi cyklu NEDC, potom i EU6AD, EU6AG, EU6AJ, které budou již na bázi cyklu WLTP a navíc s RDE, tedy vedle testu v laboratoři, bude proveden i test na silnici zatím pouze pro částice a NOx) Vůz je kondicionován min. 12 hodin na přesnou teplotu, čidla analysátorů jsou kalibrována pravidelně mnoho hodin superčistými plyny, pro pohon vozu je použito tzv. referenční palivo, řidič jede pod dozorem zástupce certifikační agentury dle předepsaného cyklu na válcích a nesmí se odchýlit od předepsané křivky, jinak by test byl zneplatněn. Testy jsou zpracovány a všechny jsou úřadům dostupné, v budoucnu dokonce veřejnosti. 97

99 Emisní měření vozu v laboratoři monitor zobrazující jízdní křivku při emisním testu sí ku du y y ík ox id vo d lo el uh lič ox id uh ox id uh itý zn e čá čišť st ují ice cí na tý analyzátory plynů PN filtr ředicího vzduchu čerpadlo pro odběr vzorku vak pro odběr vzorku výfukového plynu čerpadlo teplota plynu chladič kontrola tlaku válcová brzda pro vozidla s pohonem přední nápravy měřící zařízení Obr. 6 Laboratorní měření Porovnání NEDC a WLTP cyklu včetně RDE WLTP versus NEDC - ve významných parametrech se liší, nejde jen o rychlostní profil testu a absolutní hodnoty rychlostí, ale i o délku testu, způsob zatížení válců, kde ve WLTP nebudou válce zatěžovány setrvačnými hmotami v setrvačníkových třídách, ale budou zatíženy dle konkrétní hmotnosti vozu, což znamená investici nových válců s novou technikou ve všech zkušebnách (obr. 7). Firma Škoda Auto otevřela v květnu 2017 nové emisní centrum, které je schopno podle EU6 kritérií WLTP měřit, nejen při teplotě 20 stupňů celsia, jako v NEDC, ale i s užitím klimakomory, jež umožní měřit i při nízkých teplotách (měření při 7 stupních celsia, ale i při +14 stupních celsia, což je také nové podle WLTP. Toto centrum bude mít v budoucnu tři válcové zkušebny včetně jedné klimakomory. Investice do této zkušebny byla zhruba 15 milionů EUR a další zkušebny však musí být ještě doinvestovány, abychom měli dostatečné kapacity. WLTP měření je totiž asi pětkrát časově náročnější než NEDC a ročně musíme provádět mnoho tisíc testů. Nově přibývá i RDE test, tedy test za reálných podmínek provozu na silnici, který je nyní možný díky mobilnímu měřícímu zařízení PEMS (obr. 8). Měření v RDE je velmi striktně předepsáno předpisy a má mnoho okrajových podmínek. Před každým měřením je nutno PEMS zařízení nejen namontovat a vůz pro měření tedy vybavit, ale i zařízení kalibrovat v emisní laboratoři. Zařízení musí absolvovat tři platná měření za splněných okrajových podmínek a i při zákonem splněných podmínek postprocesingu. Proces vyhodnocení jedné motorové varianty představuje čtyři až pět týdnů, aby bylo vše splněno. 98

100 WLTP (Worldwide harmonized Light vehicles Test Procedure) NEDC WLTP Innerorts ~ 2016 NEDC nové typy od nejsou homologace na základě NEDC dále přípustné všechny vozy! WLTP Pozn. NEDC = New European Driving Cycle Obr. 7 Přechod z dosavadní metodiky (NEDC) na novou (WLTP) Zařízení PEMS v RDE měří ze zákona pouze složky NOx a partikulí, ostatní složky nejsou zákonem v RDE sledovány. Emisní měření vozu v reálném provozu pomocí mobilního měřícího zařízení PEMS Montáž měřícího zařízení na vůz PEMS Verifikace PEMS na válcové brzdě Testovací jízdy s PEMS na silnici PEMS EMISSIONEN NOX mg/km Počet částic Analýza, zpracování a vyhodnocení naměřených dat Obr. 8 Měření v reálném provozu 99

101 2.3 EOBD European On-Board Diagnostics Když mluvíme o emisním předpisu, vždy je nutno mít na paměti, že nejde pouze o dodržení limitů vybraných složek výfukových plynů, ale i o předepsané mechanismy samokontroly motoru a emisně relevantních částí diagnostickým systémem zkráceně zvaným EOBD. Popis systému, předpisy a pravidla pro tento diagnostický systém je velmi rozsáhlý - pro představu textová část popisu v manuálu pro programátory čítá asi stan. EOBD je ekvivalentem amerického standardu OBDII palubní diagnostiky. EOBD byl EU jako předpis zaveden v roce 2001 pro vozidla se zážehovým motorem a v roce 2004 pro vozidla s motorem vznětovým. EOBD European On-Board Diagnostics Příklad diagnostiky účinnosti katalyzátoru řídící jednotka porovnává napětí na lambda-sondě před a za katalyzátorem a z poměru jejich amplitud se určuje účinnost katalyzátoru v případě překročení mezí je řidič informován Obr. 9 Diagnostika katalyzátoru Hlavním úkolem je monitorovat všechny díly a systémy vozidla ovlivňující emise a při jejich závadě, nebo ztrátě účinnosti informovat řidiče zákonem předepsané kontrolky MIL a zabránit tím provozu vozidel, které by měly zvýšené množství emisí. Jedná se o několik set diagnostických funkcí, provádí monitoring základních funkcí, chrání katalysátor před poškozením, používá kódy závad standardní pro všechna vozidla, ukládá závady do paměti, včetně podmínek a času kdy k závadě došlo, o tom, že došlo a že kontinuálně dochází k provádění zákonem stanovených diagnostickým procesům ve voze, informuje tzv. readiness code, tento musí být kompletní, a to nejen po exhalačním testu, ale i po dostatečně dlouhé jízdě v běžném provozu. Funkce EOBD je nedílnou součástí schválení vozu do provozu. EOBD diagnostikuje jednotlivé snímače a akční členy, kontroluje odezvy signálů snímačů a tím diagnostikuje funkce systémů a komponent (např. katalysátor) (obr. 9). Kontroluje, zda vnitřní procesy řídicího systému probíhají správně, proto obsahuje dva procesory - řídící a kontrolní. 100

102 EOBD má své emisní limity a ty předepisuje každá emisní norma nově, při překročení těchto limitů musí dojít k rozsvícení kontrolky. Odstupy limitních hodnot EOBD od emisního limitu se stále snižují. Neklade to nároky jen na diagnostiku, ale je nutno přesněji určit účinnost systému v životnosti a je nutno jinak konstruovat emisní systém. Legislativa definuje i jak často v poměru k provozu vozidla musí být palubní diagnostiky úspěšně provedeny tzv. IUPR (In Use Performace Rate), pro úspěšné provedení testu však musí být splněny předepsané okrajové podmínky. Čítač pro jmenovatel a čitatel IUPR stanoví číslo, jak často je diagnostika systému v činnosti vzhledem k provozu automobilu. Pro představu: v EU6 J musela být provedena úspěšná samodiagnostika minimálně v každém desátém jízdním cyklu, od EU6 W to musí být v každém třetím jízdním cyklu. Readiness Code udává, které diagnostiky jsou systémem podporovány a které jsou provedeny. Tyto informace jsou v řídicí jednotce motoru a jsou ve formě osmimístného kódu, ze kterého je jasné, které diagnostiky byly provedeny a které ne. Při kontrole vozu je nutno, aby Readiness Code měl všechny testy provedeny a uzavřeny, což je důkaz pro dobrou funkci systému. 2.4 Palivo Zmínil jsem kvalitu paliva a na příkladu EU4/EU5/EU6 paliv bych Vám chtěl demonstrovat, které vybrané složky například v normách paliv byly potrefeny, tedy obsah síry, manganu, bioethanolu u benzínu, a popel, síra a nebo methylester řepkového oleje pro naftu. Vedle emisí výfukových sleduje emisní norma i tzv. SHED test, tedy test vypařování a difuse uhlovodíků ze systémů vozidla do ovzduší. Proto jsou používány stále kvalitnější pryže, vícevrstvé plasty s fluorizovanými sycenými povrchy, nádobky s aktivním uhlím s odsáváním par atd. Od září 2018, tedy od EU6 AD/AG se například zdvojnásobuje délka pobytu vozidla v SHED komoře na 48 hodin, přičemž limit pro průnik CH je stejný jako dříve pro 24 hodinový test - tedy méně jak 2g CH/test. Vše povede ke zvětšení AKF nádobek, zcela novým materiálům nádrží, nové konstrukci palivového systému s umístěním většiny dílů uvnitř nádrže, nutnosti zavedení čerpadla pro zkoušku systému v provozu. 3. CONCLUSION S životním prostředím a provozem a výrobou a recyklací automobilu a s předpisy souvisí samozřejmě mnoho, chtěl jsem dokumentovat, že bylo učiněno velmi mnoho a že technici a inženýři se velmi činí a vždy činili, aby kýžených výsledků dosáhli a věřte, že to není snadné. Cesta za dobrým životním prostředím je dlouhá a vyžaduje invenci, ale i stabilitu rozhodnutí a rozumné a přiměřené postupy. Je nutno se koukat kolem sebe a vnímat okolí, vždyť vozový park v České republice má průměrné stáří cca 15 let a 65 % vozů na našich silnicích je starší 10ti let. Tyto vozy samo sebou nejenže neplní dnes zaváděné limity, nemají ale ani ABS, ESP, airbagy ad. Jistě, že existují koncepty na nové pohony od optimalizovaných konvenčních motorů, po elektrické pohony anebo palivové články. Všechny tyto jsou v úvahách kvůli snaze o snížení emise CO2 a hledání nezávislosti lidstva na palivech fosilních. 101

103 Jsem přesvědčen, že pro dobro planety, je třeba posuzovat i emise CO 2 komplexně a tedy zohlednit výrobu energie, výrobu paliv, vznik vozidla a i jeho recyklování, abychom byli úspěšní (obr. 10) CO2 profil pohonů a pohonných hmot, analýza životního cyklu Srovnávací vůz VW Golf (provozní stav km) Elektrovůz 47 Diesel 25 CNG 24 Benzín 24 EU: 60 US: EU: Zemní plyn Norsko: g/km China: Produkce vozidel (Cradle-to-Grave) Připravenost pohonných hmot (Well-to-Tank) Využití (Tank-to-Wheel) 37 Snižování emisí automobilu a legislativa, Škoda Auto, Vývoj podvozku a agregátu, Dr. M. Hrdlička, Obr. 10 Porovnání typů hnacích ústrojí s ohledem na produkci oxidu uhličitého Ve své přednášce jsem Vám chtěl nastínit co vše se za emisemi z pohledu vývojáře a výrobce automobilu skrývá a co je nutno udělat a co je dobré v souvislostech vidět. Úsudky a případná opatření, jež se potom promítnou do legislativy, je nutné podepřít o kvalifikovanou analýzu. REFERENCES [1] TÜV SÜD CZ, 2016, Staus vozový park v ČR 102

104 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES IMPROVED IDENTIFICATION OF ENGINE PREDICTIVE MODELS Vít Doleček1, Petr Denk2 Abstract The paper is focused on optimization of combustion engine predictive model identification procedure. The training input signal is generated during simulation of vehicle driving cycle instead of random signal. The predictive model precision is higher for typical engine speed and load which improve the whole predictive control system quality. 1 INTRODUCTION The engine operation for homologation purposes is defined by time dependent function of speed and load prescribed by driving cycle. Driving cycle represents typical driving behaviour of a car. Different characteristics of traffic in various countries lead to compromise results. The real driving behaviour is very complex function, based on many independent factors like physical route properties (curvature of curves, slope profile and other route trajectory shape properties), traffic density, other random events. Furthermore, special case of the influence to the real riding behaviour is driver. Generally, driver can be represented as a mathematical controller. The natural property of this regulator is the settings to the energy save mode, based on the prescribed total driving time and predictive to the visible horizon [1]. Depending on the above facts, it is possible to model real driving behaviour as an optimization of driving behaviour based on whole route properties with respect to local and global optimization constraints. The outputs of introduces optimization process for next post-processing have to be engine torque and engine speed dependent on the given route point. Route optimization could be part of intelligent cruise control system or driver advising system that allows reduction of fuel consumption and reduction of difference between driving cycle and real driving. It can be next step to the advanced autonomous vehicle control system. Doleček Vít, CTU in Prague, Faculty of Mechanical Engineering, Technická 4, Prague 6, Czech Rep., v.dolecek@fs.cvut.cz 2 Denk Petr, CTU in Prague, Faculty of Mechanical Engineering, Technická 4, Prague 6, Czech Rep., petr.denk@fs.cvut.cz 1 103

105 Utilization of engine model is necessary for vehicle fuel consumption calculation on a given route. Engine optimal setting with predictive controller could be also provided during driving itself. Simplified engine model based only on engine maps are effective from the calculation point of view but their precision is poor for todays downsized engines and high dynamic driving behaviour. Demand for higher precision in dynamic driving cycles leads to utilization of more complex engine models. Conflict of model preciseness with computational speed calls for fast, non-physical black-box models capable of running faster than real time. Proper calibration procedure of these models is key for satisfactory results. 2 PREDICTIVE MODEL TRAINING 2.1 Predictive model Predictive models abilities for dynamic combustion engine controllers were described in [2]. The predictive controller utilizes predictive engine model describing dynamic system behaviour which is in this case combustion engine. As a predictive model LOLIMOT model was chosen for its robustness and real-time capability. LOLIMOT consists of combination of local input-output models that approximate the nonlinear behaviour of an identified system by linear functions valid in particular sub-regions of the whole domain of definition. Analogically to other black-box models, the definition of the sets of linear functions is generated during model training. This procedure is important especially in terms of model results accuracy. The model is adapted to the real system outputs according to the system input parameters Present predictive model training process The question is how the training set of input data should be arranged. All previous work described in [2], [3] and [4] used randomly generated step signal for all system state parameters displayed in Figure. Since the predictive model is applied to the combustion engine, the input parameters are system inputs (fuel mass per cycle, VGT position) and system state variables (engine speed, turbocharger speed, boost pressure ). Figure 1: Engine predictive model training signal engine speed (left) and engine torque (right) [5] If the signals for all input parameters is generated independently, it could lead to the state which is not common in engine usage or even impossible to arrange (e.g. high 104

106 injected fuel mass with very low VGT position resulting in very high turbocharger speed and boost pressure). Engine control unit uses limits and safety circuits to avoid possible danger situation and black box model learns these limitation during training process Predictive model training in drive cycle Precision of model prediction depends on quality of all possible system states examination in the multi-space of all input parameters. Randomly generated signal without subsequent evaluation of trained data cannot guarantee high model precision in whole input parameters multi-space. The highest model precision is demanded in the parameters range where engine is typically used. Which, for the vehicle, is typical driving cycle. Therefore, training process should include especially data sampled from driving cycles and not only randomly generated data. This improvement of predictive model training process could avoid extrapolation of model parameters. Nevertheless, some randomly generated signal should be used to learn limits (low and high speed and low and high load). 2.2 Vehicle Mathematical Model The simulation width of 1-D models is used in wide range of application for its high precision and calculation inexpensiveness. Multidimensional CFD simulation is simplified to 1-D, where only gas dynamics along the pipe axis is captured. Properly calibrated engine model is suitable for predictive model training process for its higher robustness in comparison to real engine measurement especially when randomly generated control signal is used (there is no risk of engine damage). Tested vehicle was medium size truck loaded to total mass of 8000 kg. Vehicle model was built in GT-Suite from Gamma Technologies displayed in Figure. Vehicle driving resistances were assumed with respect to typical values for a vehicle in this category. Figure 2: Vehicle model in GT-Suite. Vehicle uses manually shifted six speed transmission. The driver is substituted by several controllers. They are represented by ordinary PID controllers which are setting engine accelerator pedal position, brake position, clutch position and shifted gear. Shifting algorithm evaluates possible engine speed using all gears, available engine 105

107 torque at all gears and chooses the best available gear from the fuel consumption point of view. The vehicle model is directly connected to implemented 1-D engine model. Driving resistances calculated by the vehicle model results in engine load, which is controlled by the driver model to follow demanded vehicle speed. The input signal of the vehicle model is required vehicle speed (driving cycle) with road slope in dependence on route length. The output signal were engine state parameters. 2.3 Engine Mathematical Model A six cylinder diesel engine was used for the purpose of this work. The engine was equipped with a common rail fuel injection system. Turbocharger with a variable geometry turbine (VGT) was controlled by an electric actuator. The original ECU was replaced by the rcube2 control system, which is fully open and modifiable. The controller allows implementation of MATLAB/Simulink code. Engine maximum power was 194 kw / 2500 min-1 and maximum torque of 930 Nm in wide range of engine speed. Engine 1-D model was calibrated using measured data from engine test cell. The calibration process include properties setting of combustion model, engine friction model, cooling loss model and intake and exhaust pipes to fit experiment data. Detailed structure of 1-D CFD model of intake and exhaust system of an engine can be subsequently simplified to increase computation speed. This Fast Running Models are still well accurate and shows good dynamic response as original detailed 1-D model. Positive experience with FRM models designates them for driving cycle simulations requiring long simulated time. FRM engine model is displayed in Figure. Figure 3: Fast Running Model of engine in GT-Suite. rcube2 engine control system uses large set of look-up maps which are engine speedand torque-dependent. The target value of the engine torque is calculated from the accelerator pedal position with respect to the engine full load curve. The actual engine torque is not measured as a feedback, therefore only target torque value is used as input for control look-up maps interpolation. Maximum injected fuel mass is limited by the calculated air-to-fuel (A/F) ratio to suppress excessive smoke production. The 106

108 algorithm of rcube2 control system was simplified and used in 1-D engine model to assure same dynamic behaviour of engine mathematical model with real engine. 2.4 Route optimization As it was already mention, predictive model training process should cover typical usage of engine, which can be defined as a typical driving on road. Usage of real existing route will improve accuracy of predictive model in typical operation conditions. Typical real driving cycle can be generated based on tracked GPS map data or created by optimization techniques using known limits of the road. The second option was used to generate more routes with different limits to cover wider space of engine usage Route Trajectory Description The very important task in the construction of the optimization algorithm is route trajectory description and the derived route parameterization. The route trajectory description is based on the values of the maximal allowed velocity in each point of the route. The maximum allowed velocity is obtained from legal velocity limit and physical velocity limit as the minimum of them. The legal velocity limit is obtained from GPS navigation data, and the physical velocity limit is obtained from route shape classification (with respect to lateral acceleration), which is finding by turning manoeuvres based on the real trajectory shape (see Figure - the red line represents turning right manoeuvres and the green line represents turning left manoeuvres). The turning manoeuvres are finding by using the oriented angle between two consecutive navigation points [6]. Figure 4: Route trajectory sectioning by the manoeuvres The output from the route trajectory description is sectioning whole prescribed route into smaller parts named sections in basis of set of constant parameters maximum velocity limit, value of slope and the value of rolling resistance coefficient. Each section have to be consisted only from manoeuvres from Figure (it is not possible to divide the manoeuver) Vehicle Mathematical Model for Optimization Algorithm The basic requirement to the vehicle mathematical model for optimization problem solve is the time necessary to one evaluate of the objective function, based on this model. It is appropriate to build mathematical model, based on algebraic functions, Look-up tables and other basic mathematical operation. Fast vehicle model calculation is crucial for successful optimization. Therefore, mathematical model of vehicle with 107

109 electric powertrain was used for optimization minimal energy consumption. Mathematical model is described in [7]. Analogically to the powertrain with combustion engine, the efficiency maps dependent on engine load were used. Nevertheless, the electric powertrain model was less complicated and less computational demanding. The input to the mathematical model is velocity dependence on the travelled distance (output of parametrical segmented route) and the output is energy consumption, necessary torque to the engine output shaft, and necessary rotate per minute Definition of the Optimization Algorithm and Output of the Optimization task The optimization task is possible to divide into three separately parts objective function, optimization constraints and optimization algorithm. The objective function is closely related to the mathematical model of the vehicle, and the special place in the objective function occupies the set of optimization parameters. The set of the optimization parameters is based on the definition the velocity profile in each section, which has been described above. In each mentioned section is considered the same velocity profile, describe by optimization parameters OP and the optimization parameters in each section is shown in Figure. 𝑂𝑃𝑠𝑒𝑐𝑡𝑖𝑜𝑛 = {𝑣1, 𝑎𝑎, 𝑠𝑎, 𝑠𝑑, 𝑎𝑑 } (1) Figure 5: Velocity profile in each section [7] In connection to the velocity profile in section is possible the velocity profile in each section divide into smaller parts named phase, which represents one of the driving modes of the vehicle (acceleration, constant velocity, coasting and deceleration phase). The optimization parameters in each sections represents constant set of parameters, which can describe the velocity profile unambiguously and with respect to constraints shown in Figure. Parametrization allows building velocity profile along whole prescribed route. The total number of the optimization parameters is given by five times the number of sections. The introduced velocity profile along the whole route is the input to the mathematical model of the vehicle and together represent the objective function of the optimization problem. The optimization constraints is possible to divide into two separated parts. The first part is local optimization constraints, which represents constraints fixed on the each section (for example the relation between velocity v1 and v2 is possible only if 108

110 v_2 v_1). The second part of the optimization constraints represents the global optimization constraints, which define the constraints fixed on the whole considered route (for example the total travel time). The choice of the optimization algorithm is especially restricted by the design and features of the objective function. The capital property of the objective function is the continuity of the objective function, which decides to use a specific optimization algorithm. The objective function used in this optimization problem can be discontinuous, because it is possible to omit some phases in the section velocity profile between two consecutive optimization steps, based on the objective function behaviour. 2.5 Example of driving cycle for predictive model training As an example of predictive model training data, one driving cycle simulation was calculated. Demanded vehicle speed was obtained from route optimization with route constraints of road from VTP in Roztoky u Prahy to Prague Dejvice. Quite low vehicle speed was caused by quite low importance of driving time parameter in optimization setting. The simulation of driving cycle was provided with dynamic vehicle model with FRM engine model. The results are displayed in Figure. Engine state variables were used as training set for LOLIMOT models. Figure 6: Result of driving cycle simulation 3 CONCLUSION Predictive models are integral parts of predictive control system and their accuracy is key to successful application to controlled system. Benefit of predictive systems is mainly in precise control of dynamic system without oscillation and overshooting. Improved methodology of predictive model training is based on extension of training process of engine control system by training in driving cycles, which involve typical engine usage. Increased emphasis on typical engine usage will bring up rise of accuracy of the predictive model than training with randomly generated step training signal. The application of predictive control system will be tested in near future. 109

111 4 REFERENCES [1] [2] [3] [4] [5] [6] [7] CARSTEN O., From driver models to modelling the driver: what do we really need to know about the driver? P.C. Cacciabue (Ed.), Modelling Driver Behaviour in Automotive Environments, Springer, London (2007), pp DOLEČEK V., FLORIÁN M., ŠIKA Z. Model Based Nonlinear Predictive Control of IC Engine in Unsteady Operation Mode, XLVI. International Scientific Conference of the Czech and Slovak Universities and Institutions Dealing with Research of Internal Combustion Engines, 2015, ISBN SIKA Z., VALASEK M., FLORIAN M., MACEK J., POLASEK M. Multilevel Predictive Models of IC Engine for Model Predictive Control Implementation, SAE Technical Paper, 2008, paper MACEK J., POLASEK M., SIKA Z., VALASEK M., FLORIAN M., VITEK O. Transient Engine Model as a Tool for Predictive Control, SAE Technical Paper, 2006, paper DOLECEK V., MACEK J., FLORIAN M., SIKA Z. Model Based Nonlinear Predictive Control of IC Engine in Unsteady Operation Mode, JSAE Annual Congress,Yokohama 2015, paper S DENK P., STEINBAUER P., ŠIKA Z., MACEK J., MORKUS J. Route Segmentation Designed for Optimization of the Vehicle Behavior and Control by Adaptive Cruise Control, 21st Workshop of Applied Mechanics - Proceedings. Praha , pp ISBN STEINBAUER P., MORKUS J., DENK P., MACEK J., ŠIKA Z., The Development of Ride Optimization based on Vehicle Road knowledge, 18th Workshop of Applied Mechanics. Praha ISBN ACKNOWLEDGEMENT This research has been realized using the support of: o Josef Bozek Competence Centre for Automotive Industry, TE o European social fund within the frame work of realizing the project "Support of inter-sectoral mobility and quality enhancement of research teams at CTU in Prague", CZ.1.07/2.3.00/ o EU Regional Development Fund in OP R&D for Innovations (OP VaVpI) and The Ministry of Education, Youth and Sports, Czech Republic, project # CZ.1.05/2.1.00/ Acquisi-tion of Technology for Vehicle Center of Sustainable Mobility o The Ministry of Education, Youth and Sports program NPU I (LO), project # LO1311 Development of Vehicle Centre of Sustainable Mobility This support is gratefully acknowledged. 110

112 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES OPTIMUM LIMITS OF MOTOR VEHICLE DRIVING Jan Macek1, Josef Morkus2 Abstract Basic equations of road energy consumption for different motor vehicles. Model-based tool for energy and range optimization, suitable for both electric and ICE powered vehicles, if speed-time (or distance) dependence tachogram is defined as input. Regression model of brake efficiency and torque-speed constraints for all primemovers and kinetic energy recuperators. General optimization variables for route energy requirements assessment. Examples of optimization results with different constraints. Upper and lower estimates for extreme achievable times of driving for limiting speeddistance schedules minimum time or minimum energy consumption - determined by simple analytical approach for all sections of a route. Significance of tachogram approach, excluding human driver model, as an input for autonomous driving. 1. INTRODUCTION Importance of energy efficiency of vehicles has risen due to significance of greenhouse gases emitted from fossil fueled internal combustion engine (ICE) powertrains and very limited range of battery electric vehicles (BEV). On one hand, many systems for powertrain control with different level of predictive features are currently under development e.g., [6], [7] amending standard cruise control (adaptive cruise control keeping safe distance from vehicle ahead) by extended e-horizon. On the other hand, adaptive features of energy consumption optimization systems are being developed to react to changes of traffic density, weather and other external conditions for driving, especially from infrastructure. The activity of a driver is solved by some type of humanmachine interaction (HMI). Energy consumption is directly linked to carbon dioxide production using fossil fuel composition and if necessary well-to-tank (WTT) production, important although frequently neglected for BEV s. 1 Prof. Ing. Jan Macek, DrSc., Czech Technical University in Prague, Faculty of Mechanical Engineering, Centre of Vehicles for Sustainable Mobility, Technická 4, Praha 6, Czech Republic, jan.macek@fs.cvut.cz 2 Ing. Josef Morkus, CSc., as above, josef.morkus@fs.cvut.cz 111

113 Following this approach, optimization of eco-driving [8], [9] may be done by coupling simulation models of vehicle longitudinal dynamics, powertrain torque/speed and efficiency features, traffic and infrastructure models and last but not least the driver behavior. Such model is not very suitable for simulation of future autonomous driving systems, which should de-couple a driver from the whole chain of systems. Moreover, the model of a real driver behavior does not help in finding extremes of energy consumption for preferred time of trip at one side and cost of fuel (or CO 2 production) at the opposite side. The perfect copying of driving schedules according to standardized test cycles, e.g., WLTC, is not simple with a model of a real driver, as well, e.g., [9]. Other issues occur using the standard model road-vehicle-powertrain-driver, as it is, if optimization procedure is applied to the whole trip along significant distance. Such approach is important especially for full hybrids with considerable capacity of batteries. Long trips can be optimized using individual parameters of route sections only if very powerful cloud computing facilities are available. Operative adaption of optimum speed schedule cannot be done in fulfilling short real time feed-back. The goal of this contribution is to show newly developed tools for fast optimization of a trip speed schedule, which is capable of taking main energy consumers into account, namely ICE or electric powertrain, HVAC coupled to powertrain cooling under thermal management system, waste-heat recovery systems, etc. Selected parts of a complex optimization tool will be described due to lack of space here. The main ideas of a tool for energy consumption optimization presented are definition of speed-distance (or speed-time) route profile with known slopes, adhesion limit, rolling resistance, cargo load, weather (visibility, adhesion coefficient reduction, air density, wind, heat losses from a car body, sun radiation, etc.) use of generally valid optimization parameters (e.g., acceleration or power limits for speeding-up or braking, going uphill or downhill, use of adhesion limit, use of maximum speed) and constraints (e.g., time of trip) control of driving not depending on a driver model but fulfilling speed schedule known in advance in limits of above mentioned vehicle dynamic parameters, i.e., tachogram approach aiming at finding the best possible solution for energy consumption respecting constraints; it can be used even for hybrid vehicle cases - [10]. 2. OPTIMIZATION TOOL FOR ENERGY CONSUMPTION The tool for optimization, fulfilling main items of the list in introduction chapter above, has been described in some papers published in past - [4], [5]. It is based on standard vehicle motion relation, taking into account rolling resistance, slope, air drag (including wind) and acceleration or deceleration inertia forces (with rotating mass inertia impact). Thermal management of ICE or electric motor with control equipment and heat, ventilation and air conditioning of car body interior (HVAC) including heat pumps is described by energy conservation differential equations, taking into account heat sources caused by losses or waste heat recuperation (WHR) and sun radiation or heat removal from cooling or HVAC systems using convective heat exchange - [1], [3]. 112

114 The route section is defined by distance and basic maximum speed, determined as a minimum from legal speed limit, reasonable lateral acceleration considering route curvature, real speed achieved in current traffic jams, stops caused by traffic lights and stops for loading or charging/re-fueling with specified time. The internal speed limit may be reduced by available power limit. Especially for ICE s, achievable torque of ICE is respected. The gear of standard shifted gearbox is selected according to recommended gear schedule or according to the best energy consumption, downshifting if reached torque is not sufficient. Aiming at fast algorithms, the differential equations are integrated analytically in a (short) section of a route with assumption of parameters locally fixed. Between route sections, parameters depending on integrated values are changed. The route is initially treated by backwards proceeding predictor steps starting with final speed (in most cases just zero). It finds the maximum speed at outlet of every section, which ensures that the maximum speed if any of the following section would be satisfied by braking with maximum allowable deceleration. Then the forward going corrector step predicts the achievable speed respecting the generally valid optimization parameter, e.g., power limit reduction parameter, adhesion limit reduction parameter or basic speed limit reduction parameter, etc. Those normalized parameters related to nominal values for every section are subjected to optimization with constraints, e.g., time of trip, minimum or maximum car body interior temperatures, WOT torque of ICE or electric motor temperature limit. 3. EXAMPLES OF OPTIMIZATION RESULTS USING SIMPLIFIED MODEL OF A BEV The model developed for a BEV, as described above, was based on general normalized optimization parameters valid for the whole trip. The route of the length of almost 10 km with elevation difference of 45 m (going uphill) is used for illustration of this simple model possibilities. Figure 1: Testing route features slope (unfiltered from GPS data, triangles) and legal maximum speed (dotted line) together with speed achievable for durable motor power of 12 kw (circles). Speed achieved is plotted as dashed line for w 113

115 The optimization was done for different times for a trip and different cargo loads. Different air temperatures and different HVAC control strategies were tested. Going downhill, HVAC power is covered preferentially by recuperation of potential energy, considering the control range of interior temperature. The most of the following examples illustrates cases without HVAC use, however. Low power limit for a motor reduces possible positive acceleration especially at higher speeds, as visible from achieved speed w - Figure 1. Braking recuperates in the best way kinetic energy, if intensive deceleration is used. For the cases presented, it was assumed that 60% of braking energy is recuperated, the rest being braked by friction. Other trip conditions for optimum motor power and its acceleration margin for trip time of 14 min are plotted in Figure 2 - Figure 3. It is worth-while mentioning that 10 kwh of energy consumption is close to equivalent of 1 dm3 of diesel oil. The optimization was done for altogether 12 general normalized trip parameters using ModeFrontier optimization code and genetic algorithm option. It was found that the decisive parameter for a fixed trip time is relative reduction of maximum section speed and the bottom speed limit, below which the reduction of section speed is not applied. Then, the limit of motor power with pre-set margin for active acceleration should be fixed to achieve the trip time. Other section parameters, as braking deceleration or relative use of adhesion are of less importance. Figure 2: Brake power of motor P ave with temporary overload against 12 kw limit and energy consumptions from motor clutch to SOC of battery; other data from Figure 1 The resulting energy consumption lies inside the limits of the shortest trip time with the highest accelerations and the globally best energy consumption, balancing tractive effort energy (increased for higher speeds due to drag) and powertrain efficiency, which features optimum between medium (for motors) and high load (for ICE s), depending on the type of primary mover. It is analyzed below. The Pareto curve for the basic case is illustrated at Figure 4 for trip times shorter than 14 min. The curve can be used for illustration of difference between optimum case and cases selected stochastically although reasonably during genetic optimization process. Another case with load increased by 300 kg is plotted in the middle of Figure 4. The 114

116 case of heating using heat pump connected to motor cooling circuit at external temperature of -10 C is presented on right. Pareto curves can be used for assessment of improvement potential achieved by optimization, which is typically more than 10%. 4. THE SHORTEST TRIP TIME OR THE LOWEST ENERGY CONSUMPTION LIMITS - GLOBALLY VALID EXTREMES Assessing energy consumption for different route sections, the dependence on control system (or driver actions) can be significant. For using tachogram, the driver influence on longitudinal vehicle dynamics is excluded, being perturbation only in the case of future autonomous driving systems. Nevertheless, there are idealized extremes of energy consumption for a route section driving for the shortest time and driving for the best energy consumption, in both cases with or without uniform speed. The accelerations or decelerations in real cases increase or even reduce the real energy consumption and increase the shortest time or the best energy efficiency, but even both simplest extremes are worthwhile for comparison of achieved results to theoretical limits of time or of energy consumption. Figure 3: SOC development and motor/generator efficiency Figure 4: Pareto curve for dependence of SOC energy consumption in kwh/100 km of trip time in mins; constraining time 14 min. Basic case - left, extra cargo load 300 kg - in the middle, HVAC influence on energy consumption for outer air temperature of -10 C, initial interior temperature of -10 C and range for heating control 18 C-22 C - right 115

117 Whereas the theoretical shortest time can be estimated directly from maximum basic speed in sections and the distance of them, the theoretical limit of energy consumption has to be determined from wheel energy consumption and ICE/motor/generator efficiencies. On the one hand, the simple use of wheel consumption does not yield any useful optimum the lower speed, the better due to air drag. On the other hand, too low powertrain power for very low vehicle speed yields low powertrain efficiency and the task of finding the optimum speed for the tank-to-wheel energy consumption is not trivial more. The following chapter describes this process for any primary mover of a vehicle. The aim was to find general algebraic representation of all primary movers brake efficiency, which simplifies the finding of optima using generally valid formulas. 4.1 Analytical Tools for Investigating Optimum Energy Consumption Brake efficiency (mechanical power at clutch output/energy input) is described using approximation of additive friction and internal (outlet enthalpy or other thermal and internal resistance losses) he We pe Pe We W F QO p e p F q O Pe PF Q O 1 P Q O 1 F Pe,nom Pe,rel (1) Loss powers depend generally on machine power, load (forces and torques or currents) and speed. Non-dimensional form tends to be qualitatively size-independent. For this purpose, losses can be related to power (mostly relevant for electric motors) or to works (bmep or torque is relevant for ICE from experience). Using the model with additional multiplication by of relative angular speed powered to exponent xg generalizes it: y1 y Pe, rel Pe, rel A2 A3 rel AO A1 x PF Q O relg Pe,nom rel rel x z1 z A4 rel A5 Pe, rel A6 Pe, rel x 1 Pe,rel he, M Pe,rel y1 y Pe, rel Pe, rel AO A1 A2 A3 rel rel rel x z1 z A4 rel A5 Pe, rel A6 Pe, rel x 1 (2) x g rel For x1=y1=z1= xg=1 a z=0, x=y=2, Equation ( 2 ) yields friction-like standard formulas for ICE, whereas for x1=x, y1=y, xg=0 a z1=0, z=1 it yields electric motor representation, e.g., as in [2]. It is simply possible to find the best model by applying linear regression to A coefficients and optimization (non-linear regression) to exponents, which has confirmed the basic experience mentioned above. Electric machine power should be decreased by inverter or converter losses (if used), dependent simply on power. Electric generators with mechanical input at clutch but electric output are more accurately represented by reciprocal approach, which yields good results even at low generator efficiency. Input power is mechanical (clutch) one, output is electric power, which should be reduced by inverter/converter powers, as mentioned above. 116

118 hg 1 PF,G Q O,G PG,nom PG,rel PG,rel hg yg 1 yg PG,rel A A PG,rel A3,G rel A2,G OG 1G rel rel z 1G z A x G A P A6,G PG,rel G 5,G G,rel 4,G rel x 1G x g,g rel (3) 45% % % % % % eta e eta e reg EngineTorque [N.m] WOT Torque [N.m] WOT Torque reg [N.m] 15% 10% EngineTorque [N.m] Brake Efficiency PG,rel It is important for BEV and all hybrids. An example of ICE efficiency representation is in Figure 5, motor/generator efficiency fields are in Figure EngineSpeed Figure 5: ICE efficiency representation by regression and WOT torque (efficiency curves at fixed torque, dots show the values of relative torque in the total map used for calibrations of regression) 100% 100% 90% 90% 80% 60% eta G reg 50% eta G 80% eta M [1] eta G [1] 70% 20% P rel G*(1-eta G) [1] eta M reg 60% eta M 40% 30% -2 70% 0 50% 40% P rel M [1] Figure 6: Electric motor/generator efficiency representation by regression for different relative speeds in dependence of mechanical power input (generator) or output (motor) The global efficiency extreme should find for road energy consumption, i.e., it is not dependent on single powertrain efficiency but additionally on tractive energy 117

119 consumption. In the case of stepwise-shifted gearbox, the gear ratio used for the current route section with prescribed and achieved speed should be applied or just the gear for the best energy consumption may be simply investigated by try-and-error method. Again, it should be stressed here, that the aim is to find utmost limits, not just energy consumption for a specific case with different driver behavior, etc. It is important for comparing the achieved results of different strategies with the limits and for preliminary eco-routing calculations. Both cases without kinetic energy consumption/recuperation due to accelerations or just acceleration-less case may be optimized. Optimum speed satisfies minimum of K rel function Eroad. It was found, that adjustable efficiency penalty function 1 e may exclude tending of optimization to too low speeds in the case of ICE. The penalty function may be switched-off for high K, if necessary. In the case of route section of length s and rated reference (nominal) values 3 Pe,rel K 1 rel K 2 rel M e,rel Pe,rel 2 K 1 K 2 rel rel 2 Fe K 1 K 2 rel Fe,nom hem 1 htm,nom c x S x nom 2 w nom g mv eh m load,nom r cos nom sin nom 2 Pe,rel Pe,rel J xg rel (4) 1 x relg J 1 3 K 1 rel K 2 rel 3 1 A1 K 1 K 2 rel2 1 A2 K 1 K 2 rel2 J AO A5 K 1 rel K 2 rel z A3 rel x 1 A4 rel x y 3 A6 K 1 rel K 2 rel y z whereas g mv eh m load r cos sin a m red m load htm g mv eh m load r cos sin a m red m load K 1 htm i Ti F e,nom K1 K 2 2 cxsx w nom 3 2 htm i Ti F e,nom K 2 2 cxsx w nom K 2 i Ti F e, nom 2 htm i Ti2 (5) Then, road energy consumption and condition for the extreme including penalty function is 118

120 E road,kwh / 100 km x 1 relg J Fe 1 M e,rel (6) d 1 e K rel E road! y 0 d rel y 2 K 2 rel 1 e K rel Fe,i 1 e K rel hem x relg J 3 K 1 rel K 2 rel d d rel K rel KF e e (7) hem Derivatives required by the relation above can be found analytically including the derivative of y function, which is needed for fast solution of nonlinear extreme condition by Newton s method. The iteration for optimum relative engine/motor speed is rel,i 1 rel,i y rel,i dy d rel rel (8) rel,i A similar approach is applied for optimum recuperation conditions at a generator. In the case of HVAC at BEV, description of a system with thermal capacities and resistances is based on the set of differential equations. Heat pump is modelled by Carnot heating or cooling factors, depending on inlet and outlet temperatures, empirical correcting factor and mechanical efficiency of a compressor. It is of advantage, if input heat for heating is not extracted from the outer air, but from the electric equipment cooling circuit. Moderate WHR is achieved by this way. 4.2 Examples of Globally Optimized Section Speed The procedure described by Eqs. ( 7 ) and ( 8 ) has been applied to horizontal route and kg heavy car with up-to-date small diesel and five-gear transmission according to Figure 5, not taking acceleration power into account. The results are plotted in Figure 7. Figure 7: Dependence of road energy consumption, ICE efficiency and achievable optimum speed on relative speed for an ICE and five-gear transmission powertrain; arrows show the direction to higher transmission ratio (lower gear) 119

121 The overall energy consumption is very good despite the fact engine does not use its maximum efficiency due to very low load. Vehicle speed is rather low to suppress air drag. The upper branch of energy consumption is valid for high transmission ratio (low gears) and low vehicle speed associated with low engine efficiency. The best results of energy consumption are achieved at moderately loaded engine with relatively high ICE efficiency of 33%, higher gear and vehicle speed of approximately 44 km/h. Those results illustrate the well-known fact: the engine, assigned to a vehicle respecting the standard requirements of maximum power for good active safety, is too large, if the most frequent mode of operation is used. It requires low power only. Acceleration and deceleration phases during initial phases of WLTC are plotted in Figure 8 and Figure 9, using pre-defined shifting schedule (range of speeds for every gear with some overlap between subsequent gears) or shifting to gear, which ensures the best energy consumption for the current vehicle speed and tractive force. Downshifting is applied in the cases, if ICE torque is not sufficient for combination of load and engine speed, determined according to the criteria mentioned above. The ideal shifting model features no time delay and losses due to clutch slip. Figure 8: WLTC velocity schedule for the start of driving cycle; globally optimum vehicle speed w opt eff has been determined for all sections of 5-10 m length including acceleration power; standard shifting above, optimized shifting below 120

122 Acceleration power has been taken into account for both examples, which added some engine load increasing engine efficiency together with tractive wheel force. The optimum vehicle speed is the same while engine is loaded by short period of acceleration increasing vehicle speed to the prescribed value. In order to fulfill WLTC demands, the rated power limit of kw was needed, otherwise the total time of WLTC was not kept at required level. The rest of any time step of initial phase of WLTC was run with constant speed before subsequent acceleration, i.e., the needed power was low. Just due to this period there is a difference between both shifting schedules. In the case of standard shifting, the gear used for acceleration was changed to immediately higher one or even left, which reduced efficiency of temporary unloaded engine. In the case of optimum shifting, up-shifting to highest gear often occurred, which loaded the engine at low speed and increased its efficiency, keeping optimum vehicle speed high. It is clearly visible for acceleration phases in Figure 8. Braking phases have not yet been treated finally, therefore results are not comparable. Figure 9: Working points of ICE in total map for modes of operation from Figure 8; standard shifting above, optimized shifting below The same results are plotted as working points in engine speed-torque map (Figure 9). Points at and close to WOT curve used for accelerations are unchanged. The steady speed operation is compressed to low engine speed in the case of optimum shifting. 121

123 The faster acceleration may be better in some cases, if an engine is able to react without turbo-lag. More loaded engine features higher efficiency and energy consumption for overcoming inertia forces does not depend on time of acceleration to the same resulting speed. The energy consumption is determined by difference of kinetic energies, namely. The significant difference in total energy consumption during this part of WLTC can be found in description of the both figures. The examples have to show the potential of new fast method of finding limits for the best energy consumption on the road. 5. CONCLUSION New tool for finding optimum trip schedule based on optimization of simple model results has been deduced. Powertrain efficiency is described by algebraic non-linear regression model in the form, general enough for ICE, compressor or expander, emotor or e-generator. Torque WOT may be applied if necessary, especially in the case of ICE. If gear shifting is available, gear selection may use fixed shifting schedule defined by vehicle speed ranges and downshifting. Due to the fast efficiency model, shifting schedule for the best immediate road energy consumption may be simply applied, again with downshifting, if necessary. The model excludes a driver behavior, controlling engine power and speed directly according to the target vehicle speed and tractive force demands following it in a route section. Optimization variables are general for the whole trip, being based on the maximum power or use of upper speed limit, on margin against slipping at given adhesion coefficient during acceleration or braking, and some other similar parameters. Genetic algorithms are suitable to find the optimum fulfilling constraints as maximum trip time or maximum e-motor temperature. Especially for genetic algorithms, there is a necessity to find extreme levels of energy consumption. The algebraic model of energy consumption yields fast results of globally optimal speed limits. It can be used for amended eco-routing and for fast optimum distribution of power between ICE and e-motor in hybrid powertrains, which will be described in the future publication. The difference between local optimum, fulfilling constrained trip time, and general optimum offers simple assessment of further potential to improve transport efficiency. ACKNOWLEDGEMENT This research has been realized using the support of EU Regional Development Fund in OP R&D for Innovations (OP VaVpI) and The Ministry of Education, Youth and Sports, Czech Republic, project # CZ.1.05/2.1.00/ Acquisition of Technology for Vehicle Center of Sustainable Mobility, by the National Programme of Sustainability, granted by The Ministry of Education, Youth and Sports, Czech Republic, project LO1311 Development of Centre of Vehicles for Sustainable Mobility, by European Union Framework Programme 7 project IMPROVE and Horizon 2020 projects IMPERIUM and ADVICE. This help is gratefully acknowledged. 122

124 SYMBOLS A cx E F g K it J M m P Q regression coefficient drag coefficient road energy consumption [J.m-1; kwh/100 km] tractive force at wheels [N] gravity acceleration [m.s-2] coefficient variable part of transmission ratio dimensionless loss power torque [N.m] mass [kg] power or normalized (dimensionless) power [W, kw] heat flux (including outlet enthalpy for ICE) [W] Sx s w x y drag reference area distance [m] vehicle velocity [m.s-1; km.h-1] regression exponent regression exponent; unknown in non-linear equation for road energy consumption minimum regression exponent angle of road slope [deg] efficiency rolling resistance air density [kg.m-3] angular speed [rad.s-1] z h r SUBSCRIPTS e F G g i load M m nom O red rel T veh mechanical brake power or torque for engine/motor, input power or torque for generator friction generator global iteration step cargo load engine/motor mechanical, friction nominal, rated outlet (removed from thermal machine) reduced (including rotational masses for inertia) normalized (related to nominal or reference value) transmission vehicle ACRONYMS BEV battery electric vehicle 123

125 bmep bsfc ICE HVAC SOC TTW WHR WLTC WTT WTW brake mean effective pressure brake specific fuel or energy consumption internal combustion engine heating-ventilation-air-conditioning state of charge tank-to-wheel waste heat recuperation world harmonized light-duty testing cycle well-to-tank well-to-wheel REFERENCES MACEK J., STEINBAUER P., MORKUS J., BARÁK A. Procedures for determination of the current vehicle parameters, Report to IMPROVE deliverable 3.4, CTU Prague 2014 [2] ČEŘOVSKÝ Z. Regression of e-mobile motor efficiency. Internal report, CTU in Prague, 2006 [3] MACEK J., MORKUS J., DENK P., STEINBAUER P., ŠIKA Z., BARÁK A. Energy management algorithm for vehicle operation, Report to IMPROVE deliverable 3.6, CTU Prague 2015 [4] STEINBAUER P., MACEK J., MORKUS J., DENK P., BARÁK A. Dynamic Optimization of the E-Vehicle Route Profile. SAE [5] STEINBAUER P., HUSÁK J., PASTEUR F., DENK P., MACEK J., ŠIKA Z. Predictive Control of Commercial E-Vehicle. Powertrain Modelling, Control and Calibration, University of Loughborough 2016, pap. P56/N26 [6] FU MENGYIN, JIE LI, ZHIHONG DENG A practical route planning algorithm for vehicle navigation system, Intelligent Control and Automation, WCICA Fifth World Congress on. Vol. 6. IEEE, [7] MINETT, C. F. et al. Eco-routing: comparing the fuel consumption of different routes between an origin and destination using field test speed profiles and synthetic speed profiles. In: Integrated and Sustainable Transportation System (FISTS), 2011 IEEE Forum on. IEEE, p [8] SABOOHI Y., FARZANEH H. Model for developing an eco-driving strategy of a passenger vehicle based on the least fuel consumption. Applied Energy, 2009, 86.10: [9] BARTH, M., BORIBOONSOMSIN, K. Energy and emissions impacts of a freeway-based dynamic eco-driving system, Transportation Research Part D: Transport and Environment, 2009, 14.6: [10] BADIN, F. LE BERR F., CASTEL G., DABADIE JC., BRIKI H., DEGEILH P., PASQUIER M. Energy efficiency evaluation of a Plug-in Hybrid Vehicle under European procedure, Worldwide harmonized procedure and actual use, EVS28 KINTEX, Korea, May 3-6 [1] 124

126 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES MEASUREMENT WITH IMPROVED TUMBLE METER Petr Hatschbach1, Jiří Schmidt2 Abstract The goal of this article is to present a first results of the measurement with the improved design of own tumble meter. Compared to the previous tumble meter the new one is equipped with a miniature torque transducer to direct measure of torque on shaft with measuring square mesh element. The new tumble meter can be used for range of cylinder bore from 60 to 100 mm by changing of replaceable cylindrical inserts. An important dependence of drag coefficient of square mesh on Reynolds number was determined. Results of the measurement with tumble meter were compared with the results evaluated from 3D CFD data. 1. INTRODUCTION The steady flow testing is often used experimental method for experimental evaluation of intake port flow properties: discharge or flow coefficients, swirl and tumble ratio. Simple integration methods for swirl and tumble characterisation are usually based on direct measurement of some component of angular momentum flux. Unfortunately this experimental techniques use coarse simplification and measured values do not give satisfactory results particularly for tumble measurement. An example of problems with the tumble measurement was discussed in [1]. Our own tumble measuring method and evaluation procedure have been described. CTU tumble meter is shown in Figure 1. Into the cylinder is inserted a square mesh connected with the hoop with shaft perpendicular to the cylinder axis. Measured quantity is the torque required to maintain the mesh in position perpendicular to the cylinder axis. This torque is proportional to the tumble (cross) component of the angular momentum flux. Subsequently the tumble number and the tumble angular position are evaluated. In [1] described differences between experimental a simulated results were not satisfactorily explained by means of numerical simulation of experimental evaluation. 1 Petr Hatschbach, CTU in Prague, Faculty of Mechanical Engineering, Vehicle Centre of Sustainable Mobility, Roztoky, Prilepska 1920, Petr.Hatschbach@fs.cvut.cz 2 Jiří Schmidt, CTU in Prague, Faculty of Mechanical Engineering, Technicka 4, Praha 6, Jiri.Schmidt@fs.cvut.cz 125

127 As a possible source of different results were identified: influence of mesh hoop, incorrect measurement of torque on the shaft, strong dependence of drag coefficient of mesh on Reynolds number and its uncertainty. Figure 1: CTU tumble meter old design 2. NEW TUMBLE METER DESIGN Main disadvantages of the old tumble meter design were reading of measured torque on a scale, possibility of influence on the flow by the hoop and usability only for one diameter of cylinder. Figure 2: CTU tumble meter new design The new tumble meter design Figure 2 - solves these problems. Manual reading of reaction torque on a scale is replaced by miniature reaction torque transducer that 126

128 converts torque into an electrical signal. Transducer is connected to the shaft of mesh hoop using a flexible metal bellows-type coupling. Furthermore, the tumble meter has 3 replaceable cylindrical inserts that have two functions. Firstly, by replacing these inserts (and only these inserts), it is possible to measure with cylinder bore from 60 to 100 mm. And secondly, the mesh hoop is hidden between two inserts and does not affect the flow in the cylinder. The whole measurement and evaluation procedure is the same or similar as for the old tumble meter [1]. Because a mesh with slightly different dimensions was used (wire diameter 0.22 mm, wire distance mm), it was necessary to re-measure the dependence of the drag coefficient on the Reynolds number. The measurement of the drag coefficient was performed on the steady flow rig as a measurement of the pressure drop on the mesh located in the rounded hole in the settling tank. Drag coefficient was evaluated from measured difference between ambient pressure and static pressure in the settling tank. The approximation function in the form 𝑐𝐷 = 𝑎 𝑅𝑒 𝑏 + 𝑐 using MATLAB was evaluated: 𝑎 = 18.8 ; 𝑏 = ; 𝑐 = 1.49 Figure 3. This function fits measured data much better than previously used function 𝑐𝐷 = 𝑎 𝑅𝑒 𝑏. The Reynolds number is based on the wire diameter. Figure 3: Dependence of drag coefficient of square mesh on Reynolds number 3. RESULTS COMPARISON Measurements was done with a 4-valve SI engine (bore-to-stroke ratio 0.88, compression ratio 10.2) in two configurations: both intake valves opened (tumble only, no swirl) and only single valve opened (high swirl and tumble). The simulated data was used from the study of steady flow test bench simulation [3] (calculated in 3D CFD code AVL FIRE, PANS turbulence model, 4 boundary layers and 0.15 mm boundary layer thickness), As in previous work [1], 3 type of evaluation of tumble number was done: 1. from experiment: with new tumble meter - (𝑛𝑗𝑝 𝑛)𝑒𝑥𝑝2 and for comparison with old tumble meter (𝑛𝑗𝑝 𝑛)𝑒𝑥𝑝 2. direct evaluation from velocity field - (𝑛𝑗𝑝 𝑛)𝑑𝑖𝑟𝑒𝑐𝑡 3. simulation of tumble meter evaluation from CFD data - (𝑛𝑗𝑝 𝑛)𝑠𝑖𝑚_𝑒𝑥𝑝 127

129 Figure 4: Relative tumble number and tumble angular position - both intake valves open 128

130 Figure 5: Relative tumble number and tumble angular position - single intake valve open 129

131 The results are plotted in Figure 4 (both intake valve open) and in Figure 5 (single intake valve open). The top graph shows the tumble number and the bottom graph shows tumble angular position (180 correspond to longitudinal axis of the head). Compared to previous results, better qualitative and quantitative compliance was achieved. Good correspondence is in the important case of both valve open with medium and higher valve lift for new measured (exp2) and experiment simulated (sim_exp) results. But for low lift the correspondence is still not satisfactory. The positive is better qualitative compliance for experiment simulated and direct evaluated results. This is probably due to the new correct determination of dependency of the drag coefficient on the Reynolds number. 4. CONCLUSION The new design of tumble meter allows more convenient measurement, with reduced subjective factor and with mesh hoop influence elimination. For higher valve lift was achieved a better qualitative agreement of experimental and simulated results. The reason for the quantitative differences between the results of the direct evaluation and the measurement or simulation results can be in the large dependence of the drag coefficient on the Reynolds number. Hence, the future work should be aim at that. ACKNOWLEDGEMENT This research has been realized using the support of Technological Agency, Czech Republic, programme Centres of Competence, project TE : Josef Božek Competence Centre for Automotive Industry. This research has been realized using the support of EU Regional Development Fund in OP R&D for Innovations (OP VaVpI) and The Ministry of Education, Youth and Sports, Czech Republic, project CZ.1.05/2.1.00/ : Acquisition of Technology for Vehicle Center of Sustainable Mobility. This research has been realized using the support of The Ministry of Education, Youth and Sports program NPU I (LO), project LO1311: Development of Vehicle Centre of Sustainable Mobility. All the help has been gratefully appreciated. REFERENCES [1] HATSCHBACH, P.: Differences between Tumble Measurement and Simulation Methodology, In: KOKA XLVII. International Scientific Conference of the Czech and Slovak Universities and Institutions Dealing with Research of Internal Combustion Engines. Brno: Vysoké učení technické v Brně, pp ISBN [2] HATSCHBACH P.: Měření rychlostního pole ve válci spalovacího motoru pomocí integrálních metod a laserové dopplerovské anemometrie, Disertační práce, ČVUT v Praze, fakulta strojní, 1995 [3] VITEK O., TICHANEK R., HATSCHBACH P.: Application of LES, PANS and RANS to a Case of Intake Channel Steady Flow Test Bench, In: Journal of Middle European Construction and Design of Cars, Vol. 13, No. 3, 2015, pp ISSN [4] AVL FIRE Manuals, AVL List GmbH, Graz 130

132 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES DYNAMIC CYLINDER DEACTIVATION FOR A SPARK IGNITION ENGINE Ondrej Bolehovsky1, Jan Macek2 Abstract Significant efficiency loss of the spark ignition engines with stoichiometric mixture stems from the throttling that occurs at the partial load of the engine, which is the most utilized mode during the engine operation. Downsizing and downspeeding have enabled to partially decrease this loss, however, with the upcoming trend of "rightsizing", de-throttling methods are again of higher interest. One of the approaches is cylinder deactivation, which has been used in static mode yet, i. e., usually half of the even number of cylinders is switched-off during part load operation (and lower engine speed). Another approach in the cylinder deactivation is its dynamic use, i. e., all cylinders continuously switch between the activated and deactivated mode. This is suitable particularly for odd number of cylinders and does not exist in production yet. The paper deals with 1-D simulation of this kind of cylinder deactivation, compares different approaches to the gas exchange process of such deactivation, points out the benefits and advantages over other partial load control methods and also shows simulation issues that emerge with the dynamic deactivation. 1 INTRODUCTION Upcoming emission standards and related new driving cycles (WLTP) introduce higher loads of the internal combustion engine (ICE) during the drive cycle (also due to the new gear shifting procedure), which in turn leads engine manufacturers to reassess the level of engine downsizing. Highly downsized engines, especially gasoline ones, exhibit low fuel consumption at partial load, however, at higher load the emissions and fuel consumption are far from being favourable. This is why we are hearing more often terms as rightsizing, which means to find some reasonable level of downsizing, at which the engine fulfils both laws and also customers requirements. Rightsizing means going back to higher engine displacements and takes away the advantage of favourable fuel consumption at partial loads. Particularly for gasoline engine with stoichiometric charge mixture, shifting of operating points to higher specific Ondřej Bolehovský, Czech Technical University in Prague, Faculty of Mechanical Engineering, Technická 4, Prague 6, Czech, Republic, Ondrej.Bolehovsky@fs.cvut.cz 2 Jan Macek, Czech Technical University in Prague, Faculty of Mechanical Engineering, Technická 4, Prague 6, Czech, Republic, Jan.Macek@fs.cvut.cz 1 131

133 loads and thus avoiding the significant throttling losses has helped to gain fuel economy in the low demanding NEDC procedure. The more downsized the engine is, the more benefit at partial load the engine gains. When the engine displacement grows, the fuel efficiency at light loads thus deteriorates. It is therefore necessary to brush up the de-throttling measures that had been scrutinized before the advent of downsizing. Among others, these measures include variable valve timing (late/early intake valve closing), dilution of the cylinder charge (residuals, fresh air), advanced combustion concepts (HCCI) or variable displacement engine (VDE), realized usually by cylinder deactivation. Cylinder deactivation (CDA) is a technique that was introduced already in the 70 s of the 20th century. It was implemented in the first controlled automotive engines in the USA and the cylinders were deactivated via hydraulic rocker deactivation made by Eaton [2]. There have also been several other cases of such deactivation, however, expansion of this technique occurred after year 2000, mainly due to the progress in electronic control that allowed smoother reengagement of all cylinders. Cylinder deactivation has been usually used for V-type engines which allows to deactivate certain sets of even number of cylinders. With the effort to minimize fuel consumption, this deactivation was also introduced at in-line 4-cylinder engines. With the era of downsizing, in-line 3-cylinder engines became popular again and consequently questions of cylinder deactivation for them arrived as well. There are two possibilities at a 3-cylinder engine, either permanently deactivate one cylinder (static cylinder deactivation SCDA ) thus only 1/3 of the engine displacement, uneven firing sequence, NVH issues or make each cylinder work only certain multiples of the 4-stroke cycle. For an odd number of cylinders, it is advantageous (even firing sequence) that every second cycle is skipped/deactivated, which has also become called dynamic cylinder deactivation ( DCDA ) or rolling cylinder deactivation. There is not many publications dealing with the dynamic cylinder deactivation. In 2013, Tula Technology unveiled their specific cylinder deactivation called dynamic skip fire [2]. Their control system decides whether to fire or skip a cylinder immediately prior to each firing event, i. e., selects the sequence of combustion events dynamically based on torque demand and NVH characteristics. Tula together with Delphi claims 10% to 20% fuel economy improvement compared with conventional spark ignition engines depending on the engine and driving conditions (drive cycle). As presented in [4], Eaton and PSA carried out simulations that revealed fuel consumption benefit of up to 4% on the WLTP cycle and about 1.5% advantage over fixed cylinder deactivation. Ford also dealt with potential of cylinder deactivation for its 3-cylinder 1.0 liter SI engine. In [5] they compare fixed single cylinder deactivation with rolling cylinder deactivation and taking into account factors of complexity, controls efforts and cost effectiveness, they preferred single cylinder deactivation, although the rolling one gives better fuel cons. benefits [4], [5], [6]. All works agree that the dynamic cylinder deactivation have a few potential advantages over conventional cylinder deactivation: maintains uniform engine wall temperatures and load (stresses); is applicable to small number of cylinders; handles noise, vibration, and harshness well. This paper deals with the detailed thermodynamic simulation of the dynamic cylinder deactivation at light loads for a spark ignition 3-cylinder engine, focusing on the gas exchange process, and describes simulation issues that arise. To assess the fuel economy at partial load of an ICE, it has become common practice to show brake specific fuel consumption (BSFC) at an operation point of 2 bar of BMEP and 2000 min-1. Available BSFC values at this regime of contemporary production engines 132

134 (without cyl. deactivation) are presented in the Table. This point was therefore also used as a comparative point for each layout. In addition to it, load of 4 bar of BMEP at the same speed was added to capture possible trends. g/kw/h Ford 1.0 L EcoBoost 382 Hyundai 1.0 L 375 Fiat 0.8 MultiAir 375 VW 2.0 TSI Gen. 3B 325 VW 1.8 TSI Gen Daimler 1.6L Strat. SGDI TC Camtronic 290 Virtual simulation model 1.0 L (3-cylinder) 390 Ref. [7] [7] [7] [3] [3] [7] - Table 1: BSFC at 2000 rpm and BMEP = 2 bar 2 ENGINE MODEL For the simulation purposes a model of a gasoline turbocharged three cylinder engine with max. BMEP = 25 bar was used. This GT-Power model was derived from a calibrated model of a four-cylinder engine of a similar cylinder geometry. It is a detailed 1-D model with look-up maps (eng. speed vs. load) from experiments for: burn rate; ignition timing; turbine and compressor (scaled), AF ratio, intake and exhaust valve timing (newly optimized), charge air temperature, coolant and oil temperature. Wall temperature solver uses simplified FEM model. In-cylinder heat transfer coefficient is calculated by means of Woschni GT model and friction losses (FMEP) are calculated by the Chen-Flynn model (calibrated). These two empirical models appeared to be tricky in using them for the dynamic cylinder deactivation strategy and therefore were replaced or thoroughly tested, as described below. Since the simulation relates only to light load of a spark ignition engine, none of the boosting system was considered in this work and the turbocharger with the charge air cooler was removed from the model. 2.1 Friction model As noticeable in Figure (two 4-stroke cycles) by the dashed curve, during the dynamic cylinder deactivation, there occur more cylinder pressure peaks in one cycle, which makes the Chen-Flynn model for friction prediction inappropriate, as it was derived for a conventional 4-stroke cycle. Since the mechanical losses play an important role at light loads, a detailed predictive cranktrain friction model was utilized, as in detail described, tested and validated in [12]. This model consists of mechanical parts (masses) and physically-based connections representing local geometry, oil film hydrodynamics and asperity contacts. Input data was edited by available dimensions, masses and thermal properties. It is added to the thermodynamic model as a subassembly and the load is applied on the piston as a cylinder pressure signal that is transferred from the thermodynamic model. The effective torque is consequently sensed from the mechanical model and sent to all torque dependent parameters. Except the cranktrain friction losses itself, additional mechanical losses of oil and water pump and valvetrain drive were applied on the crankshaft in the form of negative torque load. Those losses were derived from reference values stated in [1]. As it was not possible to measure all geometry required for the sufficient calibration of the mechanical model, some parameters were estimated and optimized in order to 133

135 sufficiently match experimental FMEP values of the original 4-cylinder engine in the entire load vs. engine speed range. Consequently, the model was rebuilt to the 3cylinder one. The value of BSFC at 2000 rpm and 2 bar of BMEP was predicted as 390 g/kw/h (Table). 2.2 Heat Transfer Model and Wall Temperature Solution Another issue for the dynamic cylinder deactivation strategy might be calculation of the heat transfer coefficient by means of the Woschni correlation model, which is built and preferentially used in GT-Power. This model treats the heat transfer coefficients differently during the gas exchange period, i. e., where the heat transfer is increased by inflows velocities through the intake valves and by backflow through the exhaust valves and coefficients in the correlation are switched. However, during the nonworking cycle, where the gas exchange is cancelled, increase in the heat transfer within this period would be inaccurate. Therefore, two version of the Woschni correlation were tested on one period of the cycling pattern (i.e., working and nonworking cycle), namely Woschni Classic and Woschni GT. In Figure, there is their heat transfer coefficient profile together with cyl. pressure and valve lift profiles plotted for 2 cycles: first working cycle with gas exchange; second non-working compression of trapped air. It is clear from this picture that Woschni Classic(black colour) switches to higher heat transfer coefficient for the gas exchange period regardless of the valve lift opening (increased coefficient in the vicinity of the 1080 crank angle), while Woschni GT (grey colour) model respects actual valve opening and increases the calculation coefficients only on this occasion and therefore is suitable even for dynamic cylinder deactivation. Figure 1: Dynamic cylinder deactivation: cyl. pressure profile and heat transfer coefficients within two 4-stroke cycles For steady state simulation, GT-Power offers the capability to calculate wall temperatures fast, without considering thermal capacitance of the masses. This is, however, not suitable for dynamic cylinder deactivation and therefore the transient thermal solver has to be used, which requires long calculation, even better with a few iterations (more accurate initial conditions). For this reason, all layouts were calculated in this transient mode, even those without DCDA. 2.3 Residual Gas Control Throttling on the intake side of the engine and compromise valve timing is usually linked with higher internal exhaust gas recirculation at these regimes. Higher amount 134

136 of residuals positively affects pumping losses, as the charge is partially diluted. For fair comparison of all engine layouts, it is therefore necessary to assure same residual gas rate in the cylinder after closing of the intake valve. There was built an external EGR circuit with control in the model and valve timing was adjusted, so that the internal exhaust gas recirculation is always kept under the required EGR rate and controlled EGR valve offsets this rate by changing its diameter. The rate of EGR after intake valve closing was set to 10% (mass). 2.4 Results Integration and Convergence Check GT-Power is built to yield cycle averaged results at the end of each four stroke (or two stroke) cycle and also to check the convergence of solution in the way of cycle-to-cycle variation, i. e., the steady state solution is converged, when the difference between results from the previous and following cycles stays under required tolerance. However, with cycling cylinders (dynamically deactivating), the periodicity of the engine operation is not two crankshaft revolutions, but double the value, i. e., four revolutions (1440 crank angle degrees), as seen in Figure. Default cycle averaged results are therefore not useable, as it integrates results (and checks their convergence) every cycle, while we need results and their check over two four stroke cycles. For this reason, all desired average results (case RLT) must be manually integrated over 1440 deg of crank angle. It is done by means of the floating integral instantaneous signal is integrated in time by two integrating objects, one of which has delayed output about required periodicity and is subtracted from the first one. This value is then divided by simultaneously integrated time signal. Selected results, including flow and thermal quantities, are then checked for the convergence. It must be noted that in some cases the floating integral yielded confusing results and therefore a more complicated tool with external reset for the integral objects was developed. 2.5 Deactivation Control Tools For the cyclic control of the valves, injection and combustion, a control tool was built, which zeros (multiplies by 0) valve lift, combustion rate, and injection dose, when required. Another option for valve lift actuation was switching between two valve lift profiles in dependence on the simulation cycle. The control of the engine in general becomes complicated, when the engine cylinder works in a cycling pattern. Mainly those values, which are looked-up in dependence on the, e.g., load or residual rate values, simply values/results that are different for the working and non-working cycle and cannot be averaged over the engine periodicity. For instance, actual BMEP of the working cycle (different from the engine BMEP!) is therefore sensed in order to actuate burn rate profile or combustion timing, as well as residual gas rate is sensed only during the working cycle to control the residual gas amount. 2.6 Engine Conditions All layouts were simulated by means of different valve actuation with same conditions: BMEP = 2 and 4 bar, 2000 rpm, 10 % of internal EGR, lambda = 1, and basic valve timing (IVC, IVO, EVC, EVO) were kept at the same values with a few exceptions. This also means that the valve timing and valve opening duration were not optimized and there might be another potential in optimizing them. 135

137 3 VARIABLE VALVE ACTUATION METHOD Apart from charge dilution methods or LTC approaches, Miller/Atkinson cycle has been investigated for throttle loss decrease over last decades too. Such cycles use early or late intake valve closing (EIVC/LIVC), thus controlling the required amount of fresh charge in the cylinder by either cutting the flow during the intake stroke or pushing the excessive air out of the cylinder during the compression stroke. Both approaches are pictured in Figure in p-v diagram. Both approaches work under atmospheric (or slightly lower) pressure during the intake stroke, thus reducing the pumping losses. However, this benefit comes at the expense of deteriorated combustion due to the reduced compression stroke (charge temperature) and charge motion, which might subsequently even offset the gained benefit, as concluded in [9]. Increased geometric compression ratio can solve this problem and at WOT mode it can even improve brake thermal efficiency (prolonged expansion stroke), but the max. volumetric efficiency of the engine itself decreases and Otto valve timing is not possible due to knock reasons. This approach is however well handled by a few manufacturers [10]. Another era of the Miller cycle has come recently for turbocharged gasoline engines, when optimized combustion chamber and ports allowed EIVC and higher geometric CR and low volumetric efficiency is supplemented by boosting [3]. Efficiency loss due to unfavourable combustion phasing (knock control) at high load can be reduced by the prolonged expansion stroke, which helps to utilize later end of combustion. Figure 2: EIVC and LIVC (min. dynamic CR = 6) and Otto IVC at 2 bar of BMEP and 2000 rpm Above described model of a 3-cylinder gasoline engine was also tested for EIVC and LIVC approach at the designated operation points for comparison purposes. The intake valve closing was controlled in such a way that the dynamic compression ratio (IVC/TDC) equals 6, as it is the minimum value feasible in [3], [10]. It corresponds to the same value of IVC timing of 95 crank angle after BDC for LIVC and before BDC for EIVC. The throttle valve controlled the rest of the required fresh charge amount in the cylinder. In Figure, p-v diagram for EIVC, LIVC and standard IVC timing for maximum volumetric efficiency (Otto) is shown for 2 bar of BMEP and 2000 rpm together with vertical dashed lines marking intake valve closing. Due to the fact that the combustion was modelled in the form of imposed combustion rate, influence of 136

138 different conditions in the cylinder on combustion is not accounted for and the results would be then too optimistic. For this reason correlation from [13] was used to change the burn rate profile according to the temperature and pressure in the cylinder, which is lower for Miller/Atkinson, see Figure. This lengthened the combustion and worsened the BSFC. The CA50 point was maintained. Results of BSFC for the LIVC and EIVC strategies are about 18 and 24 g/kw.h lower compared to baseline layout at 2 bar of BMEP. Results of higher load point are in Figure. 4 CYLINDER DEACTIVATION (VDE) 4.1 Static Cylinder Deactivation (SCDA) In this case, only one cylinder is permanently deactivated, i. e., only 1/3 of the engine displacement, which predetermines this approach to be less advantageous in terms of fuel economy at lighter loads. As already mentioned, in [5] they decided for this kind of cylinder deactivation, although they had to cope with the crankshaft torsional oscillation, which was solved by using a dual mass flywheel with pendulum damper. Dynamic effects are not an issue in the simulation. The only question is, whether the deactivated cylinder (valves and injection are shut off) runs with trapped air, vacuum (exh. gas remain in comb. chamber) or exhaust gas. Since the deactivated cylinder might stay in the motored regime for a long time, it is better to trap fresh air in the cylinder, because too low pressure (vacuum) would lead to higher flow of blow-by gases with oil to the combustion chamber (which is not a simulated issue, of course). The thermodynamic model with the detailed friction model was edited only regarding the control of the deactivation of a cylinder. The model was first initialized running without deactivation and after convergence the deactivation of one cylinder was applied and this mode was run until steady state. Resulting values of BSFC at 2 and 4 bar of BMEP and 2000 rpm was about 38 and 19 g/kw.h lower than for the baseline layout respectively. Remarkable are the wall temperatures of the deactivated cylinder, which steady at a value that is about 30 K lower than other working cylinders. 4.2 Dynamic Cylinder Deactivation (DCDA) This kind of deactivation makes each cylinder work alternately one four-stroke cycle produces effective work, while the following four-stroke cycle is motored (partially consumes work) and this pattern repeats again. It is again possible to apply different strategies regarding the gas exchange process, when the cylinder during the deactivated cycle is filled with air, no charge (vacuum) or different amount of exhaust gases DCDA with Air and Vacuum charge For this deactivation strategy, the cylinder is filled either with certain amount of air or very little charge of exhaust gases during the deactivated work-non-effective cycle. When the whole gas exchange process occurs during the activated work-effective cycle, the cylinder is filled with fresh air (corresponding to the required load) and is transferred through the deactivated cycle, i. e., twice compressed, and then burned with fuel during the activated cycle. This strategy is shown by cylinder pressure pattern and valve lifts in the Figure by grey colour. Another option is to perform only the exhaust part of the gas exchange process during the deactivated cycle and to draw the fresh air during the deactivated cycle, as depicted in Figure by black colour. This means that only small amount of exhaust gases (corresponding to combustion chamber volume) is kept in the cylinder. As it can be 137

139 seen in the Figure, during all expansion phases of the crankshaft mechanism, the pressure in the cylinder drops almost to zero (under 0.1 bar), as well as charge temperature goes to freezing point. This might be problematic in real life due to high negative pressure drop between the oil sump and combustion chamber, but as it occurs in a cyclic pattern, it might be feasible. However, this issue is beyond the scope of this work. On the other hand, it is advantageous regarding fuel economy, because the thermodynamic work during compression strokes is lower than for the full air strategy and the friction losses are also lower due to lower pressure peaks. Figure 3: Cyl. pressure profile and valve lifts within 2 cycles for "Air transfer" and "Vacuum transfer" dynamic cyl. deactivation strategy (BMEP=4 bar, 2000 rpm) Resulting BSFC is about 43 g/kw.h and 56 g/kw.h lower than the baseline throttled engine for the Air and Vacuum deactivation strategy respectively at the point for 2 bar of BMEP and 2000 rpm. Comprehensive comparison of each layout is shown in Figure, also for the higher load of 4 bar of BMEP, where the trend is similar. It must be noted that the Air strategy did not meet the required level of the load of 4 bar of BMEP (only 3.7 bar), so the results of BSFC are not sufficiently comparable at this point DCDA with Exhaust Charge Another approach in the dynamic deactivation is to perform the whole gas exchange process as late as during the deactivated cycle. This means that most of the negative work occurs already during the activated cycle, which is noticeable in Figure by grey colour representing the cylinder pressure profile. Compression of hot exhaust gases leads to peak pressures and temperatures that are even higher than the combustion pressure and this automatically concludes that this strategy is not suitable at all. The exhaust gas can be, however, partially released just after the expansion stroke (Figure, black colour) and the pressure peaks are lowered to acceptable pressures. The charge temperature is still high, though, and the heat transfer during the cranking is higher, so there is more negative thermodynamic work than in the case of Air charge strategy and the resulting consumption is worse. 138

140 Figure 4: Cyl. pressure profile and valve lifts within 2 cycles for "Exhaust gas transfer" and "Partial exhaust gas transfer" dynamic cyl. deactivation strategy (req. BMEP=4 bar, 2000 rpm) In comparison with baseline throttled layout, the full exhaust gas charge exceeds BSFC about tens of g/kw.h (did not even fulfil req. BMEP = 4 bar), regarding the partial exhaust gas charge, BSFC benefit is about 24 g/kw.h at 2 bar of BMEP, but still less beneficial compared with Air charge or Vacuum strategies. 5 DISCUSSION OF RESULTS Figure combines all results of every single above mentioned layout for two load points, 4 and 2 bar of BMEP at 2000 rpm. Figure presents average heat transfer rate to cylinder walls to explain thermodynamic losses of the simulated layouts, since it is difficult to find a common criterion that would explain given differences between the DCDA strategies, as e. g. PMEP value is applicable only for those approaches, where the gas exchange process occurs in one cycles. These values are presented in Figure. Except the DCDA strategy with exhaust gas charge, all other layouts show better fuel consumption at both load points. At higher load, approaches without dynamic cylinder deactivation (LIVC, EIVC, Static CDA) reach similar or even better values of BSFC than DCDA. It is caused by the fact that at this load point, the throttle losses are not high enough to overcome losses that are linked with compression of air or exhaust gas for the DCDA (causing thermodynamic loss and higher friction loss). This fact is noticeable on the comparison of static CDA and dynamic CDA with air charge, where the static one shows better fuel consumption. According to Figure, on one hand, the static CDA has higher pumping losses, but on the other hand, there is less amount of air in the deactivated cylinder and the thermodynamic and friction losses are lower (see Figure for heat transfer loss). LIVC and EIVC strategies also perform well at the higher load, in addition to lower pumping losses (PMEP in Figure, Shelby def. [11]) they also profit from prolonged expansion stroke. However, results for these strategies should be taken with some uncertainty resulting from the hardly predictable combustion rate and efficiency. At lighter load (BMEP = 2 bar), pumping losses dominate the losses linked with dynamic CDA (Figure) and both air charge and vacuum DCDA approaches show best BSFC values. The approach with partial exhaust gas charge still performs worse compared to static CDA, which still has the lowest heat rejection to cylinder walls. 139

141 Figure 5: BSFC values for each engine layout for 2000 rpm and two load point: 2 and 4 bar of BMEP Figure 6: Heat transfer rate values averaged for 1 cylinder (and over two cycles for DCDA) 140

142 Figure 7: PMEP values for selected layouts (only those with complete gas exchange process in one cycle). * PMEP Shelby definition [11] 6 CONCLUSION Dynamic cylinder deactivation is a promising way for decrease of throttle losses of particularly 3-cylinder engines. At 4 bar of BMEP (2000 rpm), it can reach benefit from 3% up to 9 % of BSFC, depending on the cylinder charge of the deactivated cycle. At 2 bar of BMEP, the range of benefit is between 11% and 14%. Although the dynamic cylinder deactivation with exhaust gas charge does not look promising, it might be an interesting way for faster warm-up of engine coolant (and the cabin) with still beneficial fuel efficiency. This comes at the expense of temperature of exhaust gas and warmup of aftertreatment systems, though. Possible further benefit can be find in thorough valve profile and timing optimization, also for better fresh charge control for the DCDA strategy. These results were carried out at steady state operation and it is necessary to evaluate this benefit in driving cycles in the next steps. Further, it would be interesting to investigate the cooperation of these deactivation strategies with different boosting systems, both at steady and transient modes. Regarding simulation, a faster friction model respecting dynamic cylinder deactivation shall be built to allow less time consuming computation. 7 REFERENCES [1] HEYWOOD, J., B.: Internal Combustion Engines Fundamentals, McGrawHill,1988. ISBN X [2] Wilcutts, M., Switkes, J., Shost, M. and Tripathi, A., "Design and Benefits of Dynamic Skip Fire Strategies for Cylinder Deactivated Engines," SAE Int. J. Engines 6(1):2013 [3] Budack, R., Wurms, R., Günther, M., Heiduk, T. The New Audi 2.0-l I4 TFSI Engine, MTZ Online, Vol. 77, 05/2016, p

143 [4] Isenstadt, A., German, J., Dorobantu, M. Naturally aspirated gasoline engines and cylinder deactivation, International Council On Clean Transport, Working paper , 2016 [5] Küpper, K., Linsel, J., Pingen, B., Weber, C. Cylinder Deactivation for Threecylinder Engines, MTZ Online, Vol. 77, 12/2016, p [6] Birch, S. Ford tests cylinder deactivation on its 1.0-L EcoBoost triple, Automotive engineering, 2015, [7] Friedfeldt, R., Zenner, T., Ernst, R., Fraser, A., Three-cylinder Gasoline Engine with Direct Injection, MTZ Online, Vol. 73, 05/2012, p [8] Sellnau, M., Foster, M., Hoyer, K., Moore, W. et al., Development of a Gasoline Direct Injection Compression Ignition (GDCI) Engine, SAE Int. J. Engines 7(2):2014 [9] Ueda, N., Sakai, H., Iso, N., Sasaki, J. A Naturally Aspirated Miller Cycle Gasoline Engine - Its Capability of Emission, Power and Fuel Economy, SAE , 1996 [10] Ellies, B., Schenk, C., and Dekraker, P., "Benchmarking and Hardware-in-theLoop Operation of a 2014 MAZDA SkyActiv 2.0L 13:1 Compression Ratio Engine," SAE Technical Paper , 2016 [11] Shelby, M., Stein, R., Warren, C. A New Analysis Method for Accurate Accounting of IC Engine Pumping Work and Indicated Work, SAE Technical Paper , 2004 [12] Emrich, M., Takáts, M. Detail Engine Friction Estimation Using ExperimentallySimullation Approach. KOKA 2016 Conference Proceedings, Brno, 2016 [13] Skarohlid, M. Modelování vlivu složení paliva na vlastnosti plynového zážehového motoru, Disertační práce, ČVUT v Praze, ACKNOWLEDGEMENT This work has been realized using support of: Technological Agency, Czech Republic, programme Centres of Competence, project # TE Josef Božek Competence Centre for Automotive Industry. The Ministry of Education, Youth and Sports program NPU I (LO), project # LO1311 Development of Vehicle Centre of Sustainable Mobility. Eaton European Innovation Centre within the Student Research Services Agreement. This support is gratefully acknowledged. 142

144 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAKUNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICALUNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES THE VARIABLE VALVE TIMING MEASUREMENT Tomáš Zvolský1 Abstract The paper deals with a procedure how to measure variable valve timing of a specific combustion piston engine. The angular deflections of the intake camshaft and exhaust camshaft are graphically shown. Part of this paper is also the measurement of engine torque when the variable valve timing is disabled. The torques of the combustion engine with the enabled and disabled variable valve timing are compared. The measurement results will be used for future research camless actuation valve system of combustion piston engine. 1. ÚVOD Konstruktéři se snaží docílit co nejlepších provozních vlastností motorů, ať už se jedná o vysoký výkon, vysoký krouticí moment v širokém rozsahu otáček, nižší spotřebu paliva, nebo nižší emise výfukových plynů. Jeden ze způsobů jak toho dosáhnout je zavedení variabilního časování ventilů. U motorů s obvyklým ovládáním ventilů je plnění válců optimální pouze při určitých otáčkách. Tato situace nevyhovuje především z hlediska dosažení vyšších výkonů. Pro tento případ se totiž vyžadují velké průtokové plochy ventilů s velkým překrytím ventilů. Ovšem u takto odladěného motoru pak chybí potřebná síla při nízkých otáčkách, takže motor ztrácí náležitou pružnost. K docílení dobrého průběhu točivého momentu potřebuje motor zcela jiné časování ventilů. Sladit protichůdné požadavky (vysoký výkon a vhodný průběh točivého momentu) v širokém rozsahu provozních otáček znamená zavést variabilní časování ventilů měnící se podle aktuálních podmínek chodu motoru. Nejčastěji používaným způsobem je natáčení vačkového hřídele vůči klikovému hřídeli. Natáčením vačkového hřídele se dá posouvat fáze sání, nebo výfuku. Zdvih a doba otevření ventilů zůstává neměnná. Ing. Tomáš Zvolský, tomas.zvolsky@tul.cz 1 Technical University of 143 Liberec, Studentská 2, Liberec,

145 2. POPIS MĚŘENÍ 2.1 Měřený motor Měřený motor je zážehový přeplňovaný řadový tříválec s přímým vstřikováním paliva 1.0 TSI. Parametry motoru jsou zobrazeny v tabulce 1. Zdvihový objem Počet válců Počet ventilů na válec Maximální výkon Maximální točivý moment 999 cm kw 200 Nm Tabulka 1: Parametry motoru Ventilový rozvod je typu DOHC. Obě vačkové hřídele se můžou úhlově natáčet vůči klikovému hřídeli. V horním víku motoru jsou společně s vačkovými hřídeli umístěny elektromagnetické ventily přestavování sacího vačkového hřídele N205, výfukového vačkového hřídele N318 a také Hallovy senzory polohy sacího a výfukového vačkového hřídele G40 a G300, je vidět na obr. 1. Obr. 1: Umístění ventilů přestavování a Hallových snímačů v horním víku motoru 2.2 Příprava motoru pro měření Pro změření fázových posuvů vačkových hřídelů bylo potřeba získat signál ze snímače klikového hřídele a obou snímačů vačkových hřídelů. Všechny 3 snímače jsou Hallovy sondy, které jsou napájeny z řídící jednotky motoru napětím 5 V a na výstupu dávají logický signál. Pro získání signálu ze snímačů byla vždy mezi snímač a kabel vložena vyrobená redukce s odbočkou na BNC konektor, je vidět na obr

146 Redukce pro získání signálu ze snímačů sacího a výfukového vačkového hřídele Signálu ze snímače klikového hřídele Obr. 2: Redukce pro získání signálu z Hallových snímačů na motoru 1.0 TSI 2.3 Měřicí zařízení Pro měření úhlů natáčení vačkových hřídelů byl použit přístroj AVL 619 INDIMETER. Je možné měřit až 8 kanálů v závislosti na úhlu natočení klikového hřídele. Ozubený kotouč na klikovém hřídeli motoru má 60 zubů mínus 2 vynechané zuby a signál ze snímače klikového hřídele nelze přímo připojit k Indimetru. Pro zpracování tohoto signálu je potřeba použít přídavný přístroj AVL CRANK ANGLE CALCULATOR, který jednak dopočítá signál pro vynechané zuby a zároveň podle vynechaných zubů určí horní úvrať motoru. Měřicí přístroje AVL jsou zobrazeny na obr. 3. Obr. 3: AVL 619 INDIMETER a AVL CRANK ANGLE CALCULATOR Na obr. 4 je vidět záznam měření z Indimetru během 720 klikového hřídele. Modrá křivka je signál ze snímače sacího vačkového hřídele, červená křivka ze snímače výfukového vačkového hřídele a černá ze snímače klikového hřídele. Je vidět, že obě vačkové hřídele mají ozubený kotouč se 4 zuby (2 široké a 2 úzké) a vynechané zuby na klikovém hřídeli se nacházejí cca 90 před horní úvratí 1. válce. 145

147 Obr. 4: Záznam z Indimetru během 720 klikového hřídele Fázové posuny vačkových hřídelů vůči klikovému hřídeli lze také zjistit pomocí standardního osciloskopu. Osciloskop však zaznamenává měřené hodnoty v závislosti na čase, což je nevýhoda oproti Indimetru a při nerovnoměrném chodu motoru není měření tak přesné. Navíc nelze ze záznamu ihned odečíst fázový posun, úhel posunu se musí vypočítat z časových záznamů signálů. Pro vyzkoušení tohoto způsobu měření byl použit čtyřkanálový osciloskop typ Rigol DS1054Z. Na obr. 5 je vidět záznam měření z osciloskopu. Obr. 5: Záznam z osciloskopu Rigol DS1054Z 146

148 Tmavě modrá křivka je signál ze snímače sacího vačkového hřídele, červená křivka ze snímače výfukového vačkového hřídele a světle modrá ze snímače klikového hřídele. Žlutá křivka je trig. signál z AVL CRANK ANGLE CALCULATORu, na který je nastaveno spouštění osciloskopu, avšak pro samotné zjištění natočení vačkových hřídelů není tento signál nezbytně nutný. 3. MĚŘENÍ A VYHODNOCENÍ VÝSLEDKŮ 3.1 Měření natáčení vačkových hřídelů Měření úhlu natáčení vačkových hřídelů probíhalo od volnoběžných otáček motoru do 6500 min-1 s krokem 500 min-1. Poloha sešlápnutí pedálu akcelerace se měnila od 0 do 100 % s krokem 4 %. Celkem tedy proběhlo 338 měření. Výsledky jsou znázorněny ve dvou vrstevnicových grafech, jeden pro sací vačkový hřídel obr. 6, druhý pro výfukový vačkový hřídel obr. 7. Obr. 6: Natočení sacího vačkového hřídele v závislosti na otáčkách a zatížení motoru Při volnoběžných otáčkách motoru (bez zatížení) se obě vačkové hřídele nacházejí ve výchozí - nulové poloze. Při změně otáček a zatížení motoru se potom vačkové hřídele natáčejí. Rozsah přestavení pro sací vačkový hřídel je maximálně 50 na klikovém hřídeli, což odpovídá 25 na vačkovém hřídeli a to od základního nastavení směrem na dříve. Měřené hodnoty jsou tedy záporné, při nejvyšším natočení -50 klikového hřídele se sací ventil otevírá i zavírá nejdříve. Rozsah přestavení pro výfukový vačkový hřídel je maximálně 40 na klikovém hřídeli, což odpovídá 20 na vačkovém hřídeli, avšak na rozdíl od sací vačky směrem na později. Měřené hodnoty jsou tedy kladné, při nejvyšším natočení 40 klikového hřídele se výfukový ventil otevírá i zavírá nejpozději. 147

149 Obr. 7: Natočení výfukového vačkového hřídele v závislosti na otáčkách a zatížení motoru 3.2 Měření momentu motoru při vypnutém natáčení vačkových hřídelů Měření momentu motoru probíhalo jednak standardně se zapnutým systémem natáčení vačkových hřídelů, tak následně s deaktivovaným systémem pro natáčení vačkových hřídelů, je vidět na obr. 8 a obr. 9. Obr. 8: Moment motoru při zapnutém systému natáčení vačkových hřídelů 148

150 Obr. 9: Moment motoru při vypnutém systému natáčení vačkových hřídelů Deaktivace systému pro natáčení vačkových hřídelů byla provedena odpojením konektorů elektromagnetických ventilů přestavování sacího a výfukového vačkového hřídele (ventily N205 a N318), které jsou umístěny v horním víku motoru. Obě vačkové hřídele se tedy nemohly přestavovat a po celou dobu měření zůstaly v základní nulové poloze. V diagnostice byla zapsána závada odpojení ventilů, avšak motor i tak pracoval bez potíží. Pro porovnání je na obr. 10 znázorněn rozdíl momentu motoru při zapnutém a vypnutém systému pro natáčení vačkových hřídelů. Obr. 10: Rozdíl momentu motoru při zapnutém a vypnutém natáčení vačkových hřídelů 149

151 Z grafu na obr. 10 je vidět, že největší rozdíl momentu se projeví v rozmezí otáček motoru 1500 až 2000 min-1, naopak ve vysokých otáčkách nemá deaktivace natáčení vačkových hřídelů na moment motoru prakticky vliv. K měření momentu motoru je třeba ještě doplnit, že z organizačních důvodů probíhalo měření na zkušebně Powertrain, kde probíhá měření motoru i s převodovkou. Motor tedy nebyl přímo připojen k dynamometru, měřil se až výsledný moment na výstupu z převodovky MQ200. Moment motoru byl následně vypočítán z převodového čísla, avšak výsledek není zcela přesný, protože ve výpočtu není zahrnuta účinnost převodovky, která není přesně známa. Tato chyba však není pro stanovení vlivu natáčení vačkových hřídelů zcela zásadní. 4. ZÁVĚR U měřeného motoru se obě vačkové hřídele plynule přestavují v závislosti na otáčkách a zatížení motoru. Rozsah přestavení pro sací vačkový hřídel je maximálně 50 na klikovém hřídeli a rozsah přestavení pro výfukový vačkový hřídel je maximálně 40 na klikovém hřídeli. Sací a výfukový vačkový hřídel se od základního nastavení natáčejí různými směry, sací směrem na dříve a výfukový směrem na později. Naměřené hodnoty z Indimetru i osciloskopu jsou srovnatelné. Při deaktivaci systému natáčení vačkových hřídelů se největší pokles momentu motoru projeví v oblasti otáček min-1. Pokles však může být způsoben ještě dalšími zásahy řídící jednotky motoru. Naopak ve vyšších otáčkách motoru nemá deaktivace natáčení vačkových hřídelů na moment prakticky vliv. LITERATURA [1] [2] [3] [4] [5] [6] [7] HEISLER, H. Advanced Engine Technology. SAE Technology, 1995, s ISBN VLK, F. Vozidlové spalovací motory. Nakladatelství VLK, 1. vydání, Brno, s ISBN ZDENĚK, J., ŽDÁNSKÝ, B. Automobily 3, Motory. Avid, 1. vydání, Brno, 2000, s [online].[cit ]. Dostupné z: BEROUN, S., PÁV, K. Vybrané statě z vozidlových spalovacích motorů. Liberec, 2013, s ISBN SCHWARZ J. Automobily Škoda Octavia II, Grada, 2010, s ISBN Spark-ignition three-cylinder engine 1.0l TSI 85 kw series EA211, Dílenská učební pomůcka č. 111, Škoda Auto a. s., 9/2016 PODĚKOVÁNÍ Tato publikace byla napsána na Technické univerzitě v Liberci jako součást projektu za podpory Grantu specifického univerzitního výzkumu poskytovaného Ministerstvem školství, mládeže a tělovýchovy České republiky v roce

152 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES AREA OF UTILIZATION OF ALTERNATIVE COMPRESSED AIR DRIVE PRACTICAL APPLICATION Ľubomír Miklánek1, Marián Čučo2, Ondřej Gotfrýd3 Abstract The Compressed Air (CA) drive is classified as one of the alternative drives. Therefore it was considered in framework of the research for sustainable mobility. This theme is attractive in research area especially for young people and also for our students. Some of them are missing a practical application of this kind of alternative drive in our Research Centre. This article deals with description of a practical application of CA drive and shows areas where this concept can be applied with benefits closed industrial objects with plenty of technological compressed air. It is generally known that alternative CA drive is in area of transport not as effective as other alternative drives (e.g. Compressed Natural Gas drive). 1. INTRODUCTION Compressed Air (CA) drive belongs to alternative drives which are classified as renewable fuels. Vehicles equipped with this drive belong to a group of Local ZeroEmission Vehicles (LZEV), same as Electric vehicles (EV). It is also well known that CA drive is not effective solution in a transport area, especially not in a long distance transport. Nevertheless, CA drive is still very attractive theme for general and even professional public. It is similar like e.g. a water fueled car. A practical application of CA drive in our Research Centre is missing by some students. Another situation is at the Brno University of Technology (BUT) where students build racing vehicles with CA drive, see Figure 1, [1]. Ľubomír Miklánek, CTU in Prague, Faculty of Mechanical Engineering, Vehicle Centre of Sustainable Mobility, Přílepská 1920, Roztoky, Lubomir.Miklanek@fs.cvut.cz 2 Marián Čučo, Žilina, kralskusok@post.sk 3 Ondřej Gotfrýd, CTU in Prague, Faculty of Mechanical Engineering, Vehicle Centre of Sustainable Mobility, Přílepská 1920, Roztoky, ondrej.gotfryd@fs.cvut.cz 1 151

153 However, these vehicles are suitable rather for free-time activities and not for daily use. Therefore, this article is focused on using the CA drive in a practical everyday life. Figure 1: Racing vehicles with CA drive designed by students at Brno University of Technology, [1]. Task to be solved: Presentation of basic CA drive information. To analyze areas of using CA drives in everyday practical life. To determine the area of practical use of CA drive under specific conditions. Chosen approach:. Comparison of CA drive with other alternative drives. Presentation of the CA drive practical application in an industrial object under condition of plenty of technological compressed air. This is not any detailed Well to Wheel (WTT) nor Tank to Wheels (TTW) analysis. 2. THERMODYNAMIC POINT OF VIEW Thermodynamic point of view is primarily focused on an expansion of the compressed air in an air engine of a vehicle. Issues of air compression nor heat dissipation during air compression/expansion will not be solved. The expansion of the CA in an air engine can be replaced by thermodynamic change for simplified conditions, according to [2] and [3]: Isothermal expansion, i.e. dt = 0 Energy in compressed air can be calculated as: dq dwt dw du pdv 0 pdv vdp (1) As shown by the relationship (1), the internal energy of compressed air is zero: du cv dt 0 (2) Furthermore, this thermodynamic change never uses heat supply (e.g. isochoric heat supply), as it is in case of internal combustion engines (I.C.E.): 152

154 dqin cv dt 0 (3) Where the temperature of the working fluid could increase (dt) e.g. more than 2000K. The expansion of hot working fluid is not considered, as well. The meaning of symbols: cv Specific heat capacity at constant volume, [J.Kg-1.K-1], T Temperature, [K], q Specific energy, [J.kg-1], wt Specific technical work, [J.kg-1], w Specific volume work, [J.kg-1], u Specific internal energy, [J.kg-1], p Pressure, [Pa], v Specific volume, [m3.kg-1], As relation (1) shows, energy of compressed air is obtained only from the difference of air pressure at the beginning and the end of the expansion (usually in a cylinder of an air engine. This results in a major energy disadvantage in comparison of fuel combustion in I.C.E. with expansion of hot gases. Therefore it is necessary to bring much more working fluid (CA) into air engine to obtain the same power output in comparison of fuel, which is burnt in I.C.E. (e.g. Natural gas (NG)). 3. OVERVIEW OF CA DRIVES IN TRANSPORT AND INDUSTRY 3.1 CA drives in transport Recently, several manufacturers have dealt with the issue of CA drive (or other nonflammable gas). Presented solutions are either with CA drive as the only one drive or CA drive as an additional drive for hybrid vehicles. The company Peugeot is one of the carmakers developing a hybrid vehicle with CA drive. A vehicle prototype called Hybrid Air, [4] and [5], uses a classical I.C.E. gasoline fueled and moreover auxiliary CA drive. The auxiliary CA drive is in operation only for a short time, especially in transients start and acceleration of the car. It is also able to recover part of the vehicle's kinetic energy during braking of the vehicle. However, the auxiliary CA drive system is inactive during steady drive, e.g. during drive on the highway. This system did not get into serial production, despite various promises. On the other hand, the vehicle producer MDI (Moteur Development International) has presented vehicles with only CA drive, [6], equipped with a CAT engine (Compressed Air Technologies), [7]. Based on [6], the vehicle with CA drive is called AirPod. The body of the vehicle has the shape of a bubble and it is made of glass fibers, see Figure 2. The AirPod is not controlled by a steering wheel, but using a lever. 153

155 Figure 2: Prototype of vehicle with CA drive, named AirPod, of MDI Company, [6]. Specific technical parameters of the AirPod vehicle, stated by manufacturer, are shown in Table 1. Curb weight [kg] 220 Maximal speed [km/h] Range in city-traffic [km] Air tank volume [l] 175 Pressure of air in the tank [bar] 350 Time of tank filling [min] <2 Table 1: Specific technical parameters of vehicle AirPod with AC drive, [6] The range listed in Table 1 is the most interesting parameter: min. 180 km. The equation (1) can be used for at least theoretical verification of this value. Considering the behaviour of air as an ideal gas, the energy in compressed air can be calculated using: 2 V dv m r T ln 2 V1 1 V 2 Q W Wt p dv m r T 1 (J) (4) Where: m- Weight of working fluid, [kg], V1 - r - Specific gas constant for dry air, [J kg-1 K-1], V2 - Volume of working fluid at the begin of expansion, [m3], Volume of working fluid at the end of expansion, [m3]. Energy necessary for 1 km of vehicle drive was calculated based on the CNG consumption of the vehicle Škoda Citigo (3,2 kg/100km). Calculated theoretical value of AirPod vehicle driving range is: 20,2 km. 154

156 However, the journalist taking part on the introduction of this vehicle reporting the AirPod range about only 10km, see [7]. Obviously, the range of only 10km is nonprofitable considering the investment into the compression station (more than 350bar) and the vehicle with CA drive. 3.2 CA drives in industry At present, compressed air is used in most manufacturing plants as one of the most important energy carriers for production processes. A company produces usually compressed air itself and it can be named as technological compressed air. Compressed air is typically used to drive various production machines, but also hand tools. Advantages of pneumatic tools compared to the electric are e.g.: smaller dimensions, lower weight, lower price,... A compressed air drive can also be used with advantage in a hazardous or explosive environment - especially in mining. On the other hand compressed air is minimally or not at all used to drive any transportation means in industry. History knows cases of pneumatically-driven trucks used, for example, during the construction of the Paris subway in In this case experience of usage of airpowered tram in the Paris were applied, see [8]. The tram had its own air tank and air filling (with a pressure of up to 50bar) was carried out at the filling stations along the track. 4. OPTIONS FOR AREA OF UTILIZATION OF CA DRIVES FOR TRANSPORT The above mentioned shows that the compressed air transportation means can be effective only if the CA drive limitation is accepted: Low range i.e. the movement of the vehicle should be only in a closed area, both civil and industrial, High cost of compressed air i.e. the implementation of CA drive should be only in buildings where enough technological compressed air is and where the air consumption of transportation means would be only a small part of the total volume in the building. The carriage with the CA drive for the transport of persons or material in a closed area with enough technological compressed air would be considered. The filling of the tank in the vehicle would take place in filling stations at the beginning and at the end of the route, event. along the route. 5. VEHICLE WITH CA DRIVE IN INDUSTRIAL AREA PRACTICAL APPLICATION A prototype of the carriage with the CA drive has been built by a second author to gain experience with this type of drive. The prototype of LZEV is named: Felda-Air-9b. This is a low-cost amateur modification of a serial vehicle (Škoda Felicia) and it is designed for the transport of 1 person. The main aim was to reduce the prototype's weight. The modification consists in vehicle shortage and removal of all components 155

157 of a serial vehicle that are not needed for the operation of the vehicle with an AC drive (such as original I.C.E., seats, exhaust and fuel system, fuel tank and other), see Figure 3. Figure 3: On the left side: prototype of carriage with CA drive (LZEV) for 9-bar application in industry object. On the right side: location of the CA tank inside the prototype with dashed line. A schema of pneumatic drive of the prototype Felda-Air-9b is shown in Figure 4. Figure 4: A schema of pneumatic drive of the prototype Felda-Air-9b Within the low-cost solution, two Rotary Vane Motors (RVM), from original pneumatic impact wrenches, were used as propulsion units. These devices are used in tire repair shops for tightening wheel bolts. Two RVMs are serially arranged with a reduction gear in order to obtain increase of resulting torque. RVMs are located in front of the original 5-speed manual transmission of the vehicle. Moreover, the air supply to RVM 2 can be controlled separately, as it is visible in Figure 4. This can save air during vehicle movement at constant speed, when only one RVM is needed for driving. View of the drive units arrangement (RVMs) in the prototype Felda-Air-9b is visible in Figure

158 RVM 1 Gear of the vehicle RVM 2 Rotor of RVM Figure 5: On the left side: view on drive units (Rotary Vane Motors- RVM) placed in front of the original gearbox of prototype Felda-Air-9b. On the right side: View on rotor of RVM The air tank was filled by a hosepipe from the technological compressed air distribution in an industrial object (plug-in). Specific technical parameters of vehicle prototype Felda-Air-9b, stated by designer, are presented in Table 2. Curb weight [kg] app Maximal speed [km/h] app. 10 Full tank range [m] Air tank volume [l] 150 Maximal pressure of air in tank [bar] 9 Time of tank filling [min] <2 Table 2: Specific technical parameters of prototype Felda-Air-9b It should be noted, that above presented CA drive of prototype was not optimized for full tank range, nor for maximum efficiency. Authors believe that it could be possible by optimization to achieve both higher speed of prototype and larger full tank range. Experience gained by practical application of prototype: carriage starts with CA drive, practical full tank range, drive at constant speed, air tank filling and others. 6. COMPARISON OF CA DRIVE WITH OTHER DRIVES Comparison of specific energy of above mentioned CA drives with other alternative drives (electric and CNG drives) and also with conventional drives (Gasoline and Diesel fuelled) is presented in Figure 6. Specific energies of various types of batteries are applied based on work [9]. 157

159 Figure 6: Comparison of CA drive specific energy with other drives, (MDI=CA drive) It follows from the above that only high-pressure MDI CA drive (350bar) is approaching the specific energy of Li-Ion battery. On the other hand, application of CA drive 9bar has larger specific energy compared to lead-acid battery (Pb). However, specific energy of systems with combustion of fuel (e.g. CNG or fossil fuel) is two orders of magnitude higher, see Figure 6. Comparison of tank sizes (its water volume) of different systems for 100 km of range is presented in Figure 7. Figure 7: Comparison of tank sizes (its water volume) of different systems for 100 km of range, (MDI=CA drive) As shown in Figure 7, all applications of CA drive have very large air tank compared to CNG and also conventional drives. Especially application of CA drive 9bar has very large air tank in comparison to MDI 350bar or 200bar. However, the price of the CA drive 9bar tank is lower compared to the price of the air tank of MDI CA drive. 158

160 Comparison of the energy amount of CA drive systems is presented in Table 3. Air tank volume, [l] Usable energy of CA, [MJ] Corresponding weight of Li-Ion battery, [kg] Correspondin g amount of Gasoline, [l] Air-9bar MDI-350bar Table 3: Comparison of the energy amount of mentioned CA drives with energy of both the Li-Ion battery and Gasoline fuel From above mentioned drives only two drives are so-called Local Zero-Emission: the CA drive and the Electric drive. Moreover, without a so-called carbon footprint. It is clear from the above that it makes sense to use the CA drive only in these applications: 1/ for a short range, 2/ where enough technological compressed air is. Comparison of specific advantages/disadvantages of using of the CA drive versus the Electric drive for both of applications above is presented in Table 4. Engine efficiency Filling/ charging time Number of recharge cycles Dimensions of energy storage Air tank/ battery price Compressed Air Drive (-) (+) (+) (-) (-) 2x (+) Electric drive (+) (-) (-) (+) (-) 2x (+) Number of advantages Table 4: Comparison of specific advantages (+) and disadvantages (-) of using of the CA drive compared to the Electric drive, (for a short range). As follows from Table 4, both LZEV concepts of propulsion are roughly the same (for the short range) without considering investments in the construction of charging station or compressor station. The number of recharge cycles of Li-Ion batteries is considered app , see [9]. However, this number should be much higher in case of air tank. It is clear, that for a specific application of one of the mentioned drives, it is not possible to use simple analysis in Table 4. More technical and economic considerations have to be taken into account. 7. CONCLUSION This paper deals with an analysis of the suitability of the CA drive for transportation means in daily usage As next drive range of vehicle with CA drive of one producer has been verified by calculations. The calculations have shown a significantly lower range than the manufacturer has stated. This is because of the low energy density in the compressed air - only the entropy of the working substance is used, namely the pressure difference at the beginning and 159

161 end of the expansion. Also, the efficiency of air engine does not reach the efficiency of electric motors. The aforementioned dependences leads to conclusions that the CA drive can be used only on a small distances (because of a large air tank) and only in the case of the sufficient amount of technological compressed air. For example, it can be manipulation vehicles for the transport of persons or material inside an industrial building. For low-pressure CA drives (e.g. 9 bar), the meaningful range is few hundreds of meters. High-pressure CA drive (200 bar or more) it is a few kilometers. REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] VACEK V, NOŽIČKA J. Příručka termodynamiky s příklady, Vydavatelství ČVUT v Praze, Praha, pp MACEK J. Spalovací turbiny, turbodmychadla a ventilátory, Vydavatelství ČVUT v Praze, Praha, pp ISBN HYBRID.CZ EKOLOGICKÉBYDLENÍ.EU IDNES.CZ _ak_aktual_fdv, STOLETI.CZ WIKIPEDIA ALTERNATIVNÍ POHONY V DOPRAVĚ. 66/AltPohVDopr.ppt, ACKNOWLEDGEMENT This research has been realized using the support of The Ministry of Education, Youth and Sports program NPU I (LO), project # LO1311 Development of Vehicle Centre of Sustainable Mobility. This support is gratefully acknowledged. This research has been realized using the support of Technological Agency, Czech Republic, programme Centres of Competence, project # TE Josef Božek Competence Centre for Automotive Industry. This support is gratefully acknowledged. This research has been realized using the support of EU Regional Development Fund in OP R&D for Innovations (OP VaVpI) and Ministry for Education, Czech Republic, project # CZ.1.05/2.1.00/ Acquisition of Technology for Vehicle Center of Sustainable Mobility. This support is gratefully acknowledged. Special thanks to second author for providing backgrounds from practical application of CA drive (9 bar) in prototype vehicle. 160

162 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES INFLUENCING OF ENGINE BLOCK LOAD BY MEANS OF CRANKSHAFT DESIGN Lubomír Drápal1 Abstract The engine block load is significantly affected by crank train inertia forces and moments mainly at higher engine speed. This load can be reduced by means of crankshaft counterweights and their appropriate design. The influence of counterweights design upon the engine block load is examined on an in-line four-cylinder naturally aspirated engine in consideration of internal moments of crank train rotating parts. For the analysis are used state-of-the-art methods such as CAD, Finite Elements Method and Multi-body System tools which incorporate flexible bodies, and a computation model of hydrodynamic bearing dynamics. INTRODUCTION The engine block is loaded, mainly, at crankshaft main bearings and this load can be described as a force effect obtained by numeric integration of a pressure field at the bearing hydrodynamic film. This type of loading is investigated in the case of an aluminium deep-skirt engine block of modern 1.6-litre naturally aspirated spark-ignition in-line four-cylinder engine. The engine is considered in two variants: the standard one and the variant with reduced friction losses. The reducing of crank train friction losses is achieved by a reduced number of crankshaft main bearings from 5 to 3. The new 3-main-bearing (3mb) crankshaft is based on a 5-main-bearing (5mb) version of the standard engine. The missing main pins are replaced by sheet-metal webs due to mass and inertia moments reduction. A connection between the web and the crankshaft is created by laser welding for low thermal load of the weld. Current version of the 3mb crankshaft results from a previous design, computational, and technology studies. These studies show that potential for saving of power losses of crankshaft main bearing reaches around 33 % in comparison to the standard 5mb crankshaft [1]. However, not only friction losses are affected by the new crankshaft design, but also changes in system dynamic response must be taken into account. In order to Lubomír Drápal, Institute of Automotive Engineering, Brno University of Technology, Technická 2896/2, Brno, drapal@fme.vutbr.cz 1 161

163 investigate a dynamic response of excited crank train, the state-of-the-art computational approaches are used. 1 SIMULATION OF CRANK TRAIN DYNAMICS For simulation of crank train dynamics, a complex computational model of an engine (i.e. a virtual engine) is used. The dynamics is solved in the time domain which enables different physical problems, including various non-linearities, to be incorporated. The virtual engine is assembled as well as numerically solved in MBS (Multi-Body System) ADAMS which is a general code [2,4]. In general, the virtual engine includes all significant components necessary for dynamics analyses. The included module is a crank train, a valve train, a timing drive and a rubber damper [8]. Following analyses deal with the crank train as a main module of the virtual engine. The crank train module consists of solid model bodies, linearly elastic model bodies and constraints between them. Solid model bodies defined by location of centre of gravity, mass, and inertia tensor are a piston assembly, a connecting rod assembly, and a dynamometer rotor. The linearly elastic model bodies are modally reduced finite element models suitable for dynamic simulation [5,6]. These are: crankshaft, crankshaft pulley, flywheel, engine block, cylinder head, crank train sump, gear case. Cast-iron main bearing caps are included as an integral part of the engine block (different materials are considered) because the model of preloaded screws cannot be incorporated for this purpose. The computational model of block of the 3mb variant is derived from the standard one by necessary design modifications; therefore, changes in dynamic behaviour of the 3mb block are included in the virtual engine as well. A dumb-bell shaft connecting a flywheel with a dynamometer rotor is represented by a body with defined torsional stiffness and damping. These characteristics are adjusted on account of torsional vibration measurement. The interaction between the crankshaft and the engine block is ensured via a nonlinear hydrodynamic journal bearing model [3]. Virtual engine is excited by means of cylinder pressure, defined by high-pressure measurement, and via inertial forces from moving parts. Simulations start from 1000 rpm and are carried out to 6200 rpm. 2 MAIN BEARINGS LOAD The comparison of main bearing reaction forces for 5mb and 3mb variants, obtained from the virtual engine, is shown in Figure 1 and Figure 2, respectively. The maximum of main bearing reaction force means the greatest magnitude of this vector during a working cycle for respective rpm. The most loaded main bearings of the 5mb variant are bearings 2 and 4. However, there is an obvious a growth in the middle main bearing. The most loaded bearing of the 3mb variant is the middle bearing. More analyses also show a special importance of the middle main bearing for the engine block load; therefore, further attention is paid to this bearing. 162

164 Maximum of Main Bearing Reaction Forces Force [N] Force [N] Maximum of Main Bearing Reaction Forces Engine Load 100 % Engine Speed [rpm] Engine Load 100 % Figure 1: Maximum of main bearing reaction forces of the 5mb crank train Engine Speed [rpm] Figure 2: Maximum of main bearing reaction forces of the 3mb crank train 3 ENGINE BLOCK LOAD IN THE MIDDLE MAIN BEARING The crank train main bearing load is caused by gas forces, inertia forces, and their dynamic effects [7]. For desired power curve, the gas forces cannot be markedly changed. In order to affect the main bearing load, the only inertia forces must be considered. Inertia forces originate in: reciprocating parts, rotating parts. Inertia forces of reciprocating parts can be balanced only for the entire engine, regarding a car body vibration by a special balancing unit. These are secondary inertia forces in this case. a a a Fr1 mr η ω Mrp ω η Mrp Mr1 A T r ζ Fr4 A T Fr3 ξ Mr2 Mr3 ξ ζ 2 r F Figure 3: Inertia forces of rotating parts and its internal moments at a four-cylinder in-line engine Figure 4: Inertia moments of rotating parts at a four-cylinder in-line engine Primary inertia forces do not directly cause vibration of the engine and the car body because they are mutually disturbed by means of their moments, predominantly in the middle main bearing. This load, however, for given kinematic conditions cannot be 163

165 influenced, because piston and connecting rod assemblies of investigated engine cannot be changed. Nevertheless, modifications in crankshaft are possible. Inertia forces of rotating parts and its internal moments of discretized four-cylinder inline engine are shown in Figure 3, where mr is mass of rotating part reduced to the crank pin at crank radius r and a is a cylinder distance. If crank train deformations are neglected, angular velocity ω is considered to be constant and masses of particular crank train parts are supposed to be the same, the sum of rotating parts inertia forces is given by: 4 F h m r i r i 1 2 r mr r 2 mr r 2 mr r 2 0. (1) Resulting force is zero and reference point for resultant moment identification can be, therefore, chosen arbitrarily. By using the right-hand rule, the sum of moment with respect to the point A according to Figure 4, for instance, can be obtained by: 4 M m r i 1 i r r 2 3a mr r 2 2a mr r 2 a 0 0. (2) Zero resulting moment means that the car body is not excited by this crank train effects, however, this effects are disturbed inwards. If the crank train is divided by plane ηξ going through the crank train decision point T, see Figure 3, two twin-cylinders are virtually obtained in the mirror form. Resulting force of rotating parts of each of them is zero, while their resulting moment is not zero with magnitude: M rp Fr a mr r 2 a (3) and its vector rotates with crankshaft. Regarding the mirror form, both vectors have the identical magnitude but opposite direction and they are disturbed largely in the middle bearing. Similar effect can be observed also in the case of six-cylinder engines and their virtual triples. Permanent load of the middle main bearing, owing to inner moments of rotating parts, can be reduced by lowered rotating parts mass and by suitable counterweights. Each crankshaft web can be equipped by a counterweight, which corresponds to 8 counterweights for the standard four-cylinder crank train. For the main bearing load, this is a favourable solution due to balancing of each inertia force of rotating parts in point of action. Nevertheless, crankshaft mass is raised and its natural frequencies are lowered. The other method consists of balancing a twin-cylinder resultant moment of rotating parts; therefore, only 4 counterweights are required. This design brings load reduction for the middle main bearing in particular, however, crankshaft can be lighter. The 5mb and 3mb crankshafts are designed like this. Crankshafts of worldwide engine manufacturers used to be also designed as the above mentioned combination, where bigger 4 counterweights are situated at crank webs adjacent to the first, the middle, and last main bearing and other crank webs are equipped by smaller 4 counterweights. The discretization of the crank train rotating parts according to Figure 3 is not very suitable for purposes of the engine described due to crank throw asymmetry caused 164

166 by unilateral counterweight and a skew lightening bore of a crank pin. The splitting into two twin-cylinders is used and their product moment of inertia is further considered. The product moment of inertia of a twin-cylinder is described by: Prp M rp 2 Prp, ct Prp, cr, (4) where Prp, ct is the product moment of inertia of a twin-cylinder crank throws, determined from crankshaft CAD model, and Prp, cr is the product moment of inertia of rotating parts of a connecting rod couple given by: Prp, cr mcrrot ra, (5) where mcrrot is mass of rotating parts of a connecting rod assembly. In fact, the product moment of inertia is not a vector. However, if general moment of an inertia couple is expressed for unit angular velocity (ω = 1 rad s-1), its dimensionality corresponds to the product moment of inertia. Product moments of inertia of the 5mb and 3mb variants are determined this way. A A A Critical area of engine block x A Critical area of engine block Significant impact on load of critical area of engine block y peak-to-peak Fx peak-to-peak Fy Figure 5: Critical areas of a deep-skirt engine block and its load In order to describe influence of counterbalances size upon the middle main bearing load, a variant 3mb-c is derived from 3mb by attaching 0.05kg mass points directly on each counterweight of a modally reduced crankshaft. The influence of this modification upon the product moment of inertia of a twin-cylinder is determined by a position of these points with respect to each other and to crankshaft axis of rotation. The dynamic behaviour of variants 5mb, 3mb, and 3mb-c is simulated by the virtual engine and magnitude and direction of the middle main bearing load is investigated. 165

167 The deep-skirt engine block of the engine presented is equipped by separate main bearing caps. Each cap is connected with the block via highly preloaded screw couple. The cap position in the block is ensured by side facets with a transition fit. In spite of the screw preload, a slide side movement of the cap towards the block can occurs as a consequence of the bearing load and thermal distortions when engine operates. The movement introduced can cause a fatigue crack in the groove, which is the critical area of the engine block, see Figure 5. Only bearing load reduction is required for fatigue crack elimination on this engine and the problem is peak-to-peak value of the bearing transverse force. A polar diagram of the middle main bearing load is shown in Figure 5. The coordinate system is linked to the engine block, therefore, peak-to-peak value of Fy affects fatigue life at critical areas. Figure 6 shows a polar diagram of bearing force in the middle main bearing of all crankshaft variants at 2500 rpm. The main bearing load in the x-direction of 3-mainbearing crankshafts is more than double in comparison to the standard one. This is caused by firing in cylinders 2 and 3 which affects mainly the middle main bearing due to the crankshaft elasticity when bearings 2 and 4 are removed. The bearing load is influenced, especially, by gas forces at lower speed. A small growth in the vertical load of the 3mb-c variant is caused by a counterweight mass rising because its inertia force acts in nearly conformable direction, like gas forces near the top dead centre of pistons 2 and 3. The peak-to-peak-value of bearing transverse load is not significant at this speed. x x x 1,0 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0 1,0 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0 1,0 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0 y 143 % 166 % 100 % 5mb y 210 % 100 % 207 % y 3mb 3mb-c Figure 6: Polar diagram of a relative bearing force in the middle main load at 2500 rpm The situation at higher speed is described in Figure 7. In spite of markedly lightened piston and connecting rod assemblies, the inertia forces dominate in the middle bearing load at 6200 rpm. A wide peak-to-peak value of transverse bearing load can be observed at all variants. This is caused, particularly, by the inner moments of rotating parts. The meaning of these moments is clearly shown in Figure 7, where 166

168 x x 1,0 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0 1,0 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0 y 83 % x 1,0 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0, % 100 % significant decrease of peak-to-peak value of transverse force appears at 3mb-c variant. y 100 % 115 % 94 % 5mb 3mb 3mb-c y Figure 7: Polar diagram of a relative bearing force in the middle main load at 6200 rpm Force Load of Middle Main Bearing Along Transverse Direction R² = Initial level Fy Peak-to-peak Value [N] R² = Fully balanced internal moments of rotating parts Twin-cylinder Relative Product Moment of Inertia [ ] Figure 8: Influence of a twin-cylinder product moment of inertia upon peak-to-peak value of the transverse force in the middle main bearing 167

169 The conjoined reduction of load in vertical direction is favourable because this load affects contact pressure and then friction force between the cap and block in the negative. The influence of a twin-cylinder product moment of inertia upon peak-to-peak value of the transverse force in the middle main bearing at 6200 rpm is shown in Figure 8. The graph describes results of a parametric study in terms of the crank train dynamics simulation. The load dependence on the product moment of inertia (counterweight mass, respectively) is almost perfectly linear. A wider Fy peak-to-peak value of the 3mb variant caused by bearings 2 and 4 absence is obvious. The product moment of inertia sensitivity is, however, relatively marked, while the initial size of counterweights is small. Therefore further increase in size of counterweights can bring an important reduction of load at critical areas of the engine block. It is efficient to concentrate a counterweight material as far as possible from crankshaft axis. A counterweight size is restricted by a piston cooling spray jet and a crank case design, respectively. CONCLUSION Advanced computational methods are efficient tools for understanding complex dynamic effects concerning not only the crank train, but also other engine subsystems. The study described shows the utilization of these methods for analysing the influence of crankshaft counterweights upon the engine block load. The results reflect the meaning of inertia forces for engine block load at a naturally aspirated spark-ignition engine. The analysis output can be effectively used by crankshaft designers as it enables to reduce a number of crankshaft variants respecting the engine block load for both the standard engine and the experimental engine with lowered main bearings number. REFERENCES [1] [2] [3] [4] [5] [6] DRÁPAL L., NOVOTNÝ P., MARŠÁLEK O., RAFFAI P., PÍŠTĚK V. A Conceptual Study of Cranktrain with Low Friction Losses. MECCA Journal of Middle European Construction and Design of Cars, Vol. XI, No. 02, 2013, p. 611, ISSN NOVOTNÝ P. Virtual Engine A Tool for Powertrain Development. Inaugural Dissertation, Brno University of Technology, Czech republic, p. BUTENSCHÖN H. J. Das hydrodynamische, zylindrische Gleitlager endlicher Breite unter instationärer Belastung. PhD dissertation, Universität Karlsruhe, Germany, p. NOVOTNÝ P., PÍŠTĚK V. New Efficient Methods for Powertrain Vibration Analysis. Proceedings of the Institution of Mechanical Engineers, Part D, Journal of Automobile Engineering, Vol. 224, No. 5, 2010, p ISSN CRAIG R. R. Structural Dynamics. John Willey & Sons, p. ISBN REBBERT M. Simulation der Kurbewellendynamik unter Berücksichtigung der hydrodynamischen Lagerung zur Lösung motorakusticher Fragen. Ph.D. 168

170 [7] [8] dissertation, Rheinisch-Westfälischen Technischen Hochschule, Aachen, Germany, p. HEISLER H. Advanced Engine Technology. 1st edition. Oxford (Great Britain): Arnold, 1995, reprint 2002, 794 p. ISBN MSC.SOFTWARE. ADAMS/Engine Help. Newport Beach (CA): MSC Software Corporation, Version MD Adams R3. ACKNOWLEDGEMENT The research leading to these results has received funding from Specific research project of the Faculty of Mechanical Engineering, Brno University of Technology (FSIS ). 169

171 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES ADDRESSING THE LAST FRONTIER OF CLEAN DIESEL ENGINES: REAL DRIVING EMISSIONS OF NITROGEN COMPOUNDS Michal Vojtíšek1 Abstract Diesel engines, practical prime mover of large portion of on-road vehicles and mobile machinery, are being scrutinized for the adverse health effects of their exhaust emissions, which may become the limiting factor of their use. Following successful abatement of particulate matter emissions by diesel particle filters, the focus is shifting to nitrogen oxides (NOx) and byproducts of the efforts of their reduction, such as nitrous oxide, nitrogen dioxide and ammonia. Maintaining acceptably low levels of reactive nitrogen compounds and N2O under everyday real-world operation can be viewed as a necessary condition for future use of diesel engines in cities. 1. INTRODUCTION: ROLE OF DIESEL ENGINES Compression ignition (diesel) engines are currently a prime mover of majority of utility vehicles and mobile machinery thanks to their overall practicality, combining relatively high efficiency and reliability, relatively low investment and operating costs, and relatively compact energy storage in the form of liquid or gaseous fuels. Diesel engines dominate the sectors of larger utility vehicles and mobile machinery, including light and heavy trucks, buses, agricultural and construction machinery, and ships, dominating majority of medium and heavy duty applications except for aviation, dominated by turbines for larger aircraft, and for rail vehicles using electric drives powered from traction lines. Small machinery, motorcycles, small aircraft and majority of cars in the world are powered dominantly by spark ignition engines, except for Europe, where diesel cars account for tens of percent of car registrations and for about half of the miles driven. Very small and low-power machinery, such as power tools and bicycles, are dominated by battery electric drives. One of the key drawbacks of diesel engines are their exhaust emissions, which contain carcinogenic carbonaceous nanoparticles and nitrogen oxides, and which are released at the breathing level in the streets in the immediate proximity of people. Michal Vojtíšek, Center for Sustainable Mobility, Faculty of Mechanical Engineering, Czech Technical University in Prague, Technicka 4, Prague, CZ, michal.vojtisek@fs.cvut.cz 1 170

172 2. DIESEL EXHAUST-RELATED RISKS Diesel engines emit particles predominantly in the tens of nanometers (nm) [1], which, readily deposit in lungs [2-3]. Particles in lower tens of nm and smaller can penetrate through cell membranes into the blood, and have a wide and detrimental effect on human health [4]. Diesel particulate matter contains various compounds known to be carcinogenic, mutagenic, genotoxic, and otherwise dangerous to human health, including polycyclic aromatic hydrocarbons such as benzo[a]pyrene, but also lesser known compounds such as 3-nitro-benz[a]benzanthrone [5]. Diesel particulate matter as a whole has been found to be carcinogenic [6-7] and was declared as such by the California Air Resources Board, United States Occupational Safety and Health Administration (OSHA), World Health Organization (WHO) and as Class I carcinogen by the International Agency for Research of Cancer (IARC) [8]. Carbonaceous particles from engines are therefore more hazardous, on the equivalent mass basis, than average particles in the air [9], and reduction of diesel particulate matter has been found to be, on a mass equivalent bases, 4-9 times more beneficial than the reduction in overall PM2.5 concentrations [10]. Another pollutant of major concern are nitrogen oxides (NOx). The majority of NOx from the combustion process are emitted as NO, which oxidizes into a brownish irritant gas NO 2; NOx also participate in the formation of tropospheric ozone [11]. NO2 has been associated with mortality and hospital admissions for a range of respiratory and cardiovascular endpoints [12-14]. Proximity to sources of internal combustion engine exhaust has been historically associated with increased risks of various chronic health problems [15-17], including myocardial infarction [18], chronic cough [19], inflammatory processes [20], neurobehavioral problems [21], and increased blood pressure [22]. Diesel exhaust exposure is further connected to decreased quality of sperm [23]. Prenatal exposure can lead to premature birth or low birth weight, congenital abnormalities and an increase in the mortality of infants [24]. Oyana and Lwebuga-Mukasa [25] have found a cluster of excessively high prevalence of asthma within close proximity of truck waiting area of the U.S.-Canada border crossing in Buffalo, NY; in another study, asthma prevalence was found to be correlated to truck traffic volumes [26]. Overall, air pollution, much of which at the breathing level comes from transportation, is one of the leading causes of premature deaths worldwide, with the number of premature deaths being roughly one order of magnitude higher compared to traffic accidents. In the Czech Republic for 2013, the National Health Institute reports over 7 thousands premature deaths due to air pollution [27] compared to 583 traffic accident related deaths [28]. In the EU, particulate matter is responsible for approximately 403 thousands, and NOx for approximately 72 thousands, of premature deaths annually [29]. 3. ABATEMENT OF DIESEL PARTICULATE MATTER In the U.S., and later in the EU, production heavy-duty diesel engines have been fitted with diesel oxidation catalysts (DOC) and diesel particle filters (DPF), and have been performing with very low emissions of particulate matter [30] for over a decade. Generally, considering a limit of 6 x 1011 solid particles per kwh for Euro 6 heavy-duty on-road vehicles, diesel engine fuel economy of 250 g/kwh, air-fuel ratio of approximately 25-50, and density of air 1.22 kg/m3, yielding an equivalent of 5-10 m3 171

173 of exhaust (at normal conditions) per kwh, the maximum concentrations of solid particles in the exhaust would be thousands of particles per cm3 (#/cm3), a value not atypical for worse cases ambient concentrations along busy streets, and an order of magnitude above the average urban background concentrations of approximately 7 thousands #/cm3 in Prague and several other cities [31]. In reality, the concentrations downstream of the DPF are typically far below ambient. However, during regeneration of the DPF a significant increase about or over three orders of magnitude - of the emitted number of particles is observed, with apparent nucleation mode peaking at approximately 10 nm [32]. The DPF filtration efficiency is, obviously, dependent on the proper functioning of the DPF. The Swiss retrofit program target filter failure rate (fraction of vehicles with a DPF failure) is below 1% [33]. Given the nominal DPF efficiency on the order of 99%, a 1% failure rate would double the total fleet emissions. In Netherlands, non-functional DPF on passenger cars were estimated to increase the fleet emissions (all vehicles with DPF) by a factor of 5-6 [34-35]. In some countries, including the Czech Republic, the overall benefits of DPF are further reduced by deliberate removal of the DPF. This matter should be addressed by public outreach, improved periodical inspection procedures, and on-road enforcement. Luckily, owing to the high efficiency of the DPF, malfunction or absence of the DPF can be readily identified by looking into the tailpipe. 4. REACTIVE NITROGEN SPECIES IN DIESEL EXHAUST Nitrogen monoxide (NO) is formed at high temperatures during the combustion process from atmospheric nitrogen and oxygen. In fact, under most conditions, virtually all NOx leave the engine as NO. NO is oxidized in the atmosphere into nitrogen dioxide, NO2, a brownish irritant gas. Photolysis of NO2 by ultraviolet light (contained in sunlight) yields tropospheric (ground level) ozone. Nitrogen oxides (NO x) are the sum of NO and NO2. NOx are also precursors of N2O4, N2O5 and acids HNO2 [36-37] and HNO3, and contribute to the formation of secondary particulate matter in the form of nitrates (discussed further in [38]). On some diesel engines, oxidation catalysts are used to convert NO into NO2, as higher concentrations of NO2, around tens of percent, aid both in the combustion of soot in DPF and in the reduction of NO x in Selective Catalytic Reduction (SCR) catalysts. As a result, NO2 from new engines accounts for tens of percent of NOx [39-40], causing the concentrations of NO2 on streets to increase. Ammonia (NH3) is formed in some types of catalytic exhaust aftertreatment devices under some conditions. On spark ignition engines running typically at stoichiometric air-fuel ratio, reduction of NOX in three-way catalysts under fuel-rich conditions produces ammonia (NH3) [41], particularly at lower exhaust gas temperatures [42]. SCR systems and lean NOX traps (LNT, also called NOx storage and reduction catalysts, NSRC) employed to reduce NOX in diesel engines can produce, under some conditions, NH3 and a potent greenhouse gas nitrous oxide (N 2O) [43-47]. In LNT, NH3 is produced as a by-product during reduction of NOx, while in SCR, NH3 is used as reducing agent, and its emissions are a slip of unreacted portion. NH 3 is produced by thermal decomposition of urea, which yields isocyanic acid (HCNO) and ammonia, with HCNO further hydrolyzing into NH3 [34,48-49]. NH3 is a precursor of secondary inorganic aerosol, namely ammonium nitrate and sulphate [50-51] which are the most abundant atmospheric secondary inorganic aerosols in many regions [52-53]. Large portion of NH3 in urban air in the U.S. and China originate from motor vehicles [54]. 172

174 5. NOX ABATEMENT On spark ignition engines, NOx has been successfully abated by the combination of three-way catalysts and by maintaining stoichiometric air-fuel ratio through closed-loop control of the quantity of fuel injected. On diesel engines, where three-way catalyst cannot be used due to excess air, some level of reduction of NO x emissions has been achieved through reducing peak combustion temperatures via delaying injection timing, exhaust gas recirculation (EGR), and later by advanced combustion strategies such as low temperature combustion (LTC) and partially premixed charge compression ignition combustion [55-56]. Both EGR and advanced combustion concepts are, however, effectively limited to lower engine loads. More recently, various catalytic devices have been explored, of which LNT and SCR have shown relatively high potential [57] and are currently used in production vehicles. LNT (for overview, see [58]) operates alternately in a long (on the order of minute or minutes) lean or accumulation mode, during which NOx are stored (i.e., as barium nitrate) in the catalyst, and in a short (on the order of seconds) rich or regeneration mode, during which the engine operates with excess fuel, NOx are released, and reduced in a manner similar to a three-way catalyst. SCR (reviewed in [59]) use a reducing agent, aqueous solution of urea, which is injected upstream of the SCR, is mixed with the exhaust, and thermally decomposes into NH3, which is used to reduce NOx; NH3 is also stored in the SCR. Delaying combustion, EGR and LNT are typically associated with a slight (one to several percent) fuel penalty. SCR requires a reducing agent in quantities constituting several percent of diesel fuel usage. 6. ON-ROAD NOX EMISSIONS VS. TYPE APPROVAL LIMITS Since all of NOx reduction strategies are associated with both initial and operating costs, they offer a potential to unscrupulous vehicle operators or manufacturers to disable NOx reduction to achieve (a very small) reduction in the consumption of fuel or reducing agent. Over last two decades, various cases emerged where engine, in order to gain a minor improvement in the fuel economy, switched from compliant to excess NOx emissions during over-the-road operation; this has been the case with several U.S. heavy duty engine manufacturers settled as the Consent Decree [60]. On heavy-duty vehicles, such practice has been curbed by extending the emissions tests to real-world operation and by establishing not-to-exceed limits [61] and in-use compliance factors or conformity factors [62]. European automobile diesel engines produced throughout the last decade, however, have been reported to emit substantially, often by an order of magnitude, more NO x on the road than during the type approval test [63-66]. Marginal design and ethically problematic calibration of many light vehicle diesel engines, which exhibit much higher emissions of NO x in ordinary operation than during type approval tests, create a paradox, where per-km emissions of NOx may be lower for buses than for cars. The mentioned issues necessitate a critical assessment of exhaust emissions under real driving conditions. 7. ON-ROAD MEASUREMENT OF NITROGEN COMPOUNDS Of reactive nitrogen compounds, NO, NO2, NH3, and potentially HCNO, are the most critical ones. The measurement also needs to include N 2O, which is stable, nonreactive, non-hazardous, but is a greenhouse gas, which is approximately 265 times 173

175 as potent, on an equivalent mass basis, as CO 2 [67]. Such measurement should be facilitated not only in the laboratory, but also on the road, with the instruments mounted in the vehicle. Given the highly transient nature of engine operation, time resolution of on-board systems should be on the order of 1 Hz. NO has been historically measured by chemiluminescence analyzers, which, when preceeded by a suitable catalyst, can also measure the sum of NO x or NOx+NH3, depending on the catalyst used and its working temperature. Other methods include NDIR for NO and NDUV [68] for NO, NO2 and possibly NH3. Inspection grade instruments and some of the early on-board systems used fast electrochemical cells for NO measurement [69-71]. All mentioned compounds can also be measured in the infrared spectra, however, the presence of large amounts of water vapor constitutes a major restriction. Water can be removed by membrane dryers to measure, for example, NO, but removal of water affects the measurement of NH 3. Other approach is to use high optical resolution spectroscopy, employing either a Fourier transform spectrometer (FTIR) or laser diodes tunable to specific wavelengths, such as quantum cacsade lasers (QCL). Both FTIR and quantum cascade laser (QCL) have been successfully used for online measurements of NH3 in vehicular exhaust [72]. The QCL has also been used on-road to measure NH3 and N2O [73,74], however, its size and weight restricted such measurement for heavy vehicles. FTIR is increasingly used in vehicle exhaust measurement as the most universal gaseous pollutant measurement system, despite chemiluminescence offering lower detection limits for NO and NOx and despite the limitation of FTIR in measuring total hydrocarbon emissions, which remains the domain of flame ionization detectors. The use of FTIR for exhaust measurement has been reviewed in [75,76]. The use of FTIR in a moving vehicle has been, so far, limited due to technical challenges. Jeter [77] has used a pair of laboratory FTIR with a membrane sample dryer (a technique not compatible with NH3 measurement) and a 10-liter optical path length mounted in an instrumented vehicle to assess very low emission levels from a passenger car. A smaller but still relatively bulky and relatively slow (30 s time resolution) instrument with a 13 m optical path length was used by Reyes [78]. A portable, fast system has been used by Daham [79], however, the system worked at a relatively low (4 cm-1) optical resolution, while a substantially higher resolution at least 0.5 cm-1 has been recommended, along with the tunable laser diode approach, for NH3 measurement [41]. A compact portable FTIR setup has been developed, using industrial FTIR components, and used within the MEDETOX project at the Technical University of Liberec. The FTIR features a Michelson interferometer, ZnSe optics, liquid nitrogen cooled mercury cadmium telluride detector, and a compact 0.2 dm3, 6 m optical path length cell. The cell and the sample train run at 121 C and 15 dm3/min sample flowrate. The FTIR runs at 0.5 cm-1 optical and approximately 0.4 Hz time resolution. The system weights 35 kg, consumes W and can run for 6-8 hours on a pair of lithiumyttrium-iron-phosphate batteries (each battery V, 90 Ah, 15 kg). By deconvolution of the absorption spectra, concentrations of various pollutants of interest, including NO, NO2, NH3, N2O, CH4, CO and CO2, were calculated using a method developed in-house and using library spectra or spectra obtained by measuring calibration gases. The system has been recently compared against laboratory and other portable instruments at the European Commission Joint Research 174

176 Center [80]. The limit of detection, calculated assuming 3 times the standard deviation of the background measured, for NH3 and NO were 1 ppm. A similar setup using faster (1 Hz) industrial FTIR (Nicolet Antaris IGS) with a 0.3 dm3, 5 m path length cell, with about double the mass and double power consumption, has been built at the Czech Technical University and recently used for evaluation of emissions of compressed natural gas vehicles [81]. 8. SYNOPSIS Compression ignition (diesel) engines are currently a prime mover of majority of utility vehicles and mobile machinery thanks to their overall practicality, combining relatively high efficiency and reliability, relatively low investment and operating costs, and relatively compact energy storage in the form of liquid or gaseous fuels. One of the key drawbacks of diesel engines is their exhaust emissions, which contain carcinogenic carbonaceous nanoparticles and nitrogen oxides, and which are released at the breathing level in the streets in the immediate proximity of people. The issue of particulate matter appears to have been successfully resolved through the use of diesel particle filters, and the abatement of diesel particulate matter is becoming merely an issue of proper deployment and maintenance of the particle filters. Nitrogen oxides (NOx), formed at high temperatures during the combustion process, and contributing to the formation of ground level ozone, remained a challenge, with combustion control offering only a modest reduction, and has been addressed only recently by the introduction of selective catalytic reduction (SCR), and also to a yet unknown extent by NO x storage and reduction catalysts. Such efforts brought about an inadvertent creation of nitrogen dioxide, ammonia and other reactive nitrogen compounds, and a potent greenhouse gas nitric oxide. SCR efficiency also drops during prolonged low load operation. Marginal design and ethically problematic calibration of many light vehicle diesel engines, which exhibit much higher emissions of nitrogen compounds in ordinary operation than during type approval tests, create a paradox, where per-km emissions of NOx are often lower for buses than for cars. The mentioned issues necessitate a critical assessment of exhaust emissions under real driving conditions. This paper summarizes the underlying technical and public health issues, reviews the instrumentation available for real driving emissions measurement of nitrogen compounds, and discusses results of several test campaigns conducted with the participation of the author. Overall, it appears that clean diesel engines are within the reach of current technology, but that many vehicle owners and operators and many automobile manufacturers fail to follow the best available technology practice, making diesel engines, en-bloc, prone to their prohibition in urban areas mindful of their air quality. ACKNOWLEDGEMENT This work was supported by the The Ministry of Education, Youth and Sports program NPU I (LO), project # LO1311 Development of Vehicle Centre of Sustainable Mobility, and by the Technology Agency of the Czech Republic, programme Centres of Competence, project # TE Josef Božek Competence Centre for Automotive Industry. 175

177 REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] KITTELSON, D.B., Engines and nanoparticles. J. Aerosol Sci. 29, doi: /s (97) GEHR, P., BLANK, F., ROTHEN-RUTISHAUSER, B.M., Fate of inhaled particles after interaction with the lung surface. Paediatr. Respir. Rev. 7. doi: /j.prrv GERDE, P., MUGGENBURG, B.A., LUNDBORG, M., DAHL, A.R., The rapid alveolar absorption of diesel soot-adsorbed benzo[a]pyrene: Bioavailability, metabolism and dosimetry of an inhaled particle-borne carcinogen. Carcinogenesis 22. KÜNZLI, N., et al., Public-health impact of outdoor and traffic-related air pollution: A European assessment. Lancet 356. doi: /s (00) ENYA T., et al., 1997: 3-Nitrobenzanthrone, a Powerful Bacterial Mutagen and Suspected Human Carcinogen Found in Diesel Exhaust and Airborne Particulates. Environmental Science and Technology, 31, ATTFIELD M.D., et al., 2012: The Diesel Exhaust in Miners Study: A Cohort Mortality Study With Emphasis on Lung Cancer. J Natl Cancer Inst 104(11): SILVERMAN, D.T., et al., 2012: The Diesel Exhaust in Miners Study: A Nested Case-Control Study of Lung Cancer and Diesel Exhaust. J Natl Cancer Inst, 104, IARC International Agency for Research on Cancer and World Health Organization, IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Volume 92: Some Non-heterocyclic Polycyclic Aromatic Hydrocarbons and Some Related Exposures. Online at KRZYZANOWSKI M, KUNA-DIBBERT B, SCHNEIDER J, eds. Copenhagen: World Health Organization; Health Effects of Transport-Related Air Pollution. JANSSEN, N.A.H., et al., Black Carbon as an Additional Indicator of the Adverse Health Effects of Airborne Particles Compared with PM10 and PM2.5. Environ Health Perspect December; 119(12): SEINFELD, J. H., & PANDIS, S. N. (1998). Atmospheric Chemistry and Physics. John Wiley & Sons, Hoboken, NJ, USA. FAUSTINI, A., RAPP, R., FORASTIERE, F., Nitrogen dioxide and mortality: review and meta-analysis of long-term studies. Eur. Respir. J. 44, US EPA, Integrated Science Assessment for Oxides of Nitrogen - Health Criteria (Second External Review Draft), EPA/600/R-14/ [Accessed February 2016]. United States Environmental Protection Agency, Washington DC. BRAND, Christian. Beyond Dieselgate : Implications of unaccounted and future air pollutant emissions and energy use for cars in the United Kingdom. Energy Policy, 2016, 97: LEWTAS, J., Air pollution combustion emissions: Characterization of causative agents and mechanisms associated with cancer, reproductive, and cardiovascular effects. Mutat. Res. Rev. Mutat. Res doi: /j.mrrev MCENTEE, J.C., OGNEVA-HIMMELBERGER, Y., Diesel particulate matter, lung cancer, and asthma incidences along major traffic corridors in MA, USA: A GIS analysis. Heal. & place 14. doi: /j.healthplace BALMES, J.R., et al., Exposure to traffic: Lung function and health status in adults with asthma. J. Allergy Clin. Immunol doi: /j.jaci BARD, D., et al., Traffic-related air pollution and the onset of myocardial infarction: Disclosing benzene as a trigger? A small-area case-crossover study. PLoS One 9. doi: /journal.pone GRONEBERG-KLOFT, B., et al., Analysing the causes of chronic cough: Relation to diesel exhaust, ozone, nitrogen oxides, sulphur oxides and other environmental factors. J. Occup. Med. Toxicol. 1. doi: / DOORNAERT, B., et al., Negative impact of DEP exposure on human airway epithelial cell adhesion, stiffness, and repair. Am. J. Physiol. - Lung Cell. Mol. Physiol KILBURN, K.H., Effects of diesel exhaust on neurobehavioral and pulmonary functions. Arch. Environ. Health 55. FUKS, K.B., et al., Association of long-term exposure to local industry- and traffic-specific particulate matter with arterial blood pressure and incident hypertension. Int. J. Hyg. Environ. Health. doi: /j.ijheh

178 [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] ŠRÁM, R.J., et al., Adverse reproductive outcomes from exposure to environmental mutagens. Mutat. Res. - Fundam. Mol. Mech. Mutagen doi: /s (99) GHOSH, J.K.C., et al., Assessing the influence of traffic-related air pollution on risk of term low birth weight on the basis of land-use-based regression models and measures of air toxics. Am. J. Epidemiol doi: /aje/kwr469 OYANA T.J., LWEBUGA-MUKASA J.S., 2004: Spatial Relationships Among Asthma Prevalence, Health Care Utilization, and Pollution Sources in Neighborhoods of Buffalo, New York. Journal of Environmental Health. Apr2004, Vol. 66 Issue 8, p LWEBUGA-MUKASA J.S., OYANA T., THENAPPAN A., AYIROOKUZHI S.J., 2004: Association Between Traffic Volume and Health Care Use for Asthma Among Residents at a U.S. Canadian Border Crossing Point. Journal of Asthma, Vol. 41, No. 3, Pages PUKLOVÁ V., LUSTIGOVÁ M., KAZMAROVÁ H., KOTLÍK B., 2013: Ke vlivu znečištění ovzduší na úmrtnost v České Republice. Hygiena, 58(1), Policie ČR, Statistika nehodovosti. ( ) EEA, Air quality in Europe report. EEA (European Environment Agency), Copenhagen. See also Premature deaths attributable to fine particulate matter (PM2.5), ozone (O3) and nitrogen dioxide (NO2) exposure in 2012 in 40 European countries and the EU 28, online at HESTERBERG, T.W., et al., P.A., Particulate matter in new technology diesel exhaust (NTDE) is quantitatively and qualitatively very different from that found in traditional diesel exhaust (TDE). J. Air Waste Manag. Assoc. 61. doi: / LANZINGER, Stefanie, et al. Associations between ultrafine and fine particles and mortality in five central European cities Results from the UFIREG study. Environment international, 2016, 88: MAMAKOS A., MARTINI G. (2011): Particle Number Emissions During Regeneration of DPFequipped Light Duty Diesel Vehicles. European Commission Joint Research Center publication no. JRC 64870, EUR EN ISBN MAYER, A., CZERWINSKI, J., PÉTERMANN, J., WYSER, M. et al., "Reliability of DPFSystems: Experience with 6000 Applications of the Swiss Retrofit Fleet," SAE Technical Paper , 2004, doi: / KADIJK G., SPREEN J.S.: Roadworthiness Test Investigations of Diesel Particulate Filters on vehicles, TNO, Delft, Netherlands, KADIJK G., SPREEN J.S., VAN DER MARK P.J.: Investigation into a Periodic Technical Inspection test method to check for presence and proper functioning of Diesel Particulate Filters in light-duty diesel vehicles. TNO report no TL-RAP , TNO, Delft, Netherlands, GUTZWILLER L., ARENS F., BALTENSPERGER, U., GÄGGELER H.W., AMMANN M.: Significance of Semivolatile Diesel Exhaust Organics for Secondary HONO Formation. Environmental Science & Technology (4), DOI: /es015673b KURTENBACH, R., et al. Investigations of emissions and heterogeneous formation of HONO in a road traffic tunnel. Atmospheric Environment, 2001, 35.20: HEEB, N.V., et al. Reactive nitrogen compounds (RNCs) in exhaust of advanced PM NO x abatement technologies for future diesel applications. Atmospheric environment, 2011, 45.18: OLSEN, D.B., KOHLS, M., ARNEY, G., Impact of oxidation catalysts on exhaust NO2/NOx ratio from lean-burn natural gas engines. J. Air Waste Manag. Assoc. 60, doi: / CARSLAW, D.C., Evidence of an increasing NO2/NOX emissions ratio from road traffic emissions. Atmos. Environ. 39. doi: /j.atmosenv SUAREZ-BERTOA, R., ZARDINI, A.A., ASTORGA, C., Ammonia exhaust emissions from spark ignition vehicles over the New European Driving Cycle. Atmos. Environ. 97. doi: /j.atmosenv HEEB N.V., et al. (2006): Three-way catalyst-induced formation of ammonia velocity- and acceleration-dependent emission factors. Atmospheric Environment, 40, 31, , doi /j.atmosenv JEON, J., LEE, J.T., PARK, S., Nitrogen Compounds (NO, NO2, N2O, and NH3) in NOx Emissions from Commercial EURO VI Type Heavy-Duty Diesel Engines with a Urea-Selective 177

179 [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] Catalytic Reduction System. Energy and Fuels 30, doi: /acs.energyfuels.6b01331 JOHNSON, T., Vehicular Emissions in Review. SAE Int. J. Engines 9. doi: / CZERWINSKI, J., ZIMMERLI, Y., MAYER, A., HEEB, N., BERGER, H., D URBANO, G., Testing of SCR-systems on HD-vehicles-TeVeNOx. SAE Tech. Pap. 1. doi: / MRÁČEK, D., KOČÍ, P., CHOI, J.S., PARTRIDGE, W.P., New operation strategy for driving the selectivity of NOx reduction to N2, NH3 or N2O during lean/rich cycling of a lean NOx trap catalyst. Appl. Catal. B Environ. 182, doi: /j.apcatb SUAREZ-BERTOA, R., ZARDINI, A.A., KEUKEN, H., ASTORGA, C., Impact of ethanol containing gasoline blends on emissions from a flex-fuel vehicle tested over the Worldwide Harmonized Light duty Test Cycle (WLTC). Fuel 143. doi: /j.fuel BAIK, J.H., YIM, S.D., NAM, IS. et al. Topics in Catalysis (2004) 30:37, doi: /B:TOCA KOEBEL M., ELSENER M., KLEEMANN M.: Urea-SCR: a promising technique to reduce NOx emissions from automotive diesel engines. Catalysis Today, 59, 3-4, 2000, FINLAYSON-PITTS, B. J., AND J. N. PITTS, Chemistry of the Upper and Lower Atmosphere, Elsevier, New York. BEHERA, S.N. AND SHARMA M., Investigating the potential role of ammonia in ion chemistry of fine particulate matter formation for an urban environment. Science of The Total Environment, (17), KIM, B.M., TEFFERA, S., ZELDIN, M.D., Characterization of PM25 and PM10 in the South Coast Air Basin of Southern California: Part 1 Spatial Variations. Journal of the Air & Waste Management Association, (12), SILLANPÄÄ, M., et al., 2006: Chemical composition and mass closure of particulate matter at six urban sites in Europe. Atmospheric Environment, 40, Supplement 2 (0), SUN, K., et al. Vehicle emissions as an important urban ammonia source in the United States and China. Environ. Sci. Technol, 2017, 51.4: KIMURA, S., AOKI, O., KITAHARA, Y., AND AIYOSHIZAWA, E., "Ultra-Clean Combustion Technology Combining a Low-Temperature and Premixed Combustion Concept for Meeting Future Emission Standards," SAE Technical Paper , 2001, doi: / JACOBS, Timothy J.; ASSANIS, Dennis N. The attainment of premixed compression ignition low-temperature combustion in a compression ignition direct injection engine. Proceedings of the Combustion Institute, 2007, 31.2: JOHNSON, T.V. Diesel emission control: 2001 in Review. SAE Technical paper MRÁČEK D.: Analysis and Modelling of Spatiotemporal Concentration Profiles in NOx Reduction Automotive Catalytic Converters. Ph.D. dissertation, VŠCHT Praha, GUAN, Bin, et al. Review of state of the art technologies of selective catalytic reduction of NO x from diesel engine exhaust. Applied Thermal Engineering, 2014, 66.1: USDOJ:, United States Department of Justice: Clean Air Act diesel engine cases. [WWW Document]. URL United States Code of Federal Regulations, volume 40, part VLACHOS, T., BONNEL, P., PERUJO, A., WEISS, M. et al., "In-Use Emissions Testing with Portable Emissions Measurement Systems (PEMS) in the Current and Future European Vehicle Emissions Legislation: Overview, Underlying Principles and Expected Benefits," SAE Int. J. Commer. Veh. 7(1): , 2014, doi: / FRANCO, V., SÁNCHEZ, F.P., GERMAN, J., MOCK, P., Real-word exhaust emissions from modern diesel cars a meta-analysis of PEMS emissions data from EU (EURO 6) and US (TIER 2 BIN 5/ULEV II) diesel passenger cars. ICCT Int. Counc. Clean Transp. LIGTERINK, N., KADIJK, G., MENSCH, P. VAN, HAUSBERGER, S., REXEIS, M., Investigations and real world emission performance of Euro 6 light-duty vehicles. WEISS, M., et al., Will Euro 6 reduce the NO x emissions of new diesel cars? - Insights from on-road tests with Portable Emissions Measurement Systems (PEMS). Atmos. Environ. 62, doi: /j.atmosenv YANG, L., FRANCO, V., MOCK, P., KOLKE, R., ZHANG, S., WU, Y., GERMAN, J., Experimental Assessment of NOx Emissions from 73 Euro 6 Diesel Passenger Cars. Environ. Sci. Technol. 49, doi: /acs.est.5b

180 [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] IPCC (Intergovernmental Panel for Climate Change) Fourth Assessment Report: Climate Change, MEINEL, H.; JUST, Th. Measurement of NOx exhaust emissions by a new NDUV analyzer. In: AIAA, Aerospace Sciences Meeting VOJTISEK-LOM, Michal; COBB, J. T. Vehicle mass emissions measurement using a portable 5-gas exhaust analyzer and engine computer data. Proceedings: Emission Inventory, Planning for the Future, BRETON, Leo Alphonse Gerard. Real-time on-road vehicle exhaust gas modular flowmeter and emissions reporting system. U.S. Patent No 6,148,656, VOJTISEK-LOM, M. AND ALLSOP, J., "Development Of Heavy-Duty Diesel Portable, On-Board Mass Exhaust Emissions Monitoring System With NOx, CO2 And Qualitative PM Capabilities," SAE Technical Paper , 2001, doi: / SUAREZ-BERTOA, Ricardo, et al. Intercomparison of real-time tailpipe ammonia measurements from vehicles tested over the new world-harmonized light-duty vehicle test cycle (WLTC). Environmental Science and Pollution Research, 2015, 22.10: SUAREZ-BERTOA, Ricardo, et al. On-road measurement of NH 3 and N 2 O emissions from a Euro V heavy-duty vehicle. Atmospheric Environment, 2016, 139: MENDOZA-VILLAFUERTE, P., et al. NOx, NH3, N2O and PN real driving emissions from a Euro VI heavy-duty vehicle. Impact of regulatory on-road test conditions on emissions. Science of The Total Environment, 2017, 609: ADACHI, M., Emission measurement techniques for advanced powertrains. Meas. Sci. Technol. 11. doi: / /11/10/201 REYES, F., et al. Technical Note: Analysis of non-regulated vehicular emissions by extractive FTIR spectrometry: tests on a hybrid car in Mexico City. Atmospheric Chemistry and Physics, 2006, 6.12: JETTER, J., MAESHIRO, S., HATCHO, S., AND KLEBBA, R., "Development of an On-Board Analyzer for Use on Advanced Low Emission Vehicles," SAE Technical Paper , 2000, doi: / REYES F., GRUTTER M., JAZCILEVICH A., GONZALEZ-OROPEZA R., 2006: Analysis of nonregulated vehicular emissions by extractive FTIR spectrometry: tests on a hybrid car in Mexico City. Atmospheric Chemistry and Physics Discussions, European Geosciences Union, 2006, 6 (4), pp DAHAM, B., ANDREWS, G., LI, H., BALLESTEROS, R., BELL M.C., TATE J., ROPKINS K., "Application of a Portable FTIR for Measuring On-road Emissions," SAE Technical Paper , 2005, doi: / SUAREZ-BERTOA, R., MENDOZA-VILLAFUERTE, P., RICCOBONO, F., VOJTISEK, M., PECHOUT, M., PERUJO, A., ASTORGA, C., On-road measurement of NH3 emissions from gasoline and diesel passenger cars during real world driving conditions, Atmospheric Environment (2017), doi: /j.atmosenv (in press) VOJTÍŠEK M., BERÁNEK V., KLÍR V., JINDRA P., PECHOUT M., VOŘÍŠEK T.: On-road and laboratory emissions of NO, NO2, NH3, N2O and CH4 from late-model EU light utility vehicles: Comparison of diesel and CNG. Science of the Total Environment, 2017, in review. 179

181 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES EMISSIONS AND PERFORMANCE OF A PASSENGER CAR SIZE DIESEL ENGINE FUELLED WITH HVO -DIESEL FUEL MIXTURES Ivan Bortel1, Jiří Vávra2, Michal Takáts3 Abstract The current requirements to reduce CO2 emissions may be met via several ways. In addition to increased combustion engine thermal efficiency, the use of renewable fuels is a way to improve the balance of CO2. In case of compression ignition engines a HVO (Hydrotreated Vegetable Oil) can be an alternative to FAME (Fatty acid methyl ester). This study presents the experimental results of comparison of standard diesel fuel, pure HVO and their mixture tested on a passenger car size single cylinder compression ignition engine equipped with a contemporary common rail injection. Testing cycle was based on a WHSC (World Harmonized Stationary Cycle). Common gaseous emissions, smoke number, opacity, particulate matter (PM) and particle number (PN) were measured. Results confirm positive or neutral influence on measured emission components and performance parameters. 1 INTRODUCTION Compression ignition (CI) engines are used in a wide range of applications including passenger car engines. The main advantage of CI engines in comparison with spark ignition ones (SI) is a higher overall fuel conversion efficiency [1] and related reduction of emissions of CO2 from passenger cars. Renewable biofuels can help to reduce Green-House Gases (GHG) including CO2 with benefits in Particulate Matter (PM) emissions [2], [3] and particle number (PN) [4]. Nowadays, fatty acid methyl esters (FAME), ordinarily known as biodiesel, produced from vegetable oils, waste cooking oils, and animal fats seems to be the biofuel of choice for fossil diesel fuel replacement [4]. However, FAME fuel usage brings in inconveniences like oil aging [5], limited storage time of fuel [6] and cold properties of FAME [7]. Hydrotreated vegetable oil (HVO) is made by a refinery-based process that converts vegetable oils and animal fat into paraffinic [8], [9]. HVO fuel does not have the Ivan Bortel, ČVUT v Praze, Fakulta strojní, Technická 4, Praha 6, ivan.bortel@fs.cvut.cz Jiří Vávra, ČVUT v Praze, Fakulta strojní, Technická 4, Praha 6, jiri.vavra@fs.cvut.cz 3 Michal Takáts, ČVUT v Praze, Fakulta strojní, Technická 4, Praha 6, michal.takats@fs.cvut.cz

182 detrimental effects of ester-type biodiesel fuels, such as problems mentioned above, increased NOx emission and deposit formation. HVO is a straight chain of paraffinic hydrocarbons that are free of aromatics, oxygen and sulphur and has a high cetane number [8]Chyba! Nenalezen zdroj odkazů.. Typical density of HVO is between 770 to 790 kg/m3 [7] and consequently the volumetric heating value of HVO is 5% lower than fossil diesel fuel [9]. The article [9] states, that injection volume quantity of HVO is about 5 % higher than injection volume quantity of standard diesel fuel due the lower bulk modulus of HVO in comparison with standard diesel fuel. Other explanations are mentioned in [7] and [9]. Therefore it looks, that energy quantity in fuel per cycle is not significantly affected by using HVO in comparison with standard diesel fuel [7], [9]. High cetane number of HVO accelerates the start of combustion in low and medium loads, but is of less influence at high loads. Furthermore, the high cetane number particularly improves cold startability, reduces noise and emissions [7]. Engine's energy efficiency can be slightly improved with HVO [7], [11]. Using pure HVO or blends of HVO and standard diesel fuel reduces emissions of hydrocarbons (HC), carbon monoxide (CO) [7], [8]Chyba! Nenalezen zdroj odkazů., [14] and particle matter (PM) emissions [7], [12], [13], [9], especially at low temperatures [7]. Some authors observe small increase in PM with HVO blends, but only in several cases, not in general [4]. Particle number (PN) emissions showed reductions with HVO fuel [4], [13]. Influence of HVO on NOx emission seems to be ambiguous [5]. In several articles the reduction of NOx is stated [7], [8], while somewhere the negligible effect [4], [12] [14], or growth of NOx emissions is mentioned [13]. HVO content in regular diesel fuel does not cause clogging or premature deactivation of aftertreatment devices, like DPFs, DOCs or SCRs [7]. Authors in [15] state, that regeneration ability of DPF is improved by usage of HVO. This study extends the available results of HVO blends tests on CI engines of passenger car size instead of most commonly observed heavy duty engines. The main contributions are wide range of tested modes of engine based on WHSC, usage of contemporary common rail injection system and measurement of PN and PM emissions in addition to opacity and smoke measurement. 2 EXPERIMENTAL SETUP 2.1 TEST ENGINE The experiments were conducted on a 0.5L single cylinder research engine (AVL 5402) with common rail diesel injection, which parameters correspond to contemporary engines of passenger cars. Opened, rapid prototyping electronic control unit (ECU) allows to adjust the look-up tables of common rail pressure, injection timing and number and quantity of injections. The ECU parameterization was performed using an INCA software. The configuration of the engine testbed allows the control of intake manifold air pressure and exhaust back pressure independently by external compressor with an inlet air conditioning unit and the throttle valve downstream of the damping vessel in the exhaust. The external oil and coolant conditioning unit keeps constant temperature and pressure conditions of measurement. The engine is not equipped with any aftertreatment system. Engine specifications are given in Table

183 Bore x stroke, compression ratio 85 x 90 mm, 16:1 Number of valves 4 Diesel Injection type BOSCH Common Rail, CP 4.1, 1800 bar Diesel Injection Nozzle DLLA 162 P2160, 8 x 0.12 mm x 162 Engine control unit AVL-RPEMS CR Cylinder pressure measurement Cylinder pressure transducer AVL GU22C Piston bowl design Table 1 Engine specifications 2.2 Measurement equipment Arrangement of measurement equipment is introduced in Figure 1. Diesel fuel consumption was measured with the AVL733 dynamic fuel balance. All emissions were measured in the raw exhaust gases and all specific emissions presented in this paper were evaluated from the raw exhaust gas. Smoke number was measured with the AVL 415 smokemeter and opacity with an AVL 439 opacimeter. Particle number (PN) was measured with the AVL 489 Particle Counter connected to the diluter with rotation disks for raw exhaust gases. Gravimetric particulate matter (PM) measurements were obtained using an AVL Smart Sampler 472 with appropriate control and postprocessing software. Figure 1 Single cylinder engine setup and data acquisition schema 182

184 All gaseous emissions were measured by AVL AMA i60 with heated pre-filter. AMA i60 includes measurement of nitrogen oxides using chemical luminescence detector method (CLA), total hydrocarbons and methane emissions via flame ionization detector (FID), and carbon monoxide and carbon dioxide (CO2) via non-dispersive infrared (NDIR) measurement technique. Content of oxygen was measured by a paramagnetic detector (PMD). For in-cylinder pressure measurement and fast acquisition of injector current the AVL INDIMODUL and accessories were used. See Figure 1 for more details. The test cell is equipped with an in-house developed control and lowspeed data acquisition system based on National Instruments cdaq and crio hardware, and a LabVIEW software. Measured data was postprocessed by in-house developed algorithm in Matlab. 2.3 Methodology The aim of the study is to get results based on engine modes, which are close to realusage conditions. The tested modes of engine were defined by WHSC (World Harmonized Stationary Cycle) test [16], [17]. The WHSC test procedure was slightly modified: The stabilization for each tested modes was added and all assessed parameters were measured after this stabilizations. Therefore the influence of the ramp during the change between the modes is suppressed. Hardware configuration of the single cylinder engine testbed is inappropriate for transient modes, which was the main reason for this modification of the test procedure. The full load curve of engine was determined for commonly used speeds and consequently the 12 tested modes were derived. Test starts and stops with the same mode idle (numbers of modes 12 and 13). Each mode has its own weight factor. Full load curve and testing modes are displayed in Figure 2. Numbers of modes are displayed in the circles and diameter of each circle is proportional to weighting factor Wfi. After the fuel change, a 10 min long stabilization run in the mode number 9 was performed before each WHSC test. Figure 2 Selected tested modes of engine and WOT curve 183

185 For each operating point the cylinder pressure traces of 200 consecutive cycles were acquired. All the cylinder pressure derived quantities were evaluated from the average cycle. All emissions were measured in raw exhaust gases and sampled with 30 s average after stabilization. All specific quantities are related to engine indicated power. The appropriate value of common rail (CR) pressure and estimated appropriate manifold air pressure (MAP) was set for each mode (Table 2). The strategy with two injections per cycle, pilot and main, was used for all modes. Exhaust gas recirculation (EGR) was not used for any mode. The temperature of inlet air was kept constant at 30 C, inlet temperature of coolant at 80 C and inlet temperature of oil at 85 C. Point# Speed MAP [bara] CR pressure [bar] IMEP [bar] Wfi [1] Table 2 Test modes definition and weighting factors Change of fuel between measurements of each fuel was carried out as follows. The hoses of fuel system were disconnected and fuel residues were drained or pumped out from these hoses, fuel balance, pumps, and CR injection system. Fuel filters were disassembled and drained. Consequently, the fuel system was flushed and filled with the new fuel. The fuel from fuel balance was consumed in mode 9 of the test and balance were filled again. The stabilization and whole test followed. Three fuels were tested, regular diesel fuel without biofuel (RD), mixture diesel fuel and 30 % HVO (HVO30) and pure HVO (HVO100). Properties of fuels are mentioned in Table 4 in Appendix. 3 RESULTS AND DISCUSION The weighted average of all thirteen modes of test was calculated for each fuel and quantity. The weighting factors are mentioned in Figure 2 and Table 2. Quantities of each fuel are displayed in absolute values as well as in comparison with reference diesel fuel. 3.1 Energetic parameters Indicated power (Pi) and indicated specific heat consumption (ISHC) are displayed in Figure 3. Averaged Pi per cycle is increased proportionally to the growth of the concentration of HVO in regular diesel fuel. Growth for HVO30 (1.41 %) is negligible, 184

186 but Increase by 4.53 % for HVO100 can be considered to be provable. Impact to ISHC seems to be negligible, however slight drop of ISHC looks consistent, while concentration of HVO increases. Similar effect was observed in [7], [11]. Figure 3 Engine indicated power (Pi) and indicated specific heat consumption (ISHC) average per test 3.2 Gaseous emissions Slight drop of indicated specific emission of CO 2 is showed in Figure 4, while content HVO in regular diesel fuel is increased. HVO has higher hydrogen/carbon ratio (H/C) [9], [11], [12], [18], what helps mentioned effect together with growth of Pi mentioned above. Figure 4 Indicated specific CO2 and CO emission average per test Significant drop in emission of CO is showed in Figure 4 and similar drop in hydrocarbon (HC) emission is displayed in Figure 5, while HVO concentration in regular diesel fuel is increased. It is caused by several different properties of HVO fuel in comparison with regular diesel fuel. The lower distillation range improve the fuel 185

187 evaporation and mixing with surrounding air, the higher cetane number provides high reactivity at low combustion temperature and low loads [9] [11]. Figure 5 Indicated specific hydrocarbon (HC) and NOx emission average per test Right side of Figure 5 - shows influence of HVO on emissios of NOx. Higher HVO content in regular diesel fuel fuel shows slight decrease of NOx emission. This effect is observed in [7], [8]. Slight or ambiguous influence of HVO to NOx emissions seems to be positive in comparison with growth of NOx emission with FAME blends [4]. The ambiguous influence to NOX emission can depend on combined effects between ignition delay and fuel injection quantity [11]. Figure 6 Indicated specific particle number (PN) and particulate matter (PM) emission average per test Results for particle number (PN) and particulate matter (PM) are showed in Figure 6. The very significant drop in PM emission and significant drop in PN emission with the pure HVO fuel (HVO100) are achieved. Results achieved with 30% HVO content in regular diesel fuel shows, that significant reduction of PM with lower concentrations of 186

188 HVO is attainable. Effect of HVO30 to PN and PM emission is more than proportional to HVO concentration in comparison with results achieved with HVO100. Figure 7 Opacity and smoke (FSN) average per test Figure 7 shows traces of opacity and filter smoke number (FSN), which represent aproximately proportional decrasee to HVO concentration in regular diesel fuel. Drops in PN, PM, opacity and FSN is positively affected by the absent of aromatic hydrocarbons in HVO, lower distillation range and higher cetane number [9] [11] in comparison with regular diesel fuel. It seems, that benefits of usage HVO in PN, PM, CO and HC emissions and marginal effect to NOx open new possibillities for optimalization of ECU strategy for PM-NOx trade-off. Nevertheless, pressented positive results are attained without any optimization for HVO30 of HVO100 fuel. Furthermore, analysis of emissions of individual modes shows, that HVO content in regular diesel fuel improves CO and HC emissions significantly and improves and NO x at idle and light load modes. It is expected, that HVO content in regular diesel fuel reduce unburned fractions at low combustion temperature [11]. 4 SUMMARY AND CONCLUSIONS Regular diesel fuel without biofuels, blend consists from regular diesel fuel and 30 % hydrotreated vegetable oil (HVO30) addition and pure HVO (HVO100) were tested on single cylinder compression ignition engine of passenger car size equipped with contemporary common rail injection system. Test was based on WHSC test and results were presented as the weighted average per test. Content of HVO in regular diesel fuel slightly increased indicated power (4.53 % for HVO100). Positive, but not statistically significant improvement of engine efficiency was observed. Emission of CO2 was slightly improved (3.56 % for HVO100) with increase of HVO content. Significant drops in CO and HC emissions were achieved (61.8 % and 64.3 %. for HVO100) in comparison with regular diesel fuel. 187

189 NOx emission was slightly decreased. The drop of 4.2 % was achieved with HVO100. Drop of PM emission is very significant for pure HVO, drop of 80.2 % was achieved, whereas decrease of PN, opacity and smoke was about 25% for the same fuel. The lower distillation range, the higher cetane number and zero aromatic content in HVO fuel are probably responsible for the mentioned improvement in smoke, PN, PM, opacity, CO and HC emissions. Usage of 30 % addition of HVO in regular diesel fuel has effect, which is proportional or more than proportional to usage of 100 % HVO. HVO mixtures have positive or neutral effect on all acquired energetic and emission parameters without any changes of ECU settings. Benefits in PM, PN and NOx emissions open a new possibilities for optimization of ECU strategy for NOx-PM trade-off. 5 REFERENCES [1] Heywood, J. B., Internal combustion engine fundamentals, New York: McGrawHill; 1998, ISBN [2] Malhotra, R.K., and Sarin, R., Bio-diesel for Energy Security, Environment Protection and Employment Generation, SAE Technical Paper , [3] McCormick, R.L., Williams, A., Ireland, J., Brimhall, M. and Hayes, R.R., Effects of biodiesel blends on vehicle emissions, Int. J. Eng. Res. 2006: 2:249-61, 2006 [4] Karavalakis, G., Jiang, Y., Yang, J., Durbin, T. et al., Emissions and Fuel Economy Evaluation from Two Current Technology Heavy-Duty Trucks Operated on HVO and FAME Blends, SAE Int. J. Fuels Lubr. 9(1), 2016, doi: / [5] Singer A., et al., Aging studies of biodiesel and HVO and their testing as neat fuel and blends for exhaust emissions in heavy-duty engines and passenger cars, FUEL, vol. 153, pp , 2015 [6] Ohshio, N., Saito, K., Kobayashi, S., and Tanaka, S., Storage Stability of FAME Blended Diesel Fuels, SAE Technical Paper , 2008, doi: / [7] Hartikka, T., Kuronen, M. and Kiiski, U., Technical Performance of HVO ( Hydrotreated Vegetable Oil ) in Diesel Engines, SAE Technical Paper , 2012, doi: / [8] Hannu Aatola, Matti Larmi, Teemu Sarjovaara, and Seppo Mikkonen. Hydrotreated Vegetable Oil as fuel for heavy duty diesel engines, SAE , 2008 [9] Sugiyama, K., Goto, I., Kitano K., and Mogi, K., Effects of Hydrotreated Vegetable Oil ( HVO ) as Renewable Diesel Fuel on Combustion and Exhaust Emissions in Diesel Engine, SAE Technical Paper , 2011 [10] Crepeau, G., Gaillard, P., van der Merwe, D., and Schaberg, P., Engine Impacts and Opportunities of Various Fuels, Including GTL and FAME: Toward Specific Engine Calibration?, SAE Technical Paper , 2009, doi: / [11] Jaroonjitsathian, S., Saisirirat, P., Sivara, K., Tongroon, M. et al., Effects of GTL and HVO Blended Fuels on Combustion and Exhaust Emissions of a CommonRail DI Diesel Technology, SAE Technical Paper , 2014, doi: /

190 [12] F. Millo et al., Experimental Investigation on the Effects on Performance and Emissions of an Automotive Euro 5 Diesel Engine Fuelled with B30 from RME and HVO, SAE Technical Paper , 2013, doi: / [13] N. Nylund, T. Hulkkonen, A. Tilli, S. Mikkonen, P. Saikkonen, and A. Amberla, Emission performance of paraffinic HVO diesel fuel in heavy duty vehicles, SAE Technical Paper , 2011, doi: / [14] Pellegrini, L., Beatrice, C., and Di Blasio, G., Investigation of the Effect of Compression Ratio on the Combustion Behavior and Emission Performance of HVO Blended Diesel Fuels in a Single-Cylinder Light-Duty Diesel Engine, SAE Technical Paper , 2015, doi: / [15] Rodríguez-Fernandez, J., Lapuerta, M., Sanchez-Valdepenas, J., Regeneration of diesel particulate filters : Effect of renewable fuels, Renewable Energy journal vol. 104, pp , 2017 [16] [17] [18] Bhardwaj, O., Lüers, B., Holderbaum, B., Koerfer, T. et al., Utilization of HVO Fuel Properties in a High Efficiency Combustion System: Part 2: Relationship of Soot Characteristics with its Oxidation Behavior in DPF, SAE Int. J. Fuels Lubr. 7(3), 2014, doi: / APPENDIX - TESTED FUELS Results of fuel analysis Start of distillation [ C] 250 C [% v/v] Distilled 350 C [% v/v] until 360 C [% v/v] Temperature of 95 [% v/v] distilled [ C] Total distilled volume [% v/v] End of distillation [ C] Flash point [ C] Cold filter plugging point (CFPP) [ C] Cloud point (CP) [ C] Polyaromatics [% m/m] Fatty acid methyl ester (% v/v) Water Karl Fischer [mg/kg] Sulphur [mg/kg] Ash [% m/m] Impurities before filtration [mg/kg] Conradson Carbon residue (CCT) 10 % [% m/m] Copper corrosion [class] ČSN EN 590:2009 RD Regular Diesel Distillation test < Min > 55 < HVO30 (Diesel + 30% HVO) HVO100 (100% HVO) < < < < 0.1 < < 3.0 < < 6.0 < 6.0 < Class 1 Class 1 Class 1

191 Lubricity HFRR [μm] Kinematic viscosity at 40 C [mm2/s] Oxidation stability Insoluble deposits [g/m3] Oxidation stability Rancimat (110 C) [hours] Oxididation stability PetroOxy [min] Cetane number Cetane index Density 15 C [kg/m3] Heating value, lower [MJ/kg] > Table 4 Test fuel properties 7 ACKNOWLEDGEMENT The study has been performed within the framework of the project of The Ministry of Agriculture of the Czech Republic NAZV QJ This research has been realized using the support of EU Regional Development Fund in OP R&D for Innovations (OP VaVpI) and Ministry for Education, Czech Republic, project # CZ.1.05/2.1.00/ Acquisition of Technology for Vehicle Center of Sustainable Mobility. This support is gratefully acknowledged. This research has been realized using the support of The Ministry of Education, Youth and Sports program NPU I (LO), project # LO1311 Development of Vehicle Centre of Sustainable Mobility. This support is gratefully acknowledged. Authors would like to thank to colleague Ondřej Gotfrýd for support of ECU parametrization. 190

192 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES MĚŘENÍ PASIVNÍCH ODPORŮ PÍSTOVÉHO MOTORU PROTÁČENÍM SE ZVYŠOVÁNÍM TLAKU VE VÁLCI Robert Voženílek1, Stanislav Beroun2, Aleš Dittrich3 Abstrakt Measurement of mechanical losses of the combustion engine by spinning is characterized by the fact that in the cylinder of the spinning engine there is less than half the pressure against the cylinder pressure in the real engine operation. A lower cylinder pressure reduces the piston and crank mechanism load force and the mechanical losses in the engine are significantly reduced. The paper shows the results of measuring the mechanical losses of the combustion engine by spinning with cylinder pressures in the cylinder, which are comparable to cylinder pressures in standard engine operation and explains the engine modification and the required equipment for such measurement. 1. ÚVOD Významným hlediskem při jízdních testech silničních vozidel jsou měrné emise CO 2, které v případě vozidlových pístových spalovacích motorů (PSM) závisí na celkové účinnosti PSM. Celková účinnost je potom určena tepelnou účinnosti pracovního oběhu a mechanickou účinností PSM. Zvyšování tepelné účinnosti PSM a snižování mechanických ztrát v motoru je proto při výzkumu a vývoji motorů předmětem trvalého úsilí. Často používaným postupem pro zjišťování účinku různých opatření ke snižování mechanických ztrát v PSM je měření pasivních odporů motoru protáčením na zkušebním motorovém stanovišti s dynamometrem nebo na speciální zkušební stanici. V obou případech je potřeba vybavení zkušebního zařízení kvalitním měřením točivého momentu, stabilizací teplot chladicí kapaliny i mazacího oleje v motoru a automatizovaným záznamem i vyhodnocováním sledovaných veličin 1. Nedostatkem při měření mechanických ztrát v motoru jeho protáčením je však velký rozdíl mezi tlakem ve válci při protáčení motoru a tlakem ve válci při provozu motoru (tj. se Ing. Robert Voženílek, Ph.D., Technická univerzita v Liberci, robert.vozenilek@tul.cz prof. Ing. Stanislav Beroun, CSc., Technická univerzita v Liberci, stanislav.beroun@tul.cz 3 Ing. Aleš Dittrich, Technická univerzita v Liberci, ales.dittrich@tul.cz

193 spalováním směsi) a tím se při protáčení motoru významně snižují mechanické ztráty na pístové skupině a na klikovém mechanizmu. Jiné způsoby zjišťování mechanických ztrát v motoru (např. výpočtem středního efektivního tlaku a středního indikovaného tlaku pracovního oběhu motoru z termodynamické analýzy vysokotlaké indikace motoru nebo extrapolací průběhu spotřeby paliva v zatěžovacích charakteristikách) poskytují při správně provedených měřeních spolehlivé údaje o celkových mechanických ztrátách v motoru, pro hledání nových technických řešení na snižování mechanických ztrát v PSM jsou však taková měření časově i provozně nákladná. Vzhledem k tomu, že největší podíl (cca 40 až 50 %) na mechanických ztrátách v PSM mají ztráty třením na pístové skupině (píst + těsnící pístní kroužky), je výzkumu vlivu konstrukčního i materiálového provedení pístové skupiny na mechanické ztráty v PSM věnována velká pozornost. Výzkum je založen jak na teoretickém vyšetřování a modelových výpočtech mechanizmu tření pístních kroužků na stěně válce, tak na experimentech. Mechanizmus třecích ztrát na pístních kroužcích souvisí s přítlakem pístních kroužků na stěnu válce, který je určen jak montážním předpětím pístních kroužků, tak změnou přítlaku tlakem ve válci motoru. Pohyb a zatížení pístního kroužku jsou v průběhu zdvihu pístu ve válci motoru velmi proměnlivé: potřebné vůle v uložení pístu ve válci a pístních kroužků v drážkách pístu umožňují průnik tlaku z válce do drážek pro pístní kroužky a významně tak zvyšují přítlak pístních kroužků na stěnu válce. Proměnlivost zatížení pístních kroužků tlakem v drážce jednotlivých kroužků pro případ přeplňovaného vznětového motoru ukazují grafy na obr.1, které byly zjištěny výpočtovým modelováním 2. Obr.1: Vypočtené průběhy tlaku v drážkách jednotlivých pístních kroužků a jejich třecí síly v závislosti na poloze klikového hřídele přeplňovaného vznětového motoru (tj. na tlaku ve válci motoru) podle 2. Výše uvedené skutečnosti, vlastní zkušenosti s experimentálním vyšetřováním mechanických ztrát v PSM na speciální zkušební stanici k protáčení motoru a snaha potlačit dosavadní nedostatek měření mechanických ztrát v PSM protáčením vedly k návrhu nového způsobu měření pasivních odporů PSM protáčením na zkušební stanici v laboratoři pohonných jednotek na TU v Liberci. Podstatu nového způsobu měření na zkušební stanici KVM a výsledek takové měření ukazují následující odstavce. 192

194 2. PROTÁČENÍ PSM SE ZVÝŠENÝM TLAKEM VE VÁLCI Pro nový způsob měření pasivních odporů pístového motoru protáčením na zkušební stanici v laboratoři katedry vozidel a motorů TU v Liberci má motor na všech válcích zavřené ventily (vypojené ovládání ventilů). Místo zapalovacích svíček jsou v hlavě válců upevněny multifunkční šrouby se samočinnými jednocestnými ventily, tlakovými snímači a dekompresními šrouby. Nákres a foto na obr.2. ukazují uspořádání válcové jednotky motoru s multifunkčním šroubem a partii hlavy válců s přípojkami tlakového vzduchu k jednocestným ventilům. Obr.2: Schéma multifunkčního šroubu (5) se speciálním samočinným jednocestným ventilem (7), tlakovým snímačem (6) a dekompresním šroubem (8). Pohled na horní partii hlavy válců se zamontovanými multifukčními šrouby s připojením jednocestných ventilů k přívodu (71) tlakového vzduchu s nastavováním potřebného tlaku vzduchu v přívodu regulátorem (9) 4. Při protáčení 4dobého motoru s trvale zavřenými ventily ve válci je z pracovního cyklu odstraněna výměna obsahu válce a kompresní i expanzní fáze probíhají s dvojnásobnou frekvencí. Požadovaný maximální tlak ve válci se nastaví tlakem v přívodu tlakového vzduchu k jednocestnému samočinnému ventilu v multifunkčních šroubech. Samočinným ventilem je při měření průběžně doplňována náplň ve válci (kompenzace úbytku náplně profukem a poklesu tlaku v důsledku tepelných ztrát) a nastavený tlak ve válci je tím automaticky udržován na požadované hodnotě (obr.3). 4,5 4 3,5 p [MPa] 3 2,5 2 1,5 fil 1 l 0, [deg] Obr.3: Vypočtený průběh tlaku ve válci motoru s počátečním tlakem 200 kpa a zavřenými ventily. K doplňování vzduchové náplně válce samočinným jednocestným ventilem je využitý úsek pootočení klikového hřídele kolem DÚ ( KH). 193

195 2.1 Výsledky ověřovacího měření Ověřovací měření byla provedena na speciální zkušební stanici KVM pro měření pasivních odporů protáčením 4válcového zážehového motoru (Vz = 1,6 dm3), na kterém byly vyřazen mechanizmus ovládání ventilů a v hlavě válců byla místo zapalovacích svíček plně funkční multifunkční šroubení (s tlakovým snímačem pouze v jednom válci). Vzhledem k přestavbě zkušební stanice se měření uskutečnila protáčením neprohřátého motoru v relativně nízkých otáčkách. Výsledek ověřovacích zkoušek nového způsobu měření pasivních odporů motoru protáčením prokázal plnou funkční schopnost samočinného jednocestného ventilu pro nastavování i udržování požadovaného průběhu tlaku ve válci motoru při jeho protáčení. K dosažení maximálních (kompresních) tlaků při protáčení, které budou na úrovni spalovacích tlaků v zážehovém motoru lze odhadnout velikost tlaku vzduchu v přívodu k samočinným jednocestným ventilům do cca 10 bar (se vzrůstem otáček protáčení motoru se začnou projevovat i dynamické vlastnosti samočinného ventilu). Změřené průběhy tlaku ve válci při různých úrovních naplnění válců vzduchem ukazují obr. 4 až 6, ve kterých jsou vyznačeny i hodnoty maximálních tlaků ve válci v průběhu 100 otáček ( cyklů ) motoru. Zjištěnou závislost ztrátového točivého momentu motoru na maximálním tlaku ve válci ukazuje graf na obr.7. Změřený ztrátový točivý moment motoru vykazuje prakticky lineární závislost na maximálním tlaku ve válci, zahrnuje ztráty třením v motoru a negativní práci cyklu účinkem rozdílů v průbězích kompresního a expanzního tlaku a jeho velikost je určena dvojnásobnou frekvencí komprese a expanze proti skutečnému 4dobému motoru. hodnoty pmax ve válci v průběhu měření 100 otáček Obr. 4: Průběh tlaku ve válci při vstupním tlaku vzduchu k jednocestnému ventilu 2 bar. 194

196 Obr. 5: Průběh tlaku ve válci při vstupním tlaku vzduchu k jednocestnému ventilu 3 bar. Obr. 6: Průběh tlaku ve válci při vstupním tlaku vzduchu k jednocestnému ventilu 4 bar. Výsledky měření pasivních odporů motoru protáčením tímto novým způsobem nemohou být přímo porovnávány s výsledky zdánlivě podobných měření podle současných metodik. Podle dat z měření [3] pasivních odporů fyzicky stejného motoru protáčením obvykle používanou metodikou na zkušební stanici KVM byl v režimu nízkých otáček (n = /min) neprohřátého motoru (t 35 0C) kompresní tlak ve 195

197 válci pmax = 19 bar a ztrátový točivý moment Mt/ztr = 14,5 Nm. Z ověřovacího měření pasivních odporů novým postupem je zjištěný ztrátový moment neprohřátého motoru pro stejný kompresní tlak ve velikosti Mt/ztr 23 Nm. Závislost ztrátového točivého momentu na maximálním tlaku ve válci motoru Mt/ztr [Nm] n = 990 1/min pm ax [bar] Obr.7: Protáčení motoru s různě vysokým tlakem (počátečním nebo maximálním) ve válci poskytuje možnost zjistit proměnlivost třecích ztrát pístové skupiny (včetně ztrát čepového tření ložisek v klikovém mechanizmu) v závislosti na tlaku ve válci motoru. Příčiny rozdílu jsou určeny podstatnými změnami podmínek při měření současným a novým způsobem: komprese a expanze ve válcích jsou v uzavřených objemech bez výměny obsahu válců a frekvence cyklů je pro měřený otáčkový režim u nového způsobu měření dvojnásobná proti současnému způsobu. K porovnání a vysvětlení rozdílu výsledků obou způsobů měření lze využít hodnoty z grafu na obr.7: - Výfukový a sací zdvih probíhá prakticky s konstantním tlakem ve válci p 1 bar. Extrapolací závislosti v grafu na obr.7 lze pro tento tlak ve válci odhadnout Mt/ztr 6 Nm. - Zahrnutím vlivu frekvence a střídání p max = 1 bar a 19 bar v navazujících otáčkách, potom střední hodnota Mt/ztr = (6 + 23)/2 = 14,5 Nm. Výsledek se až překvapivě shoduje s dřívějším měřením [3]. Zvýšené tlaky ve válci motoru zvyšují i termodynamické ztráty přestupem tepla z náplně do stěn ve válcové jednotce a zvyšují se i ztráty netěsností (profuky). Výpočtovým zpracováním změřených průběhů tlaku pomocí SW AVL Concerto 4.6 byla zjištěna velikost středních indikovaných ztrátových tlaků p_ ztr(ind) při protáčení v závislosti na tlaku ve válci: výsledek ukazuje graf na obr.8. Vypočtené hodnoty středních indikovaných ztrátových tlaků p_ ztr(ind) při měření motoru novým způsobem protáčení byly stanoveny pro jednu otáčku motoru, aby číselné hodnoty p_ztr(ind) více korespondovaly se ztrátami při kompresi a expanzi 4dobého motoru (tj. bez fáze výměny obsahu válce). Na obr.8 je zakreslen i střední ztrátový tlak p_ztr(mtztr), vypočtený ze změřeného ztrátového točivého momentu: i v tomto případě jsou hodnoty p_ztr(mtztr) vztaženy na jednu otáčku motoru. Podíl tepelných ztrát a netěsností ve válci na celkovém ztrátovém momentu při protáčení v závislosti na tlaku ve válci zpočátku exponenciálně vzrůstá, průběh však naznačuje i možný efekt vyššího tlaku ve válci na zvětšení přítlaku pístních kroužků s omezováním profuků (ztráty p_ztr(ind) na celkovém ztrátovém tlaku p_ztr(mtztr) pro vyšší tlaky vykazuje spíše lineární závislost na tlaku ve válci). 196

198 Střední tlak ztrát v motoru při protáčení 1,6 0,1 1,4 0,09 1,2 0,08 1,0 0,07 0,8 0,06 0,6 0,05 0,4 0,04 0,2 0,03 0,0 0, pmax [bar] p_ztr(ind) / p_ztr(mtztr) p_ztr(ind) / p_ztr(m(tztr) [bar] p_ztr(mtztr) pztr p_ztr(ind) 45 Obr.8: Závislosti středního indikovaného ztrátového tlaku p_ztr(ind), celkového středního tlaku ztrát p_ztr(mtztr) a podílu těchto ztrátových tlaků na tlaku náplně ve válci při protáčení motoru. Způsob měření pasivních odporů pístového motoru protáčením při zvýšených tlacích ve válci spolu se zařízením k provádění tohoto způsobu byl po úspěšném ověření přihlášen na Úřadu průmyslového vlastnictví v Praze k patentovému řízení ZÁVĚR Přes malý rozsah ověřování nového způsobu vyšetřování pasivních odporů pístového motoru protáčením zjištěné poznatky prokazují, že měření pasivních odporů motoru protáčením se zvyšováním tlaku náplně ve válci poskytuje významné informace o vlastnostech a chování pístních kroužků v podmínkách blízkých reálnému provozu motoru. Nový způsob měření pasivních odporů motoru protáčením se tak může stát účinným nástrojem při výzkumu (konstrukčním i technologickém) pístních kroužků. Po dokončení přestavby zkušební stanice KVM (nové silové, měřicí a ovládací elektrovybavení stanice, snímač točivého momentu s větším rozsahem, měření profuků) k měření pasivních odporů motoru protáčením budou provedena další měření s podrobnější analýzou výsledků většího souboru dat. V článku je popsán výzkum, na kterém pracuje KVM v rámci pracovního balíčku WP08 projektu TE Centrum kompetence automobilového průmyslu Josefa Božka. 197

199 REFERENCES [1] [2] [3] [4] BEROUN S., VOŽENÍLEK, R.: Měření pasivních odporů motoru (kompletního motoru i jeho vybraných skupin) protáčením. Výzkumná zpráva SM 617/2009. KVM TU v Liberci, NOVOTNÝ P., PÍŠTĚK V., DRÁPAL L.: Modelling of Piston Ring Pack Dynamics. MECCA IX/02, ČVUT v Praze, 2011, pp ISSN ZVOLSKÝ T.: The Valve Train Mechanical Losses Measurement. XLVI. International Scientific Conference of the Czech and Slovak Universities and Institutions dealing with Research of Internal Combustion Engines. STU in Bratislava, Slovak Republic. Kočovce ISBN VOŽENÍLEK R., BEROUN S.: Způsob měření mechanických ztrát pístového motoru jeho protáčením a zařízení k provádění tohoto způsobu. PV , ÚPV Praha,

200 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES WAYS OF RECHARGING ELECTRIC BUSES AT STOPS Josef Morkus1 Abstract The basic problem of battery powered electric buses is the range. If the vehicle is equipped with all-day batteries, due to their weight, the permitted axle pressures can be exceeded and the number of passengers has to be reduced. One possible solution is to use smaller batteries and to charge them at the final and/or intermediate stops. The article deals with various charging systems using pantographs, charging heads, inductive charging and so called partial electric buses, their advantages and problems. 1. INTRODUCTION At present, many cities are trying to introduce environment-friendly urban public transport and the often idea is to use purely electric buses. The main problems that hinder the mass introduction of these vehicles are their high prices, short range and limited passenger capacity due to high battery weight and permissible axle load. For example, a Chinese electric bus BYD of 12 meters length has a battery of 324 kwh, range about 250 km, but the capacity of only 54 passengers [1]. Such a vehicle can be suitable for low-load lines only. If it is used on the line where the number of passengers corresponds to the use of a standard bus with a capacity of 90 people, the transport company has to buy almost double number of vehicles. Otherwise, if it is used for example an electric bus SOR EBN 10.5 with an appropriate capacity of 85 passengers, but with range of approximately 120 km (battery 172 kwh) [3], the transport company must provide additional vehicles for the second half of the day and the result is similar to the previous case. A car with low range is suitable for loaded short lines or for operation in peak hours only, when is a possibility to recharge batteries in meantime. What are the options to ensure all-day operation of electric vehicles? If we omit the use of hybrid or plug-in hybrid buses, then may be considered partial trolleybuses or electrobuses recharged on final or intermediate stops. Here it is advisable to add that if fast recharging at stops is used, an additional slow night charging is also Ing. Josef Morkus, CSc., Ústav automobilů, spalovacích motorů a kolejových vozidel, Fakulta strojní ČVUT v Praze,Technická 4, Praha 6, josef.morkus@fs.cvut.cz 1 199

201 necessary to recharge the battery at full capacity and to balance the individual battery cells. Figure 1: Electric bus BYD - battery location [2] 2. PARTIAL TROLLEYBUSES A partial trolleybus (sometimes also for business reasons called a partial electric bus) is a hybrid trolleybus with batteries that allow ride on a part of the line without a trolley. Max. length of part without trolley is usually 1/3 of total line length. The batteries are recharged while the vehicle is driving or standing in stops under trolley and by energy regeneration when braking. The term "partial trolleybus" means a vehicle which allows standard operation with passengers in parts of the line without a trolley, not only to detour slowly some obstacles on a track etc. Partial trolleybuses are currently operated in more cities. An example is the Solaris Trolino 18 with electrical equipment Cegelec in town Eberswalde [4], where it is operated on a line 15 km under the trolley and 5 km without the trolley. The 70 kwh battery in this vehicle weighs 1020kg, i.e. approximately the same as the Diesel drive, and thus the number of passengers is not limited. In addition, a 0.88 kwh supercapacitor is used to improve the energy regeneration. Figure 2: Partial trolleybus Solaris Trolino 18 in Eberswalde 200

202 If a partial trolleybus is operated on longer sections without trolley such as Solaris Trolino 12 with Škoda electrical equipment in the Swedish town Landskrone [5], it is necessary to use the appropriate battery size and reduce the number of passengers. Partial trolleybuses use electrical energy from the trolley preferably for the traction electric motor, the batteries are recharged only by surplus trolley energy and by regeneration. The SwissTrolley plus vehicle from Hess tested in Zurich [6] is solved in reverse: The energy for the traction electric motor is taken exclusively from the battery, and the current from the trolley serves only to recharge the batteries. There is also the possibility of energy flow from the battery to the trolley and thus balancing the peaks in the net. From the technical point of view, therefore, this vehicle is a partial electric bus. Figure 3: Partial electric bus SwissTrolley plus Partial trolleybuses/electric buses are suitable for cities where the trolleybus infrastructure is already built. These vehicles can be used to extend the trolleybus lines without the need for the construction of another overhead contact line. It is also possible to cancel most expensive parts of the overhead trolley network during its maintenance or reconstruction e.g. crossroads with trams lines, turnouts etc. - and vehicles will pass these sections using batteries. 3. ELECTROBUSES WITH CHARGING AT STOPS In cities where there is no trolleybus network or in terms of substitution a purely bus line, there is an effort to use electric buses with the possibility of recharging at terminal stations or at intermediate stops. 3.1 Charging using split pantograph The easiest and cheapest solution is recharging from a trolleybus overhead network (it can be used or built a short section only, for example at the final station) using a split pantograph on the roof of the vehicle. In this case it is advantageous to connect charging points (with voltage stabilization) to the tramway or trolleybus network, if any. However, the charging current must be low due to heating the pantograph and the trolley contact points when the vehicle is stationary. Therefore, the timetable must allow enough time to charge even if the vehicle is delayed on the line. Because such vehicle is not grounded, the electrical equipment must be double-insulated. An example is the small 7.7 m long Siemens Rampini ALE EL electric bus [7] which has the charging 201

203 current about 65 A. This vehicle is operated on short inter-city lines in Vienna. It has also been tested in several Czech cities, including Prague where the test on the more demanding line 216 in the winter conditions ended unsuccessfully. In Prague on lines 203 and 163 is from 2015 tested electrobus SOR EBN 11 [8] with a capacity of 93 passengers (battery 172 kwh, range without recharging about 120 km, electric heater 20 kw) recharged using a split pantograph from short trolleybus overhead network at one terminal station. Rechargeable currents are 150 A in summer and 200 A in winter and recharging takes place every second ride on the line. After the next ride, the stall time can be used for recharging if there is a delay on the line or if energy consumption for heating is high. The test results are positive and Prague is preparing to introduce a regular transport by electric buses on line 207. Figure 4: Electric bus SOR EBN 11 at terminal station A major problem in the conditions of the Czech Republic is heating in the winter months, which increases the energy consumption up to double and decreases the range proportionally. Therefore, some other designs use diesel heating, but such vehicles are not pure electric. 3.2 Charging using charging head To achieve higher charging currents and faster recharging, different types of charging head systems on the vehicle and charging stand at stops are used. This is not a whole new idea, recharging stands were already used by the so-called gyrobuses in Switzerland in the 1950s [9]. Figure 5: Charging arm on the roof of Solaris Urbino 12 Electric connected to charging rail 202

204 An example of a modern solution can be the Solaris Urbino 12 Electric bus with 5 poles head operated on a short 5.2 km line in Dresden [10]. The charging stand is connected to a tram network and is equipped with a recharging rail to which the electric bus is connected by means of an automatically guided lifting arm on the roof. The arm carries a charging head, the charging current is 270 A (max up to 1000 A) at 748 V and charge time is 3 minutes. Higher charging currents of about 1000 A use the ABB TOSA system [11] operated on line 5 in Geneva and newly built on line 28 across the city. The system called flash charging uses a 4-pole rechargeable head with laser guidance on the vehicle roof. Stands with recharging rail (groove) are at selected stops and have supercapacitors that allow rapid energy transfer to the vehicle in seconds. Next advantage is the possibility to connect the stand to standard 3-phase network. Figure 5: Electric bus Hess with charging system ABB TOSA at a stop with charging stand A similar system uses electrobus Škoda 26 BB HP Perun, tested in Pilsen [12]. The Škoda charging stand is equipped with an automatically-guided arm. In contrast to previous systems, the recharging head is located on the arm and the charging rails are on the roof of the vehicle. LTO batteries (Li-ion battery with titanium dioxide on the anode) are used in the vehicle to allow high charging currents up to 1000 A. The charging stand located at the end station is connected to a high voltage network and the charging time is 5-8 minutes. Figure 6: Electric bus Škoda 26 BB HP Perun at Škoda charging stand 203

205 Somewhat different way is used by Volvo, the leading manufacturer of hybrid buses. The bus Volvo 7900 Electric Hybrid [13] is a combination of an electric bus and a hybrid bus. These vehicles are operated in several cities (Stockholm, Hamburg, Luxembourg...). The Volvo system uses recharging stands at the end station. The component of the stand is an automatically guided pantograph with the charging head triggered to the vehicle from above and the charging rails are on the roof of the vehicle, like at the Škoda system. The advantage is less weight on the vehicle and fewer number of pantographs. The charging current is 200 A, charging time 6 minutes and a range for one charge is 7 km, which is sufficient to pass through the city center. Out of the city center the vehicle can continue with hybrid propulsion (Diesel + battery) without any distance restrictions. Figure 7: Volvo 7900 Electric Hybrid at charging stand [14] The basic problem of recharging systems with charging head are the high costs of the rechargeable infrastructure and its connection to public electricity nets. The use of high recharging currents allows shorter recharging times, but its influence on battery life is not yet sufficiently known. Individual systems are still being tested and developed. A pantograph system with an auto-guided 4-pole charging head and charging rails on the roof of the vehicle have been agreed as a standard a between major manufacturers of electric buses and major infrastructure manufacturers [15]. 3.3 Inductive charging An alternative to roof charging is inductive charging using non-contact transmission of energy between the transmission coil under the road surface and the receiving coil at the bottom of the vehicle. Testing operation is again in more cities. For example, Solaris Urbino 12 Electric electrobuses with electric heating 20 kw and capacity 70 passengers have been tested with Bombardier PRIMOVE induction charging on a 6.1 km long route in Berlin [16]. Recharging stations with a power of 200 kw were located at both terminals, the charging time was 4-7 minutes. However, operational reliability was low, about 40% [17]. A similar system is tested by Scania for plug-in hybrid buses [18]. 204

206 Also in London on line 69 are tested 4 double-deckers Alexander Dennis Enviro 400H [19]. These are buses with a BAE hybrid drive, equipped with an induction system to charge the batteries, allowing most of the route to go in pure electric mode. A GPS is used to navigate buses to a charging plate, the recharging power is 100 kw and charging time is 10 minutes. Figure 8: Alexander Dennis Enviro 400H hybrid bus and inductive charging plate The general problems of inductive charging are costs, efficiency and limited power, 4. CONCLUSION Electric vehicles with once a day charging ("night charging") are suitable for short and less loaded lines. If there is no significant development in the field of battery energy density, charging at stops ("opportunity charging") is the only way to ensure all day operation of electric vehicles, while retaining sufficient passenger capacity in the vehicle. The high cost of charging infrastructure can be reduced by using of hybrid vehicles with a pure electric mode of driving only in the city center. In cities where the trolleybus infrastructure is built, partial trolleybuses/electric buses is also a perspective solution. When designing an optimal transport system, it is necessary to consider all the loads and choose the most appropriate combination of systems for given conditions. 5. REFERENCES [1] [2] [3] [4] [5] [6] [7] L. Hinčica: Elektrobus BYD ebus na zkouškách v Ostravě Československý Dopravák 3/2014 J.Slavík: Elektrobus BYD v Ostravě L. Hinčica: Elektrobus SOR 10,5 v Jeseníkách Československý Dopravák 2/2015 Z. Vytouš, L.Hinčica: Parciální elektrobus řešení pro elektromobilitu Československý Dopravák 3/2014 P.Gabriel: Autobusy a trolejbusy Solaris ve Švédsku v r Československý Dopravák 1/2014 L. Hinčica: SwissTrolley plus nový parciální trolejbus pro Zürich Československý Dopravák 1/2017 L. Hinčica, J. Černý, L. Podivín, R. Filip, J.Šurovský: Elektrobus Rampini ALE EL na testech v České republice 205

207 [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] Československý Dopravák 1/2014 L. Hinčica: Elektrobus SOR EBN 11 s vrchním nabíjením na zkouškách v Praze. Československý Dopravák 3/2015 Gyrobus L. Hinčica: (Nejen) Drážďanský elektrobus Československý Dopravák 4/2015 B. Warner: TOSA e/bus enables breakthrough environmentally friendly commuting in Geneva L. Hinčica: Škoda 26 BB HP Perun elektrobus s rychlonabíjením Československý Dopravák 2/2015 L. Hinčica: Volvo 7900 Electric Hybrid Československý Dopravák 4/2014 Volvo 7900 Electric Hybrid Firemní prospekt Volvo BusPress O. Pavlůsek: Elektrobusy Solaris a indukční dobíjení v Berlíně BusPress ACKNOWLEDGEMENT This research has been realized using the support of The Ministry of Education, Youth and Sports program NPU I (LO), project # LO1311 Development of Vehicle Centre of Sustainable Mobility. This support is gratefully acknowledged. 206

208 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES ZEMNÍ PLYN JAKO AUTOMOBILOVÉ PALIVO Josef Laurin1 Abstrakt Článek uvádí palivářské a ekonomické vlastnosti zemního plynu v porovnání s vlastnostmi automobilového benzinu a motorové nafty. Popisuje některé konstrukční a provozní parametry, zejména ekologické, motorů a vozidel na zemní plyn. Poukazuje na probémy spojené s provozem vozidel na zemní plyn a naznačuje očekávaný vývoj budoucího využití automobilového paliva zemní plyn v České republice. 1. ÚVOD Podle Evropské strategie pro nízkoemisní mobilitu [3] produkuje odvětví dopravy téměř jednu čtvrtinu evropských emisí skleníkových plynů a je hlavním zdrojem znečištění ovzduší ve městech. Ke snížení výfukových emisí může přispět mj. také přechod z klasických automobilových paliv motorové nafty a automobilového benzinu na některá alternativní paliva, např. na zemní plyn. Širší využití zemního plynu jako automobilového paliva v ČR je v souladu s Národním akčním plánem čisté mobility [8] (dále jen NAP ČM) i se Státní energetickou koncepcí ČR [9] a má podporu ve Směrnici Evropského parlamentu a Rady o zavádění infrastruktury pro alternativní paliva [6]. Zemní plyn může být uložen na vozidlech buď jako stlačený (CNG = Compressed Natural Gas), nebo zkapalněný (LNG = Liquefied Natural Gas). Palivářské parametry zemního plynu, benzinu a nafty a jejich ceny uvádí tabulka PALIVO ZEMNÍ PLYN Hlavní složkou zemního plynu je metan, jeho obsah v plynu z různých míst těžby je v rozmezí od 80 do 99 %. Do ČR je přiváděn ruský plyn s obsahem přibližně 97 až doc. Ing. Josef Laurin, CSc, Technická univerzita v Liberci, Studentská 2, Liberec, josef.laurin@tul.cz 1 207

209 98 % obj. metanu, norský plyn obsahuje přibližně 85 % obj. metanu. Dalšími složkami zemního plynu jsou vyšší uhlovodíky, dusík a oxid uhličitý a stopově i jiné plyny. Hustota kapaliny při 15OC (kg.m-3) Zemní plyn 415 Hustota plynu při 15 OC (kg.m-3) 0,69 Výhřevnost (MJ. kg-1) 49,4 42,7 42,5 Výhřevnost (kwh. kg-1) 13,72 11,86 11,80 Výhřevnost 9,46 kwh. m-3 8,83 kwh. dm-3 9,7 kwh. dm-3 Oktanové číslo - výzkum. metoda Cena paliva 24,7 Kč.kg-1 30 Kč.dm-3 29 Kč.dm-3 Cena paliva (Kč.kg-1) 24,7 40,27 35,27 Cena 1 kwh energie (Kč) 1,80 3,39 2,99 Spotřební daň 2,0 Kč.kg-1 12,84 Kč.dm-3 10,95 Kč. dm-3 Spotřební daň (Kč. kg-1) 2,0 17,23 13,32 Spotřební daň za 1 kwh energie (Kč) 0,15 1,45 1,13 Parametr/palivo Benzin BA Motorová nafta 822 Tabulka 1: Průměrné parametry zemního plynu, automobilového benzinu a motorové nafty. Ceny paliv u čerpacích stanic v ČR jsou z června 2017 Zemní plyn je vhodným palivem zejména pro zážehové motory. Jak je patrno z tabulky 1, má zemní plyn vysokou výhřevnost a vysokou odolnost proti klepání oktanové číslo měřené výzkumnou metodou přibližně 125. Při jeho spalování vzniká méně skleníkových plynů než při spalování benzinu nebo nafty. Stechiometrickým spálením 1 kg benzinu vzniká 3,13 kg oxidu uhličitého, spálením 1 kg nafty vzniká 3,21 kg oxidu uhličitého a spálením 1 kg zemního plynu pouze 2,75 kg oxidu uhličitého. Přechodem z benzinu na zemní plyn může produkce CO 2 vztažená na dráhovou spotřebu vozidel klesnout až na 76 %, přechodem z nafty až na 89 %. Oxid uhličitý patří mezi skleníkové plyny a podle nařízení Evropského parlamentu a Rady (ES) [5] nemá jeho průměrná specifická emise vztažená na jedno vozidlo z celkové produkce lehkých užitkových vozidel výrobce automobilů překročit v letošním roce hodnotu 130 g/km a v roce g/km. Bílá kniha EU o dopravě [2] uvádí výhledově do roku 2050 hodnotu přibližně 35 g/ km. Vozidlové zážehové motory na zemní plyn spalují zpravidla stechiometrickou palivovou směs a pomocí třísložkového katalyzátoru se snižují plynné výfukové škodliviny oxid uhelnatý, uhlovodíky a oxidy dusíku. Tyto motory produkují v porovnání 208

210 s motory naftovými a benzinovými méně plynných výfukových škodlivin i pevných částic. Zemní plyn může být použit i pro vznětové motory, kdy se pod vysokým tlakem vefukuje přímo do válců motoru a zapaluje se vstřikem malé dávky nafty. Jedná se o vznětový pracovní oběh označovaný HPDI (High Pressure Direct Injection), u kterého se dosahuje vyšší celkové účinnosti motoru než u motoru zážehového. Hlavními přednostmi použití zemního plynu jako automobilového paliva v porovnání s benzinem a naftou jsou z obecného pohledu úspory ropy, téměř nulová tvorba pevných částic, již zmíněná mírně nižší produkce oxidu uhličitého a v důsledku nízké spotřební daně též jeho relativně nízká cena, viz cenové údaje v tabulce 1. Z hlediska vzniku a průběhu požáru při havárii vozidla je zemní plyn v porovnání s benzinem díky svým fyzikálním vlastnostem (zejména vysoká teplota zapálení, nižší měrná hmotnost než vzduch) poměrně bezpečným palivem. Jako nevýhody paliva zemní plyn lze uvést zejména vyšší ceny vozidel s hmotnějšími palivovými nádržemi a v případě CNG kratší dojezd vozidel, v ČR řídkou síť plnicích stanic, delší dobu plnění nádrží a zákaz parkování vozidel v podzemních hromadných garážích a parkovacích domech. Vyšší jsou nároky na servis vozidel. Zemní plyn se vzhledem ke svým palivářským vlastnostem v současnosti jeví jako perspektivní alternativa k benzinu a naftě. Lze očekávat, že ekonomicky těžitelná ložiska ropy, tj. suroviny pro výrobu klasických automobilových paliv nafty a benzinu, budou během tohoto století převážně vyčerpána. Pravděpodobné ekonomicky těžitelné světové zásoby zemního plynu bývají odhadovány na 130 a více let. Dalšími zdroji metanu mohou být hydráty nacházející se pod dny oceánů a plyn získávaný z břidlic. 2.1 CNG CNG je zemní plyn na vozidlech přechovávaný v tlakových nádobách konstruovaných na pracovní tlak obvykle 20 MPa, méně často na vyšší až 30 MPa. Počátky využití zemního plynu pro vozidlové motory na území ČR spadají do doby druhé světové války, kdy byl stlačený zemní plyn používán k pohonu motorových vozidel na jižní Moravě, tam se nacházely čtyři plnicí stanice. Koncem června 2017 byly v ČR 153 veřejné plnicí stanice CNG a v provozu na CNG více než 18 tisíc vozidel, z tohoto počtu přibližně autobusů. Podle směrnice Evropského parlamentu a Rady 2014/94/EU o zavádění infrastruktury pro alternativní paliva by na hlavních evropských komunikacích v zemích EU měly být do roku 2025 vybudovány plnicí stanice tak, aby průměrná vzdálenost mezi nimi byla nejvýše 150 km. Podle NAP ČM by mělo být v roce 2025 přibližně 340 veřejných plnicích stanic CNG. Většina zejména osobních vozidel na CNG vznikla dodatečnou úpravou vozidel s původně benzinovými motory. Vozidla umožňující provoz jak na CNG, tak na benzin vyrábí řada automobilek, k nejvýznamnějším výrobcům osobních automobilů na CNG v Evropě patří Dacia, Fiat, Iveco, Opel, Mercedes-Benz, Seat, Škoda-Auto a Volkswagen. Škoda-Auto od roku 2013 dodává na trh osobní automobil Citigo CNG s tříválcovým motorem, objem válců 1 litr, výkon 50 kw, emise CO 2 79 g.km-1, nádrž na 12 kg CNG a nádrž na 10 litrů benzinu, dojezd 340 km na CNG a 220 km na benzin. Dalším automobilem Škoda-Auto na CNG je Škoda Octavia G-TEC s přeplňovaným 209

211 motorem 1,4 TSI o výkonu 81 kw, emise CO2 97 g.km-1, nádrž na 15 kg CNG a nádrž na 50 litrů benzinu, dojezd na CNG 410 km a 920 km na benzin. Užitkové automobily na CNG a tahače vyrábějí evropské firmy Iveco, Man, MercedesBenz, Renault, Scania a Volvo. Autobusy na CNG vyrábějí evropské firmy Iveco, Man, Mercedes, Scania, Solbus a v ČR firmy SOR Libchavy a Iveco Česká republika. 2.2 LNG LNG je bezbarvá kapalina, která má při atmosférickém tlaku teplotu minus 160 oc a zaujímá zhruba 570krát menší objem než zemní plyn v plynné fázi. To je významná výhoda pro jeho uskladnění na vozidlech v kryogenních nádržích, v nichž se LNG přechovává při teplotě v rozmezí minus 160 až minus 150 oc. Nádrže bývají konstruovány na maximální provozní přetlak do 0,8 MPa a musí mít velmi dobrou tepelnou izolaci. Kvalita izolace je rozhodující pro množství odpařeného plynu, a tedy i pro tzv. dobu zádrže, za kterou vzroste tlak v nádrži na nejvyšší přípustnou hodnotu, při níž pojistný ventil vypustí z nádrže část plynné fáze a tlak poklesne. Doba zádrže bývá delší než týden. V současné době je v ČR pouze jedna plnicí stanice LNG provozovaná firmou Spolgas, s.r.o., v Lounech. Slouží k plnění nádrží tahače návěsů Iveco Stralis NP 400 LNG/CNG, jediného vozidla provozovaného na LNG v ČR. Nejbližší veřejné plnicí stanice LNG se nacházejí v přibližně 320 km vzdáleném Berlíně a zhruba 500 km vzdáleném Ulmu. V Polsku, v Rakousku ani v Maďarsku veřejné plnicí stanice LNG nejsou. V Evropě jezdí na LNG pouze několik set vozidel, převážně těžkých nákladních automobilů nebo autobusů, např. v Holandsku, Polsku, Norsku, Rusku, Španělsku, Švédsku a ve Velké Británii. Problematické je využití LNG jak pro vysokou energetickou náročnost zkapalňování, tak pro skladování za velmi nízkých teplot. Zemní plyn je v některých místech těžby zkapalňován a loděmi dopravován do evropských přijímacích terminálů, odkud je možná jeho doprava k plnicím stanicím ve vnitrozemí zpravidla po silnici nebo po železnici. Podle směrnice Evropského parlamentu a Rady 2014/94/EU o zavádění infrastruktury pro alternativní paliva by na hlavních evropských komunikacích v zemích Evropské unie měly být do roku 2030 vybudovány plnicí stanice tak, aby průměrná vzdálenost mezi nimi byla nejvýše 400 km. Otázkou zůstává efektivita používání LNG pro pohon silničních vozidlech, a tudíž i efektivita případného budování sítě plnicích stanic. NAP ČM předpokládá následující počty veřejných plnicích stanic LNG v ČR: 2 v roce 2020, 5 v roce 2025 a 14 v roce Budoucnost pohonu vozidel zemním plynem ČR Jak již bylo uvedeno, je v současné době významnou výhodou CNG v porovnání s benzinem a naftou jeho nízké zatížení spotřební daní. Spotřební daň připadající na 1 kwh energie je u CNG 0,15 Kč, u benzinu 1,45 Kč a u nafty 1,13 Kč. Nízká spotřební daň je zafixována až do roku 2019 a není známo, jak se bude dále vyvíjet. Očekávání, že dojde ke zvýšení daně, již dnes nepříznivě ovlivňuje zájem o nákup vozidel na CNG. Odhad budoucího počtu vozidel na CNG je v NAP ČM [8] proveden ve čtyřech variantách scénářů, závislých hlavně na chování státu a plynárenských společností. Ideální optimistický scénář předpokládá pro využití CNG jako automobilového paliva mj. zachování nízké spotřební daně, nulovou silniční daň a výraznou finanční 210

212 podporu infrastruktury a nákupu některých vozidel na CNG ze strany státu a dotacemi z Evropské unie. Středně optimistický scénář platí při zachování podpory pro rok 2020 až do doby dosažení 10% podílu spotřeby zemního plynu na celkové spotřebě pohonných hmot (nyní pod 1%) a zachování nulové silniční daně pro vozidla na CNG. Pesimistický scénář počítá se zachováním zvýhodněné spotřební daně na CNG i po roce 2020 pouze na úrovni cca 50% výše spotřební daně pro klasická automobilová paliva s nulovou silniční daní a s podporou nákupu vozidel pro vozidla státní správy a místní samosprávy. Katastrofický scénář by nastal v případě zrušení výše uvedených zvýhodnění. Očekávané počty vozidel na CNG v ČR v letech jsou v tabulce 2. NAP ČM uvádí, že za podmínek ideálního optimistického scénáře by v ČR mohlo být v roce 2030 až vozidel na LNG. Roky Ideální optimistický scénář Středně optimistický scénář Pesimistický scénář Katastrofický scénář Tabulka 2: Očekávané počty vozidel na CNG (v tisících) v ČR v letech Vývoj motorů na zemní plyn na Technické univerzitě v Liberci Na katedře vozidel a motorů byl v minulosti proveden mj. vývoj řady plynových zážehových motorů na zemní plyn, při němž se vycházelo z konstrukcí motorů původně naftových i benzinových. Výsledkem byly 24 typy plynových zážehových motorů. V 19 případech se vycházelo z motorů původně naftových a v 5 případech z motorů benzinových. Přestavba naftových motorů na plynové zážehové spočívala zejména v úpravě spalovacího prostoru v pístu vedoucí ke snížení kompresního poměru, v náhradě naftového palivového příslušenství plynovým, ve vybavení motoru elektrickým zapalováním a v instalaci elektronického řízení motoru a katalytického systému, ve většině případů s třísložkovým katalyzátorem. Pro vozidlové využití (autobusy, nákladní automobily, osobní automobily, terénní vozidla, dopravní vozíky) bylo vyvinuto 18 typů plynových motorů. Plynové motory určené pro autobusy a nákladní automobily byly úspěšně homologovány z hlediska plynných výfukových emisí podle předpisu EHK 49 a s velkou rezervou vyhověly limitům stanoveným pro EEVs (enhanced environmentally friendly vehicles). Průmyslových plynových motorů pro elektrická zdrojová soustrojí, kogenerační jednotky a kompresory tepelných čerpadel bylo vyvinuto 6 typů. 3. ZÁVĚR V porovnání s klasickými automobilovými palivy benzinem a motorovou naftou je zemní plyn relativně ekologickým palivem. Zdroje zemního plynu mají podstatně delší životnost než zdroje ropy a je pravděpodobné, že bude docházet ke snižování objemu výroby a zvyšování cen automobilových paliv ropného původu. Pohony automobilů pak budou orientovány především na motory spalující zemní plyn. S budoucím rozvojem 211

213 uplatnění zemního plynu jako automobilového paliva v ČR počítá Státní energetická koncepce [9] i Národní akční plán čisté mobility [8], ale skutečnost bude záviset především na nepředvídatelných politických rozhodnutích státu. LITERATURA [1] Český plynárenský svaz [online]. Praha. Dostupné na: a na [2] Evropská komise: Bílá kniha - Plán jednotného evropského dopravního prostoru vytvoření konkurenceschopného dopravního systému účinně využívajícího zdroje. Brusel [3] Evropská komise: Evropská strategie pro nízkoemisní mobilitu (Sdělení komise Evropského parlamentu). Brusel [4] Evropská hospodářská komise: Předpis EHK č. 83 jednotná ustanovení pro schvalování vozidel z hlediska emisí znečišťujících látek podle požadavků na motorové palivo. [5] Evropský parlament a Rada (ES): Nařízení Evropského parlamentu a Rady (ES) č. 443/2009 ze dne 23. dubna 2009, kterým se stanoví výkonnostní emisní normy pro nové osobní automobily v rámci integrovaného přístupu Společenství ke snižování emisí CO2 z lehkých užitkových vozidel. Brusel [6] Evropský parlament a rada: Směrnice Evropského parlamentu a Rady 2003/30/ES ze dne 8. května 2003 o podpoře užívání biopaliv nebo jiných obnovitelných pohonných hmot v dopravě. Brusel [7] Evropský parlament a rada EU: Směrnice evropského parlamentu a rady 2014/94/EU ze dne 22. října 2014 o zavádění infrastruktury pro alternativní paliva. Brusel [8] Ministerstvo průmyslu a obchodu ČR: Národní akční plán čisté mobility pro období s výhledem do roku Praha [9] Ministerstvo průmyslu a obchodu: Státní energetická koncepce (aktualizovaná v květnu 2015). Praha Příspěvek byl napsán na katedře vozidel a motorů Technické univerzity v Liberci za institucionální podpory Ministerstva školství, mládeže a tělovýchovy v roce

214 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES EVALUATION OF THE IMPACT OF DIESEL FUEL ADDITIVE ON SELECTED COMBUSTION ENGINE PARAMETERS Ivan Janoško1, Juraj Jablonický2, Ľubomír Hujo3, Peter Kuchar4 Abstract The paper deals with the impact assessment of the additive used to add diesel fuel to the power and emission parameters of the vehicle and its consumption. Measurements were performed under laboratory conditions on the MAHA MSR 500 test bench where we performed driving cycles with pre-selected external conditions and time intervals. Comparison of the resultant values of the driving cycles measured before and after the application of the additives to the fuel was compared. We focused on tracking the vehicle's external revolutions and measuring the selected parameters at constant engine speeds and constant load. Due to the fuel additive, the measured parameters of the combustion compression engine were gradually improved as power, torque, fuel consumption and fume. Key words: additive, diesel engine, fuel, engine speed characteristic 1. INTRODUCTION The European Union annually orders automakers to reduce the production of harmful pollutants into the air. Therefore automakers are trying to reduce the amount of pollutants generated by combustion. Methods of reducing emissions are provided by various technologies such as particulate filters. Emissions can also be reduced by reducing of engine volume, reducing fuel consumption or using charged engines. Making of pollutants is in direct proportion to the fuel consumption of a car. Ivan Janoško, Department of Transport and Handling, Faculty of Engineering, Slovak University of Agriculture in Nitra, Tr. A. Hlinku 2, Nitra, Slovak Republic, ivan.janosko@uniag.sk 2 Juraj Jablonický, Department of Transport and Handling, Faculty of Engineering, Slovak University of Agriculture in Nitra, Tr. A. Hlinku 2, Nitra, Slovak Republic, juraj.jablonicky@uniag.sk 3 Ľubomír Hujo, Department of Transport and Handling, Faculty of Engineering, Slovak University of Agriculture in Nitra, Tr. A. Hlinku 2, Nitra, Slovak Republic, lubomir.hujo@uniag.sk 4 Peter Kuchar, Department of Transport and Handling, Faculty of Engineering, Slovak University of Agriculture in Nitra, Tr. A. Hlinku 2, Nitra, Slovak Republic, xkuchar@is.uniag.sk 1 213

215 Fuel consumption can be affected by driving, but also with the right choice of fuel or additives to fuels. Since diesel fuel has a biofuel content, which also has negative characteristics such as clogging injectors, it also causes rapid oxidation in the fuel system, formation of deposits in the fuel tank and in the winter months reduces the fuel filtration point. The main goal of the additive is to improve the basic properties such as increasing the cetane number, improving the lubricity properties, avoiding oxidation, which ultimately has a positive effect on the life of the engine as well as on the lifetime of the fuel system. Depending on the type of additive, it can positively influence parameters such as engine power, emissions production and fuel consumption. 2. METHODICS The aim of the contribution is to evaluate the effect of the additive on the power and emission parameters along with the fuel consumption monitoring and the subsequent comparison of the measured values. The tested vehicle is Škoda Octavia category M1 with the 1.9 TDI diesel engine with a rotary injection pump. The vehicle's main parameter is contained in Table 1. Vehicle Škoda Octavia 1. generation Motor type ASV Cylinders capacity cm3 Highest engine power/speed kw / 4150 min-1 Maximum vehicle design speed 191 km.h-1 Operating weight 1275 kg Number of driven axles 1 / front Year of production 2002 Tires 195 / 65/ R15 - Barum Polaris 3 number of driven kilometers Table 1: Main parameter of tested vehicle The vehicle had during the test winter tires of size 195/65 / R15 by manufacturer Barum with model name Polaris 3 and 7.2 mm tread. Before the measurement, all the tires were checked and sucked at the same manufacturer's prescribed pressure at 220 kpa 2.1 Characteristics of working mediums The engine of the examined vehicle used engine oil from company SHELL, which is sold under the brand name Helix Ultra Diesel. It is a fully synthetic, easy-to-use engine oil. The oil viscosity is SAE 5W40 and meets VW norm. 214

216 During measurements were used fuels from brand Slovnaft. It was a basic range of fuels without the additive with the trade name Tempo plus diesel. In the tank contained approximately 25 liters what was the half capacity of the tank. Pumped diesel fuel met the requirements of standard EN 590 and also satisfies the conditions of the World Association of Automobile Manufacturers. During the testing was used winter diesel fuel, which has a lower temperature of filterability compared to summer diesel fuel. As tested additive was chosen known and in practice often used Super diesel additive made by VIF in a plastic bottle with 125 ml. It is a product constructed on the basis of 2-ethylhexyl nitrate, the manufacturer indicates improvement in the cetane number by 5 units, better combustion, reduced engine noise and lower fuel consumption by 5%. The additive serves to disposable one for 40 to 60 litres of fuel. In the tests, the entire volume of the additive was used, ensuring a dosing ratio of up to 1: 200 or, l: 25 l. 2.2 Characteristics of the instruments Fuel flowmeter: Due to inaccurate measuring of fuel consumption by a vehicle on-board computer, it was necessary to use a different, more accurate system. We used an AIC-5004 Fuel Flowmaster external fuel meter from AIC SYSTEMS AG, which joined the car's fuel system in its engine compartment. Due to the fact that the device is connected to a performance dynamometer, it does not need a display unit, but datas are displayed along with the other measured variables on the roller dynamometer. This makes it possible to read the exact values of a specific amount of fuel in specific engine operating modes. Figure 1: Fuel flowmeter Performance roller dynamometer: To measure the performance of the vehicle was used performance dynamometer by the German manufacturer MAHA with the designation MSR 500 with the possibility of measuring 4-wheel drive. 215

217 Figure 2: Performance roller dynamometer MAHA MSR 500 Parameter Value Cylinder diameter 504 mm / 20" Cylinder lenght 750 mm Minimum and maximum axle spacing 700 / 2,200 mm Rotating mass of the cylinder unit 280 kg Maximum axle load 2,5 t Measurable wheelbase 220 / 3200 mm Maximum speed 300 km.h-1 Maximum traction force 7000 N Table 2: Technical parameter of performance tester Exhaust gas analyser: To detect quantity of emissions in the exhaust gas was used analyser from brand MAHA and model MGT 5 / MDO2 - LON. It is a dual instrument to record the production of both petrol and diesel emissions. As well as the fuel flow meter, the device is connected to a performance dynamometer, what allows recording and displaying, in addition to the normal emission tests, the quantity of produced emissions depends on the combustion engine operation mode. 216

218 Figure 3: Exhaust gas analyser 2.3 The course of measurement The measurement process itself consisted of several important steps. The first task was to fix the car on the roller seat to avoid collision with help of steel ropes and chain links. It followed the connection of the flowmeter to monitor fuel consumption. The next step before the measurement was to connect the exhaust gas analyser, to monitor the exact values of exhaust and other harmful substances. Than was inserted oil temperature probe into the oil level gauge. The last step was to pair the vehicle via OBD diagnostics with a computer to record all values from the control unit and the devices to the computer. After completing these steps, the vehicle was ready for measurement. After completion of the initial steps, the vehicle had to be conditioned on the operating temperature to ensure the most accurate and trusted results. And then was necessary to calculation and analysis measured parameters used the following relationships: Calculation of performance 𝑃 = 𝑀𝑘. 𝜔 = 𝑀𝑘. 2. 𝜋. 𝑛 [kw] Calculation of torque 𝑃 𝑀𝑘 = 2.𝜋.𝑛 [Nm] The quantity of fuel consumed for the selected period of time 𝑉 = 𝑉7 𝑉1 [dm3.(30 s-1)] Hourly fuel consumption 𝑉= 𝑉.𝜌𝑝𝑎𝑙𝑖𝑣𝑎 𝑡 [kg.hod-1] 217 (1) (2) (3) (4)

219 3. RESULT Based on the external engine speed characteristics obtained from the MAHA performance dynamometer, it was possible to assess and compare the performance parameters of the vehicle. As can be seen in Table 3, after comparing the results, it was concluded that the engine power increased by 8.2 kw. Before adding the additive to diesel, the maximum power was 80.0 kw at 3455 rpm and km.h -1. Upon addition, the maximum power increased to 88.2 kw at 3920 rpm and km.h-1. Before using the additive After using the additive 80,00 kw [108,8 PS] 88,2kW [119,9PS] 81,8 kw [111,2 PS] 89,3 kw [121,5 PS] 3455 min-1 / 117,6 km h min-1 / 136,2 km h-1 Torque [Mnorm] 241,2 N.m-1 246,7 N.m-1 Max. torque in speed / velocity 2225 min-1 / 75,8 km.h min-1 / 79,6 km h-1 Corrected performance [Pnorm] Engine performance [Pmot] Max performance in speed / velocity Table 3: Comparison of performance parameters before and after adding additive to diesel For a more complex assessment, not only the highest values, but also the overall performance of the speed characteristic, we present a graph with the overlap of the power curves before and after the addition of the additive in Figure

220 Figure 4: Comparison of performance curve before and after adding additive to diesel Another parameter examined was the influence of fuel additive on engine smoke. This was performed as a sequence of 5 consecutive measurements before and 5 measurements after addition additives within the specified time range. The values of these measurements were averaged and the graphics shown in Figure 5. 0,14 Smoke [m-1] 0,12 0,1 0,08 0,137 0,06 0,097 0,04 0,02 0 Smoke without added additive Smoke with added additive Figure 5: Comparison of smoke before and after applying the additive 219

221 Similarly, to the measurement of smoke, we measured the fuel consumption in the following steps. Fuel consumption was measured by using a flow meter that was connected to the fuel system of the car. There was 5 measurements before and after the addition of the fuel additive, which was carried out in the constant load mode made by the performance tester, while the vehicle speed was kept constant. The periods of measurement are consistent with the measurement of smoke. Average fuel consumption figures can be seen in Figure 6. 6,6 6,55 Mph [l.h-1] 6,5 6,45 6,57 6,4 6,35 6,357 6,3 6,25 Consumption without added additive Consumption with added additive Figure 6: Comparison of fuel consumption before and after application of the additive 3. CONCLUSION The aim of the paper was to evaluate the impact of the additive on the vehicle's power and emission parameters along with fuel consumption. Experimental measurements were performed in a test laboratory on a preselected vehicle. For detection was chosen vehicle Skoda Octavia on volume of 1896 cm3, 81 kw performance, manual transmission and front-wheel drive. In the measurements, we confirmed the manufacturer's claims and that the vehicle even after 15 years fulfils all the parameters measured in the experimental part of the work. After using the diesel additive, a difference was noted during torque and engine power. During the measurement at the selected constant load, five measurements were made before adding the additive and five measurements after adding the additive and then compared the results. The results of the experimental measurements show the positive effect of the selected additive on fuel consumption and the smoke of the tested passenger car. However, it is necessary to take into account the high age and number of driven kilometers of a vehicle in which the combustion engine and its accessories may contain a greater amount of settling and the function of the individual mechanisms may no longer be in perfect condition, which may affect the measurement results. 220

222 REFERENCES [1] JABLONICKÝ, J. TKÁČ, Z. MAJDAN, R. UHRINOVÁ, D. HUJO, Ľ. VOZÁROVÁ, V. Hodnotenie vlastností biopalív a biomazív, 1. vyd. Nitra : Slovenská poľnohospodárska univerzita, s. ISBN [2] KRÁLIK, M. JABLONICKÝ, J. TKÁČ, Z. HUJO, Ľ. KOSIBA, J. Dymivosť vznetového motora pri aplikácii nekonvenčného paliva, In XXXVIII. mezinárodní konference kateder dopravních, manipulačních, stavebních a zemědělských strojů. Plzeň: University of West Bohemia, 2012, s ISBN [3] KRÁLIK, M. JABLONICKÝ, J. TKÁČ, Z. HUJO, Ľ. UHRINOVÁ, D. KOSIBA, J. TULÍK, J. ZÁHORSKÁ, R. Monitoring of selected emissions of internal combustion engine. In Research in agricultural engineering. 2016, s. S66--S70. ISSN [4] ŠVEC, J. DUREC, J. CHRASTINA, J. TULÍK, J. Evaluation of inspected vehicles in reporting period at technical inspection and emission control station, In Acta technologica agriculturae. 16, 2013, s ISSN [5] UHRINOVÁ, D. JABLONICKÝ, J. HUJO, Ľ. KOSIBA, J. TKÁČ, Z. KRÁLIK, M. CHRASTINA, J. Research of limited and unlimited emission effect on the environment during the burning of alternative fuels in agricultural tractors, In Journal of Central European Agriculture. Zagreb : University of Zagreb. 14, 4,2013, s ISSN [6] UHRINOVÁ, D. JABLONICKÝ, J. HUJO, Ľ. TKÁČ, Z. KUČERA, M. KOSIBA, J. Measurement of operating parameters and emissions of tractor with diesel oil and biofuel. In TEAM. 4, 1, 2012, s ISSN ACKNOWLEDGMENT This work was supported by AgroBioTech Research Centre built in accordance with the project Building AgroBioTech" Research Centre ITMS The contribution was made under the grant project of the Ministry of Education of the Slovak Republic VEGA 1/0464/17 Monitoring of the impact of ecological fuels derived from agricultural production and impurities in hydrocarbon fuels to technical and environmental performance of internal combustion engines used in agricultural and transport technology. 221

223 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES DIESEL ENGINE EMISSIONS IN REAL WORLD DRIVING: LABORATORY LIMITS ACHIEVED ON THE ROAD? Jan Skácel1, Martin Pechout2, Michal Vojtíšek3 Abstract Latest regulation for vehicle emissions requires manufacturers to meet stringent limits on certification drive cycle in laboratory by using advanced engine and catalyst technology. It has been debated that real world driving performance is different. The objective of this study is to understand how diesel powered vehicle behaves on the road and why. Using portable emission measurement system, road data were collected from a Euro 5 diesel vehicle. Results show that some road driving emissions were very low and close to laboratory limits, whereas others are much higher. Control strategy and catalyst sizing seem to be the key factors. 1. INTRODUCTION Vehicle emission regulation is a key factor in reducing air pollution and protecting environment and human health. Vehicle manufacturers are required to comply with the regulation in order to certify and sell their vehicles. The regulation specifies emission limits for laboratory test over defined driving cycle, durability and in use on-board diagnostic requirements. It has been debated how relevant defined drive cycles and laboratory testing is compared to real world driving. Various studies have shown that real driving emissions can differ significantly from the certification drive cycle in laboratory environment [1,2]. For example, in 2015 German ADAC tested 32 Euro 6 diesel vehicles [1]. The results show that under certification conditions under laboratory testing all vehicles demonstrated emissions within the declared limits; however same tests run on track Jan Skácel, Czech Technical University in Prague, Faculty of Mechanical Engineering, Department of Automotive, Combustion Engines and Rail Vehicle Engineering, Technická 4, Prague 6, Czech Republic, jan.skacel@fs.cvut.cz 2 Martin Pechout, Technical University of Liberec, Faculty of Mechanical Engineering, Department of Vehicles and Engines, Studentská 2, Liberec, Czech Republic, martin.pechout@tul.cz 3 Michal Vojtíšek, Czech Technical University in Prague, Faculty of Mechanical Engineering, Department of Automotive, Combustion Engines and Rail Vehicle Engineering, Technická 4, Prague 6, Czech Republic, michal.vojtisek@fs.cvut.cz 1 222

224 outside the laboratory showed increased emissions. Under real driving conditions the emission limits were significantly higher. Newly proposed European regulation challenges the auto industry by requiring use of real driving emissions (RDE) as supplemental test to laboratory testing [3]. By using portable emission measurement system (PEMS), vehicle manufacturers are going to measure and declare vehicle emissions from a road test as well as from the laboratory. At this time the use of RDE is in evaluation phase with anticipated implementation in 2020/21. Expectation is that RDE test will have to meet the same limits as the laboratory test. The objective of this evaluation is to gain experience with RDE and to find out how to meet the challenging goal of RDE compliance. That is also the objective of this study: to explore, on example of a diesel vehicle, how the vehicle behaves on the road in various driving conditions, why exactly there is variation in emission performance and what can be done to improve the results. 2. CATALYST TECHNOLOGY OVERVIEW Modern engines face challenging requirements on performance, fuel efficiency and emissions. Because meeting one requirement usually leads to deterioration of other, it is not possible to meet all of them with only engine design and control. Therefore combination of engine measures and aftertreatment catalyst technology is used. Emission catalysts provide extremely efficient way to abate harmful pollutants - CO, HC, NOx and soot that are harmful to human health and/or environment. To eliminate all of the pollutants, stoichiometric air-fuel ratio in the exhaust is needed. That is easily achieved on gasoline engine, which almost exclusively uses three way catalyst as the only single aftertreatment component. Its application is straightforward with the only challenge being the cold start. However that can be solved by thermal management and/or catalyst technology. Diesel engine is much more challenging to make clean. Its exhaust gas is lean, i.e. with excess oxygen, which makes oxidation of CO, HC and soot easier, but reduction of NOx is a difficult challenge. Beside reducing engine-out NOx emissions to meet the target, which is increasingly difficult, the only solution is to use some reducing agent together with NOx catalyst. One approach uses additional reductant, urea, in selective catalytic reduction (SCR). Another solution is to use NOx storage catalyst lean NOx trap (LNT) - that adsorbs NOx during normal lean engine operation and is purged by artificial fuel rich occasions. Heavy duty vehicles typically use diesel oxidation catalyst (DOC) for hydrocarbon (HC) and CO abatement, catalyzed soot filter (cdpf) for soot removal and SCR with urea (AdBlue, DEF) for NOx. Light duty segment is more diversified with various systems and combinations. For NOx abatement all three options are used - engine only measures, LNT or SCR. 223

225 SCR is continuous operation system, i.e. continuously works throughout normal operation of the engine. The only exception is very infrequent soot regeneration of the DPF and de-sulfation cycle. SCR system requires urea dosing and management systems, urea tank, injector, heating and cooling, but offers best efficiency. LNT on the other hand is less efficient and has discontinuous operation, but is easier to implement and does not require any additional hardware or medium. LNT function during both NOx storage and regeneration phase is described below [4]. Loading - NOx storage (lean condition) Pt NO oxidation NO 12 O 2 NO2 -CO 2 3 NO 2 BaCO 3 Ba(NO 3 ) 2 NO NOx storage (1) (2) Regeneration N2 release (rich condition) CO 2 Ba(NO 3 ) 2 3 H 2 BaCO 3 2 NO 3 H 2 O NOx oxidation (3) NOx release 2 NO 2H 2 N 2 2 H 2 O (4) For LNT regeneration excess fuel is typically used. That means fuel penalty of few %. 3. DIESEL ENGINE EMISSIONS FROM REAL DRIVING To evaluate real driving data measurement of tailpipe emissions Škoda Octavia III 1.6 TDI Green line was equipped with PEMS. This vehicle uses combination of close coupled LNT and cdpf catalysts. Parameters of the vehicle and aftertreatment layout are shown below. Škoda Octavia Combi III 1.6 TDI GreenLine 1.6 TDI 4 cylinder diesel 1598 cm3 EURO 5 EGR + LNT- > cdpf Engine Engine volume Emission limit Aftertreatment system Fuel consumption urban/extra-urban/combined 3.8 / 3.0 / 3.2 l/100km Emissions of CO2 Transmission Maximum power Maximum torque Engine speed at 130km/h 85 g/km 6-speed manual 81/3250 kw/min-1 250/1500 N.m/min min-1 Table 1: Test vehicle specification 224

226 ECU Exhaust counterpressure valve p CDPF LNT 1 T1 2 T3 T2 Figure 1: Engine aftertreatment layout Skoda Octavia III 1.6l TDI green line Catalysts: Close-coupled LNT and catalysed DPF with counter-pressure valve to assist with regeneration downstream. Instrumentation: lambda sensor engine out & tailpipe, pressure drop across cdpf and temperatures on inlet and outlets 3.1 Measurement Equipment and vehicle instrumentation The concentrations of gaseous pollutants were measured by a prototype industrialgrade Fourier Transform Infra-Red spectrometer (FTIR). The instrument features a Michelson interferometer, zinc selenide optics, liquid nitrogen cooled mercury cadmium telluride detector, running at optical resolution of 0.5 cm-1, one scan per spectra, with a 0.2 dm3, 6 m optical path length cell heated to 121 C. It uses a nominal flow of 15 dm3 min-1 and is preceded by a 3-meter sample line and filter, both heated to 121 C. The instrument provided infrared spectra every 2-3 s. By deconvolution of the absorption spectra, concentrations of various pollutants of interest, including NO, NO2, NH3, N2O, CO and CO2, were calculated using a method developed in-house and using library spectra or spectra obtained by measuring calibration gases. The system weights 35 kg, and consumes about W. The concentrations of NO and NH3 were found to correlate well with laboratory during tests on two Euro 6 cars [5]. The instrumentation also included engine on-board diagnostics (EOBD) interface, a laptop computer, and a 26 V 90 Ah bank of electric vehicle LiFeYPo batteries and a 1.5 kw charger-inverter. The power consumption has been minimized to ensure approximately 8-hour runtime. Intake air flow has been obtained from the (OBD) mass air flow (MAF) sensor information. Vehicle and engine speed and intake manifold pressure were also recorded. All data have been synchronized in time and normalized by interpolation to 1 Hz. The exhaust flow has been determined from measured exhaust gas composition, assumed fuel composition, and intake air mass flow obtained from the EOBD. Molar flows, fuel consumption and other parameters were calculated from measured values at local atmospheric conditions. Molar weight of NO was calculated as NO 2. Vehicle installation of PEMS is shown on Fig. 2 below. FTIR is placed in the trunk whereas battery pack is in rear seats. 225

227 Figure 2: Portable emissions measurement equipment used in this study: Left FTIR with filter and heated sampling lines in the trunk, Right battery pack on rear seats 3.2 Results During the test, total of over 500 km were measured during trip from Liberec, CZ to JRC in Ispra, Italy. Two measurement segments were recorded: 4 hours of driving in Germany and 1,5h driving in Switzerland, see table 2 below. Distance [km] Travel time [h:m:s] Average speed [km/h] Fuel consumption [l] Avg. fuel cons. [l/100km] CO2 emissions [g/km] Cumulative TP NOx [g] Avg. NOx [mg/km] Comparison to laboratory EURO 5 NOx limit 180 mg/km TOTAL DE CH 516,5 5:57:33 86,7 30,17 5,84 154,7 315, ,7 4:33:58 94,6 25,68 5,95 157,5 296, ,7 1:23:35 60,8 4,49 5,30 140,5 18, ,39 3,82 1,22 Table 2: Test Results Overview of Total, German and Swiss segment of the trip Overall details of the recordings are plotted on Fig. 3 and Fig. 4. The record consists of three particular parts: a/ Highway driving in Germany in heavy traffic b/ Main roads in Switzerland c/ Highway in Switzerland Please note that the data has been edited to eliminate extended gaps due to safety breaks, shutdown of the instruments during passage through customs, etc. As a result 226

228 all segments of the record have been attached together. Fig. 3 shows vehicle speed, NOx flow and cumulative NOx emissions in time domain, Fig. 4 similar data but distance based. It is apparent that German segment had higher speed and higher NO x emissions, especially during very high speed around 160 km/h. Only slightly lower speed (~140 km/h) yielded significantly lower NOx emissions. Switzerland with top speed of 120 km/h and long slow segment showed significantly lower NO x emissions. Switzerland b/ main road c/ highway Germany a/ highway R1 + R2 + Figure 3: Vehicle speed, tailpipe NOx flow and cumulative NOx emissions time based Germany Switzerland EURO 5 limit 0.18 g/km Figure 4: Vehicle speed, tailpipe NOx flow and cumulative NOx emissions distance based 227

229 Where German segment had average NOx emissions almost 690 mg/km, which is almost 4 times higher than laboratory limit for Euro 5 of 180 mg/km, Swiss segment was only ~220mg/km, which is only 20% higher that Euro 5 limit. Although on-road emissions are not directly comparable Euro 5 limit which only applies to the Type I approval test (including NEDC drive cycle), it reveals how real world emissions compare to laboratory. The correlation between speed and NOx emissions is clearly visible from Fig. 5, which shows NOx emissions per 1km in terms of vehicle speed and the excessive emissions at high speed. EURO 5 limit 0.18 g/km Figure 5: Tailpipe NOx emissions per 1km based on average vehicle speed German and Swiss segment of the trip In order to understand this trend, whether it is due to high engine out emissions, regeneration strategy, or diminished NO x storage capacity due to high temperature, catalyst performance needs to be investigated by comparing engine-out and tailpipe emissions. Without additional sampling in engine out position, we can only work with tailpipe data. Fig. 6 shows typical LNT loading and regeneration cycle under steady state conditions Swiss 120 km/h cruise control. It can be seen that tailpipe NOx improve after each regeneration and then gradually increase until the next one. From air/flow ratio spiking into rich mode and simultaneous CH 4 spike it is also clear when regeneration is triggered. It is obvious that NOx slip starts to occur immediately after each regeneration. This is an indication of poor catalyst performance as a good catalyst should have period of no slip, with the slip gradually increasing with increasing amount of stored NO x. Cumulative NOx slip between two regenerations (area A) and frequency of regenerations indicates performance of the catalyst, see Fig. 7. Good catalyst and 228

230 control would keep NOx slip below certain threshold without excessive regeneration frequency. It can be seen that during Swiss segment, NO x slip was kept low and regeneration interval varying accordingly as needed. During the German segment though the NO x slip was not kept low despite number of frequent 2 min. regenerations. There were also two outstanding events, R1 & R2, where regeneration was delayed and resulted in substantial NOx slip. R1 & R2 only combined contributed to 56g of NO x, about 20% of total NOx. It is not clear what caused those two events. Details of the R1 event are plotted on Fig. 8. There is ~6min gap in regenerations that resulted in 22g of NOx slip. Before and after R1 the regenerations interval was ~2 min. CO2 (A/F ratio) regeneration CH4 regeneration A Figure 6: LNT Regeneration during constant vehicle speed of 120 km/h, trail segment C Cumulative NOx slip between two regenerations area A R2 R2 R1 R1 Figure 7: LNT Regeneration performance: Tailpipe NOx slip between two regenerations vs. regeneration time interval & fuel consumed within the regeneration time interval Overview of the 123 regenerations (of which 99 in DE & 24 in CH) 229

231 NOx slip 22.3g R1 No regen.~6 min Figure 8: Regeneration point R1 4. CONCLUSION Experiment with on-road emissions showed that despite no legal requirement, road emissions can be very low and comparable to the laboratory certification cycle limits for EURO5. That conveys a positive and encouraging message that maintaining emissions comparable to the type approval limit can be done on the road. However, high speed in particular tends push engine out NO x and hence the tailpipe NOx to much higher levels. The results suggest that to improve emission performance better control of NO x slip and regenerations will be needed. That will necessitate implementing NO x sensor downstream of LNT and perhaps bigger catalyst volume and/or better LNT technology. Outlook for the successful implementation of RDE into European regulation looks positive. RDE can approach certification limits partially today, therefore, it must be possible to make it work to meet the laboratory standards. 230

232 REFERENCES [1] [2] [3] [4] [5] Liuhanzi Yang, Vicente Franco, Alex Campestrini, John German, and Peter Mock : NOx Control Technologies for Euro 6 Diesel passenger cars, International Council on Clean transportation, white paper, September 2015 Department for transportation, UK: Vehicle Emissions Testing Programme, government publication, April 2016, ISBN Mock, P.: Real-Driving Emissions test procedure for exhaust gas pollutant emissions of cars and light commercial vehicles in Europe, ICCT policy update publication, January 2017 Ronald M. Heck, Robert J. Farrauto, Suresh T. Gulati: Catalytic Air Pollution Control: Commercial Technology, Wiley; 3 edition (March 7, 2016) ISBN-13: Suarez-Bertoa, R., Mendoza-Villafuerte, P., Riccobono, F., Vojtisek, M., Pechout, M., Perujo, A., Astorga, C., On-road measurement of NH3 emissions from gasoline and diesel passenger cars during real world driving conditions, Atmospheric Environment (2017), doi: /j.atmosenv (in press) ACKNOWLEDGEMENT This research has been realized using the support of The Ministry of Education, Youth and Sports program NPU I (LO), project # LO1311 Development of Vehicle Centre of Sustainable Mobility and by Technological Agency, Czech Republic, programme Centres of Competence, project # TE Josef Božek Competence Centre for Automotive Industry. This support is gratefully acknowledged. 231

233 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES SIMULACE JÍZDY HYBRIDNÍHO VOZIDLA Pavel Brabec1, Daniel Pešek2, Robert Voženílek3 Abstrakt The paper describes a model of a hybrid vehicle using simulation software Ricardo Ignite. A comparison of a conventional vehicle with an internal combustion engine and a vehicle with a parallel hybrid drive arrangement was made using simulations. The result of fuel consumption for both types of propulsion was compared to the NEDC and WLTP driving cycles. Values for the 1.0 TSI combustion engine and the ZERO ZFORCE 75-7 synchronous electric motor were used in simulations. 1. ÚVOD Uspořádání paralelního hnacího hybridního ústrojí vozidel se vyznačuje tím, že spalovací motor předává výkon kolům přímo. Do tohoto systému je paralelně zapojen i elektromotor a generátor, přičemž primární pohon může být spalovacím motorem a při okamžitém zvýšeném výkonovém požadavku se připojí také elektromotor, který je s tímto systémem spojen mechanickou vazbou. Spalovací motor proto nemusí být dimenzován na celkový výkon automobilu. Mohou ovšem nastat i jízdní režimy, kdy je upřednostněn pouze pohon elektromotorem. To je dáno strategií řízení pohonů. Oproti sériovému uspořádání není snížena celková účinnost a to vlivem přímého spojení spalovacího motoru s koly. 3. KOMPONENTY HNACÍHO ÚSTROJÍ V simulačním modelu byly použity hodnoty reálného tříválcového zážehového agregátu 1.0 l TSI s maximálním výkonem 85 kw při /min a maximálním točivým momentem 200 Nm při /min. Do modelu byla zadána změřená mapa spotřeby paliva. Ing. Pavel Brabec, Ph.D., Technická univerzita v Liberci, pavel.brabec@tul.cz Ing. Daniel Pešek, Technická univerzita v Liberci, daniel.pesek@tul.cz 3 Ing. Robert Voženílek, Ph.D., Technická univerzita v Liberci, robert.vozenilek@tul.cz

234 Obr. 1: Vnější otáčková charakteristika spalovacího motoru 1.0 TSI 6. Pro elektrickou část paralelního hybridního pohonu byl použit elektromotor ZERO ZFORCE 75-7, který se používá u motocyklů značky ZERO. Tento synchronní vzduchem chlazený elektromotor je úspěšně využíván v laboratořích pracoviště pro pohon experimentálního elektrického vozidla a na zkušebních stanovištích proběhlo ověření jeho vlastností. Elektromotor je řízen regulátorem Sevcon size 4 a pohon může pracovat ve třech režimech. Pro případ simulačního modelu práce byl zvolen režim, kde je elektromotor schopen dodávat elektrický proud 300 A po dobu 120 s, než se regulátor přehřeje. V tomto režimu elektromotor vytváří maximální moment 72 Nm při /min a maximální výkon 30,6 kw při /min, vnější otáčková charakteristika je ukázána na obrázku 2. Elektromotor je také schopen krátkodobě (po dobu 10 s) dodávat 360 A. Poslední režim, jenž umožňuje elektrické hnací ústrojí, je trvalý režim při elektrickém proudu 120 A. Obr. 2: Vnější otáčková charakteristika elektromotoru ZERO Z-FORCE

235 4. SIMULAČNÍ MODEL Pro simulaci hnacího ústrojí vozidla byl využit software Ricardo Ignite. Simulace byla provedena pro vozidlo s klasickým hnacím ústrojím využívající pouze spalovací motor a dále pro model s paralelním uspořádáním hybridního hnacího ústrojí vozidla. Byly posuzovány různé varianty připojení elektrického motoru v rámci hybridního uspořádání hnacího ústrojí. Pro potřeby příspěvku byla vybrána varianta s připojením elektromotoru za spojku zobrazeného na obrázku 3. Obr. 3: Simulované hybridní uspořádání hnacího ústrojí v softwaru Ricardo Ignite. Model vozidla v sobě obsahoval parametry vozidla Škoda Octavia III a vozidlo projíždělo definované testy v našem případě NEDC a WLTP test, který je předepsán platnou legislativou. Parametry testů a rámcové porovnání jsou zobrazeny na obrázku číslo 4. Obr. 4: Porovnání jízdních cyklů NEDC a WLTP

236 5. VÝSLEDKY SIMULAČNÍCH VÝPOČTŮ Vozidlo s hybridním uspořádáním využívalo v softwaru předdefinovanou jízdní strategií pro využívání jednotlivých pohonů ( připojení a odpojení elektrického motoru ke spalovacímu motoru, regeneraci brzdné energie, Start-Stop systém). Počátek dobíjení akumulátoru byl nastaven při 58 % kapacity baterie a konec dobíjení při 80 % kapacity baterie. Maximální dobíjení nastávalo při /min a minimální dobíjení je při /min. Dále byla nastavena strategie řazení (mapa řazení) a řídící jednotka převodovky. Výsledky simulovaných hnacích ústrojí musely v definovaných testech také splňovat podmínku tolerančního pole rychlosti dané normou. Obr. 5: Využívání jednotlivých typů pohonů během simulovaného NEDC testu. Obr. 6: Nabíjení a vybíjení baterie elektrického pohonu během NEDC testu. 235

237 Obr. 7: Využívání jednotlivých typů pohonů během simulovaného WLTP testu. Obr. 8: Nabíjení a vybíjení baterie elektrického pohonu během WLTP testu. Vozidlo využívající pouze spalovací motor mělo na simulované trati NEDC testu přepočtenou spotřebu paliva 6,4 l/100 km. Pokud bylo využito zvolené koncepce hybridního pohonu, tak se přepočtená spotřeba paliva u spalovacího motoru snížila na 4,1 l/100 km. Pro simulovaný WLTP test zvolená hybridní koncepce pohonu vozidla vykázala přepočtenou spotřebu 4,9 l/100 km. 3. ZÁVĚR Pro simulační výpočty byly použity hnací agregáty 1.0 l TSI 85 kw a ZERO Z-FORCE ,6 kw, které jsou v současné době k dispozici v laboratoři pracoviště. Hybridní uspořádání pohonu při simulačních výpočtech prokázalo významný potenciál při snižování spotřeby paliva s tím souvisejícím poklesem produkce CO2. V dalším kroku se budou práce soustředit na zpřesnění hodnot simulačního bloku vtaženého k baterii a dále budou posuzovány další možnosti v nastavení řídící jednotky v oblasti strategie řízení pohonů, kde nebyl pravděpodobně vyčerpán celý potenciál. 236

238 V článku je popsán výzkum, na kterém pracuje KVM v rámci pracovního balíčku WP25 projektu TE Centrum kompetence automobilového průmyslu Josefa Božka. REFERENCES [1] [2] [3] [4] [5] [6] [7] PEŠEK, D.: Metodika redukce CO2. Diplomová práce Technická univerzita v Liberci, Úřední věstník Evropské unie. Jednotná ustanovení pro schvalování vozidel z hlediska emisí znečišťujících látek podle požadavků na motorové palivo. [Online] Březen (EHK OSN) č. 83. UN ECE. Proposal for a new UN Global Technical Regulation on Worldwide harmonized Light vehicles Test Procedures (WLTP). Září ECE/TRANS/WP.29/AC.3/26. MOCK, P., KUHLWEIN, J., TIETGE, U., FRANCO, V., BANDIVADEKAR, A., GERMAN, J.: The WLTP: How a new test procedure for cars will affect fuel consumption values in the EU. THE INTERNATIONAL COUNCIL OF CLEAN TRANSPORTATION. [Online] 29. Říjen SCHÖPPE, D., ZHANG, H.,RÖSEL, G., ACHLEITNER, E., KAPPHAN, F., DUPONT, H.: Next Generation Engine Management Systems for Gasoline Direct Injection. 34. Internationales Wiener Motorensymposium DVOŘÁK, F.: Auto.idnes.cz. [Online] Zero Motorcycles. zeromotorcycles. [Online] 237

239 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES PŘENOS VÝKONU A ŘÍZENÍ PARAMETRŮ Pavel Brabec1, Miroslav Malý2 Abstract Vozidla automobily i samojízdné pracovní stroje - užívají k vlastnímu pohybu hnací ústrojí různých typů. Ústrojí s touto funkcí je systém, který zajišťuje dodávku energie k překonání jízdních a dalších odporů při pohybu vozidla. Podle formy energie, která se při přenosu využívá, můžeme charakterizovat typ přenosu, resp. převodu jako mechanický, hydraulický, elektrický a případně též kombinovaný. Vývoj hnacích ústrojí vozidel a mobilních strojů se stále více orientuje na využití výhod takových konstrukcí ústrojí, která umožňují velkou variabilitu v ovládání a která splňují stále přísnější požadavky na provoz. Mimořádná pozornost je pak věnována plnění kritérií, jež mají osvědčit jednak efektivní využití energie, ale zejména pak udržet v mezích tzv. emisní stopu. 1. ÚVOD - PŘENOS ENERGIE A PŘENOS VÝKONU Přirozený tlak na úspory a v různých regionech světa měnící se požadavky, s tím související legislativa s emisními limity určují další vývoj hnacích ústrojí, ale i souvisejících systémů. Hnací ústrojí musí plnit řadu požadavků, které mohou být ve svém důsledku i protichůdné. Splnění hlavních (jako např. vysoký výkon a úspora energie, komfort, bezpečnost a hygiena, ochrana životního prostředí před negativními důsledky provozu ad.) vyžaduje výrazný zásah do tradičního pojetí mechanických a kombinovaných systémů. Takovým zásahem může být i aplikace vhodných řídicích systémů, které mohou korigovat parametry tak, aby bylo dosaženo požadovaných provozních poměrů při přijatelných dopadech na životní prostředí. To ale není nic nového, takový úkol je před konstruktéry již více jak půl století. A technický pokrok je znatelný, potvrzuje to mj. i nezanedbatelný pokles spotřeby i emisí ve výfukových plynech. Ústrojí vozidel i kolových samojízdných pracovních strojů plní různé funkce, k vlastnímu pohybu slouží ústrojí sestávající se z pohonné jednotky a dalších částí včetně kol. Hnací ústrojí s touto funkcí je systém, který zajišťuje přenos a transformaci energie a výkonu k překonání jízdních odporů při pohybu vozidla nebo stroje. Podle formy energie, která se při přenosu využívá, můžeme charakterizovat typ přenosu, 1 2 Ing. Pavel Brabec, Ph.D. odborný asistent, katedra vozidel a motorů, Technická univerzita v Liberci Doc. Ing. Miroslav Malý, CSc. katedra vozidel a motorů, Technická univerzita v Liberci 238

240 resp. převodu jako mechanický, hydraulický, elektrický a případně též kombinovaný. Vývoj hnacích ústrojí mobilních strojů se stále více orientuje na využití výhod takových konstrukcí ústrojí, která umožňují velkou variabilitu v ovládání a splňují stále přísnější požadavky na provoz. Přenos výkonu je doprovázen různými formami nežádoucích přeměn. Jistá část energie (byť se přeměny účastní) není efektivně využita k požadovanému účelu. Tato část energie, resp. výkonu bývá často charakterizována jako "ztrátový výkon", tedy výkon k požadované operaci nevyužitelný. Pro energetickou bilanci v kvazistatickém režimu je nutné stanovit podíl nežádoucích přeměn energie, v přechodovém stavu je třeba dále zohlednit i nezanedbatelný vliv akumulátorů energie. Chceme-li charakterizovat převodový systém z hlediska přenosových vlastností, musíme stanovit odpory proti přenosu energie. Odpory zpravidla rozdělujeme na ty, které souvisí s nevratnou změnou a na ty, které souvisí s vratnou přeměnou energie. I když obě skupiny nelze zcela oddělit, je třeba poukázat na charakteristické rysy jednotlivých odporů a na možné ovlivnění přenosu energie. V neustáleném přechodovém stavu je tento děj provázen jevy, jež jsou ovlivněny účinky akumulátorů kinetické a potenciální energie. Výměna energie mezi nimi děj komplikuje, může docházet ke kmitání. V ustáleném nebo kvazistatickém stavu se sice hlavní parametry nemění, přesto přenos výkonu není dokonalý. Část je disipována, účinnost není ideální. V souvislosti s negativními jevy si vedle celkové účinnosti h připomeneme dva charakteristické parametry vyjadřující přenos hlavních veličin. Přenos rychlosti je definován jako poměr výstupní a vstupní rychlosti (kinematický převodový poměr ), přenos momentu je definován jako poměr momentu na straně výstupu a momentu na straně vstupu (momentový převodový poměr ). Oba parametry vyjadřují kvalitu přenosu - jednak rychlosti jakožto žádané veličiny, ale i zatížení, jež je veličinou poruchovou. Na základě nich lze usoudit, oč je skutečná rychlost nižší oproti teoretické, a jak se navyšuje zátěž vlivem pasivních odporů. 2. ŘÍZENÍ PŘEVODOVÉHO SYSTÉMU Cílem řízení je dosažení určitého stavu a provozního bodu v reálné soustavě tak, aby byly zajištěny optimální přenosové poměry. Účelovou funkcí a často uváděným kritériem bývá účinnost, pro spalovací motor může být vyjádřena specifickou spotřebou paliva. Bude-li pro použitý spalovací motor definovatelná křivka optimální spotřeby, tedy taková množina provozních bodů, v nichž pro požadovaný výkon dosahuje motor minimální specifické spotřeby, mohlo by být cílem řízení nastavit provozní bod co možná nejblíže této křivce. To s ohledem na účinnost spalovacího motoru. Pro takový zásah máme k dispozici volné řídicí veličiny: na straně spalovacího motoru i na straně převodu. Jejich variací bychom chtěli dosáhnout požadovaného (optimálního) provozního stavu celého hnacího ústrojí. Je však málo pravděpodobné, že se optima převodové části (tedy pracovní bod s optimální účinností převodu) a spalovacího motoru (pracovní bod s optimální měrnou spotřebou) promítnou do odpovídajícího pracovního bodu. 239

241 2.1 PŘÍKLAD VARIACE PŘEVODŮ Hledat optimální pracovní bod hnacího ústrojí s mechanickou stupňovou převodovkou (klasickou hřídelovou s ozubenými koly) lze ukázat na příkladu variance převodových stupňů. V uvedeném příkladu byla užita zjednodušení. Byl zvolen automobil a jízdní režim, jízdní odpory byly stanoveny pro ustálenou rychlost 90 km/h, resp. 120 km/h. Celkový převod mezi koly a spalovacím motorem byl určen zařazeným nejvyšším převodovým stupněm převodovky a stálým převodem, tedy u dvou pětistupňových variant s výslednou hodnotou 3,484 a 3,063, resp. u šestistupňové varianty 2,906. Tabulka 1 Rychlostní charakteristika (točivý moment, výkon a hodnoty točivého momentu pro výkon při rychlosti vozu 90 km/h a 120 km/h, tedy pro konstantní výkon P1 = 15 kw, resp. P2 = 30 kw) n (ot./min) rad/s) MMAX (N.m) 125,0 132,0 137,0 138,0 140,0 150,0 146,0 144,0 138,0 126,0 PMAX (kw) 19,6 27,6 35,8 43,3 51,3 62,8 68,8 75,4 79,4 79,1 M90 (N.m) 95,5 71,7 57,3 47,8 40,9 35,8 31,8 28,7 26,1 23,9 M120 (N.m) 191,1 143,3 114,6 95,5 81,9 71,7 63,7 57,3 52,1 47,8 Pro zjednodušení výpočtu byla uvažována konstantní (a pro uvedené varianty stejná) na parametrech nezávislá hodnota účinnosti převodu. Potřebný výkon spalovacího motoru byl tak určen hodnotou P1 = 15 kw, resp. P2 = 30 kw. Z níže uvedené charakteristiky (obr. 1) a polohy pracovních bodů je patrná dostatečná výkonová rezerva pro případnou změnu charakteru zatěžování (změna stoupání, případná potřeba zrychlovat). To ostatně není překvapivé, neboť parametry spalovacího motoru, ale i uvedených převodovek jsou relevantní. Tabulka 2 Parametry pracovních bodů převodový poměr I ve třech variantách (pro rychlost vozu 90 km/h, resp. 120 km/h) V rychlost i přev. st. km/h nsm nk MSM mp M Q ot./min. ot./min. N.m rad/s g/kw.h kg/100km l/100km PA(15 kw) PB(15 kw) , ,9 753,8 54,6 274, ,66 6,1 3, ,5 753,8 62,1 241, ,50 5,8 PC(15 kw) 90 2, ,5 753,8 65,4 229, ,43 5,7 PD(30 kw) PE(30 kw) PF(30 kw) , ,2 1005,0 81,9 366, ,67 8,7 3, ,0 1005,0 93,1 322, ,30 8,2 2, ,6 1005,0 98,1 305, ,23 8,1 A i když volba jízdních režimů, která v uvedeném příkladu posloužila pouze pro porovnání, představuje velké zjednodušení, a to i proti laboratorním zkouškám 240

242 s rozmanitým charakterem rychlostních a zatěžovacích cyklů, přesto i tento jednoduchý příklad (s charakteristikou spalovacího motoru) ukazuje značnou citlivost účinnosti vyjádřené specifickou spotřebou na volbě provozních bodů kw M [Nm] 100 F E 15 kw D 80 C B 60 A n [1/min] Obr. 1 Pracovní body v charakteristice zážehového spalovacího motoru 2.2 VARIABILNÍ PŘEVODOVKY Variabilní převody mají mnohé přednosti. Široký rozsah plynule měnitelných převodových poměrů je pro aplikace v mobilní technice velkým přínosem, v kombinaci s inteligentním řídicím systémem poskytuje nové možnosti. Ale zatímco mechanické převodovky s ozubenými koly mají zařazením příslušného rychlostního stupně kinematickou vazbu jednoznačnou, a tedy pracují bez tzv. skluzu a ztráty rychlosti, převodovky s proměnnými poměry jsou typické variabilní kinematickou vazbou. Přenosové poměry jsou složitější. Kinematická vazba ale ani přenos zatížení nejsou hodnoty konstantní na parametrech nezávislé. Jak kinematický převodový poměr, tak i momentový převodový poměr na parametrech závisí, a nejen na geometrii, ale i na parametrech rychlosti a zatížení. Geometrický parametr je zpravidla proměnný a závislý na volné řídicí veličině (lze měnit geometrii - objem hydrostatických převodníků, geometrické poměry řemenových variátorů ad.). Potom přenos rychlosti můžeme vyjádřit vztahem 2 1 kde kinematický převodový poměr není konstantní, ale závisí na parametrech = G,, M, přenos zatížení (momentu) lze popsat analogicky M1 = M2 /, kde = G,, M, a celková účinnost charakterizuje kvalitu přenosu výkonu, tedy jak rychlosti, tak i zatížení h., resp. je také funkcí parametrů h = f G,, M, 241

243 Zatímco mechanické převody s ozubenými koly jsou zatíženy pasivními účinky, které celkovou účinnost sice snižují, ale účinnost přenosu rychlosti je v ustáleném stavu absolutní (rychlost se v pevné kinematické vazbě neztrácí ). Celková účinnost převodů proměnných (ať řemenových variátorů, hydrodynamických, hydrostatických nebo elektrických systémů) je na parametry velmi citlivá a rozptyl hodnot je obvykle velký. Přenos obrazu pracovního bodu do charakteristiky spalovacího motoru je po zpřesnění poněkud posunut (od teoretického je více vpravo a výše: spalovací motor tak musí překonat větší odpor, který odpovídá součtu zatížení a pasívních účinků, a rovněž musí generovat vyšší otáčky o tzv. skluz, který vzniká vlivem nedokonalé kinematické vazby). 3. ZÁVĚR Vozidla i pracovní stroje jsou konstruovány tak, aby svými parametry co nejlépe vyhověly požadavkům přímých uživatelů, ale musí plnit i podmínky stanovené legislativou. Požadavky jsou to rozmanité, a i kdyby byla pozornost soustředěna jen na ty výkonové (které v minulosti převažovaly), pak jejich spektrum s ohledem na rozdílné provozní cykly je tak široké, že provoz pohonných jednotek a převodových systémů nelze definovat jedním (optimálním) pracovním bodem. Nelze předpokládat identitu souřadnic extrémů několika různých kritérií, úlohou konstruktérů je hledání takových řešení, která poskytnou oblasti provozních parametrů, v nichž požadavkům vyhoví. LITERATURA [1] Naunheimer H., Bertsche J., Ryborz J., Novak W. B. Automotive transmissions. Springer-Verlag, 2011, ISBN [2] BRABEC, P. HO, H. H. MALÝ, M. - VOŽENÍLEK, R.: Hnací ústrojí s hydromotory. In: 18th International Conference on Hydraulics and Pneumatics. Praha ISBN [3] BRABEC, P. MALÝ, M. VOŽENÍLEK, R.: Vehicle Model for Simulation of Driving Properties. In: Polytransport Systems, IV Russian Scientific and Technical Conference. Krasnoyarsk ISBN

244 XLVIII. INTERNATIONAL SCIENTIFIC CONFERENCE OF CZECH AND SLOVAK UNIVERSITIES DEPARTMENTS AND INSTITUTIONS DEALING WITH THE RESEARCH OF COMBUSTION ENGINES SEPTEMBER , 2017 KLÁŠTER HRADIŠTĚ NAD JIZEROU, CZECH REPUBLIC TECHNICAL UNIVERSITY OF LIBEREC DEPARTMENT OF VEHICLES AND ENGINES VLIV VYBRANÝCH BIOPALIV NA VÝKONOVÉ PARAMETRY MOTORU ŠKODA ROOMSTER 1.4 TDI Jakub Mařík1, Martin Kotek2, Petr Jindra3, David Marčev4, Vladimír Hönig5 Abstract S rostoucí silniční dopravou po celém světě se biopaliva dostávají do popředí veřejného zájmu. Biopaliva umožňují zcela nebo částečně nahradit stávající neobnovitelné energetické zdroje ropného původu. Využívání biopaliv je obzvláště výhodné, pokud jde o snížení závislosti na fosilních palivech (ropě) a následně snížení negativních dopadů na životní prostředí. Podle čl. 3 směrnice 2009/28/ES musí každý členský stát zajistit, aby energetický podíl z obnovitelných zdrojů ve všech druzích dopravy dosáhl do roku 2020 nejméně 10 % konečné spotřeby energie v dopravě. Přidávání biopaliv je jednou z metod, které mají členské státy k dispozici, aby splnily tento cíl. Povinný procentuální podíl biopaliv, jakožto zdroj energie z obnovitelných zdrojů, bude pravděpodobně stále obtížnější dosáhnout udržitelným způsobem, neboť celková poptávka po energii v dopravním segmentu nadále stoupá. Tento článek analyzuje a porovnává vybraná biopaliva, jejich chemické vlastnosti a jejich vliv na výkonové parametry motoru. Experimenty byly provedeny na vozidle Škoda Roomster 1.4 TDI. Na válcovém dynamometru byly zjišťovány výkonové parametry při provozu na různé druhy směsných biopaliv (motorová nafta, HVO a butanol). Dále byly měřeny chemické vlastnosti těchto vybraných biopaliv, jmenovitě hustota, kinematická viskozita, cetanové číslo, bod vzplanutí a vlastnosti paliv při nízkých teplotách. 1. INTRODUCTION Spotřeba fosilních paliv stále roste zároveň s rostoucím prodejem naftových aut v Evropě. Emise, které tato vozidla produkují, mají negativní vliv na lidské zdraví a životní prostředí (EAMA, 2015; Dockery et al., 1992; Pourazar et al., 2004; Goyal et Ing. Jakub Mařík, Ph.D. Česká zemědělská univerzita v Praze, Katedra vozidel a pozemní dopravy, marikj@tf.czu.cz 2 Ing. Martin Kotek, Ph.D. Česká zemědělská univerzita v Praze, Katedra vozidel a pozemní dopravy, kotekm@oikt.czu.cz 3 Ing. Petr Jindra Česká zemědělská univerzita v Praze, Katedra vozidel a pozemní dopravy, jindra.petr2@gmail.com 4 Ing. David Marčev, Ph.D. Česká zemědělská univerzita v Praze, Katedra vozidel a pozemní dopravy, marcev@tf.czu.cz 5 doc.ing. Vladimír Hönig, Ph.D. Česká zemědělská univerzita v Praze, Katedra chemie, honig@af.czu.cz 1 243

245 al., 2010; Jacobson et al., 2001; Koch et al., 2011). Evropská Unie reguluje produkci emisí zaváděním přísných limitů. Evropská Unie take zavedla náročnější jízdní cyklus cycle WLTP (Worldwide Harmonized Light Vehicles Test Procedure). Cílem regulací, které Evropská Unie zavádí je zvýšit podíl obnovitelné energie na 10 % do roku 2020 a aby se snížila produkce skleníkových plynů, především CO 2 (Directive 28/2009/CE). Směsná biopaliva jsou jednou z metod členských států, aby splnili tento cíl. Obrázek 1: Škoda Roomster 1.4 TDI na válcové zkušebně Bioethanol a methyl ester řepky olejné (MEŘO) jsou nejčastěji využívaná biopaliva v Evropě. Rostlinný olej může být využit i jiným způsobem než transesterifikací, kterou vzniká MEŘO. Čistý rostlinný olej může být přidán do nafty v poměru 20 % olej a 80 % nafta a může být spalován bez nutnosti úprav motoru. (Yilmaz & Morton, 2011), některé zdoje uvádějí i 30% oleje (Masjuki et al., 2001). Další možnosti využití rostlinného oleje se neobejde bez úprav motoru, protože je nutné olej předehřívat ke snížení viskozity (Pexa, 2014). Mezi hlavní nevýhody MEŘA patří vysoká cena vstupní suroviny a nízká skladovací a oxidační stabilita. Ve srovnání s čistou naftou má FAME obecně nižší hmotnostní výhřevnost, vyšší hustotu a vyšší viskozitu (Hönig, 2014; Rahman, 2014) Dalším způsobem zpracování rostlinného oleje je hydrogenace (hydrogenovaný rostlinný olej - HVO) (Stumborg 1996, Kuronen, 2007). K výrobě HVO může být kromě čistého rostlinného oleje - použito i jiné druhy triglyceridů bohatých odpadních materiálů (živočišné tuky, použitý kuchyňský olej atd.) (Huber, 2007). Výroba HVO z těchto odpadních tuků je méně náročná než výroba bionafty. Proto je HVO někdy označován jako biopalivo druhé generace. Produkce HVO je založena na reakci triglyceridů s vodíkem (da Rocha Filho, 1993). Je složena z parafinových uhlovodíků s lineárním řetězcem, bez aromátů, kyslíku a síry (Aatola, 2008). HVO má nižší hustotu ve srovnání s dieselovým palivem a srovnatelnou kalorickou hodnotu (Lapuerta, 2011). HVO má také nízký obsah síry, což vede ke snížení emisí SO2, NOx, částic a aromátů (Kuronen, 2007). HVO má nižší spotřebu paliva, nižší 244

246 ztrátu výkonu a vyšší účinnost motoru než běžná bio nafta (Duckhan, 2014, Kim, 2014). Cílem této studie je porovnání čisté nafty a vybraných biopaliv (HVO a D50H30B20), jejich chemických vlastností a jejich vlivu na výkonnostní parametry a opacitu motoru. 2. MATERIÁL A METODY V tomto pokusu bylo použito vozidlo Škoda Roomster 1.4 TDI s 3 válcovým turbodmychadlem s kompresním vznětovým motorem (CI). Podrobné technické specifikace vozu jsou shrnuty v tabulce 1. SPALOVACÍ MOTOR Konstrukce Přeplňovaný vznětový motor Vstřikovací systém unit injector system Počet válců a ventilů 3 válce v řadě, 6 ventilů Palivo Objem válců Výkon nafta 1,422 ccm 59 kw při 4,000 ot./min Točivý moment 195 Nm při 2,200 ot./min Emisní norma EURO 4 Rok výroby 2006 Stav tachometru 102,000 km KAROSERIE Poh. hmotnost 1,240 kg Celková hmotnost 1,755 kg JÍZDNÍ VLASTNOSTI Max. rychlost 165 km h-1-1 Zrychlení km h 14,7 s Spotřeba paliva 5,1/3,76/4,26 (l/100 km) Tabulka 1: Technické parametry Škoda Roomster 1.4 TDI Výkonové parametry a opacita výfukových plynů byly měřeny u čisté motorové nafty, HVO a směsných biopaliv D50H30B20 (50% motorová nafta, 30% HVO a 20% butanol). U vybraných paliv byly měřeny následující chemické parametry: Hustota při 15 C dle EN ISO Kinematická viskozita při 40 C dle EN ISO CFPP - Bod zapouzdření za studena dle EN 116. Cetanové číslo podle EN ISO Cetanový index dle EN ISO Bod vzplanutí dle EN

247 V první fázi byly výkonnostní parametry vozidla měřeny na dynamometru, jehož technické parametry jsou uvedeny v tabulce 2. Zařízení Technické parametry DC Maximální brzdná síla 56 kw motorgenerátor Maximální brzdný moment 305 Nm Maximální rychlost ot/min Vířivý Maximální brzdný moment 125 kw dynamometr Maximální brzdný moment 478 Nm Maximální rychlost ot./min Tensometr Měřící rozsah: 2 kn Přesnost: 0,5% z měřícího rozsahu Nastavení Simulovaná hmotnost vozidla: 680 kg setrvačníků Přiřaditelné hmotnosti: 900 kg, 450 kg, 225 kg, 120kg, 112,5 kg. Tabulka 2: Parametry válcové zkušebny Ve druhé fázi experimentu byla změřena opacita výfukových plynů pro každé zkoušené palivo pomocí opacimetru Atal AT-605 (technická specifikace viz tabulka 3). Metoda měření kouře opacimetru byla v souladu se směrnicemi ECE R24 pro měření metodou volné akcelerace. Měřené veličiny Absorbční koeficient (k) Opacita (N) Teplota Otáčky Rozsah 0 16 m % C 400 2,000 min-1 2,001 9,999 min s Acceleration time RV = of reading value; *) in range 0.0 to 2.5 m-1; **) in range 2.5 to 4.0 m-1 Resolution Přesnost 0.01 m-1 ± 0.15 m-1*); ± 0.30 m-1**) 0.1% ± 2% absolute 1 C ± 2 C min ± 20 min-1 ± 2% RV 0.1 s ± 0.2 s Tabulka 3: Technické specifikace Opacimetru Atal AT

248 3. VÝSLEDKY A DISKUZE Hustota směsi (tabulka 4) je ovlivněna nižší hustotou HVO a butanolu ve srovnání s dieselovým palivem. Jelikož je palivo dávkováno objemově, můžeme očekávat malou ztrátu výkonu a zvýšení spotřeby vzhledem k poměru butanolu. Biobutanol jako alkohol s krátkým uhlovodíkovým řetězcem má nižší výhřevnost ve srovnání s dieselovým palivem a HVO, které se ve směsi projeví. Čistý butanol má také velmi nízké cetánové číslo ve srovnání s dieselovým palivem, ale přítomnost HVO pozitivně kompenzuje tuto hodnotu. Množství cetánového čísla HVO udávané výrobcem je> 70,0 jednotek. Proto působí HVO ve směsi jako vhodná přísada, aby zvýšila cetanové číslo. Butanol i HVO mají velmi dobré vlastnosti při nízkých teplotách, které jsou charakterizovány hodnotami bodu zákalu a ztrátou filtrovatelnosti (CFPP). Hodnota CFPP čistého motorového nafty dosáhla -17 C a parametru bodu zákalu byl -5 C. Použití biobutanolu při nízkých teplotách se vyznačuje krystalizační teplotou, která je mnohem nižší než CFPP zimní motorové nafty. Hodnota CFPP čistého HVO obvykle odpovídá hodnotě zimní motorové nafty a nejsou zde žádné komplikace při nízkých teplotách, jako v případě konvenční bionafty ve formě FAME. Parametr Cloud Point je v případě čistého HVO velmi blízko CFPP. Testovaná směs byla také stabilní při velmi nízkých teplotách, a proto nevyžaduje použití stabilizátorů. Toto je velmi důležité zjištění, protože komerčně používaný bioethanol vykazuje obtížnou stabilitu ve směsích při nízkých teplotách. Proto se zdá, že použití biobutanolu je upřednostňované, což má také za následek, že biobutanol ve srovnání s bioetanolem není hydroskopický. Přísady biobutanolu mají významný vliv na hodnotu bodu vzplanutí. Zatímco motorová nafta je charakterizována jako III. Třída nebezpečnosti hořlavý produkt s hodnotou bodu vzplanutí > 55 C, hodnota bodu vzplanutí směsi byla pouze 35 C. Bod vzplanutí se však používá pro zařazení do tříd nebezpečnosti pro kapaliny. Výsledná směs je však charakterizována jako II. Třídy hoření v hořlavém produktu, takové snížení bodu vzplanutí není pro proces spalování relevantní. Tento parametr je nutno vzít v úvahu, aby bylo zajištěno bezpečné skladování a manipulace se směsí. Diesel Hustota při 15 C Kinematick8 viskozita při 40 C Bod vzplanutí Bod tuhnutí Ztráta filtrovatelnosti Cetanové číslo Cetanový index Diesel 50% + HVO 30% + BUT 20% C C C > kg m-3 mm2 s- HVO 1 Tabulka 4: Naměřené parametry směsných paliv 247

249 Výsledky výkonových parametrů pro každé palivo jsou uvedeny na obr. 2. Je zřejmé, že nejlepší výkonnostní parametry byly dosaženy čistým naftovým palivem, zatímco ostatní biopaliva dosáhly trochu nižší hodnoty výkonu a točivého momentu. Jak je uvedeno v tabulce 5, nafta dosáhla 60 kw a točivého momentu 188 Nm, po němž následoval HVO, kde výkon byl mírně nižší s hodnotou 57 kw a točivým momentem 184 Nm. Nejnižší výkonnostní parametry byly měřeny smíšeným palivem D50H30B20, kde výkon dosáhl 56 kw a točivým momentem 180 Nm. Maximální hodnoty Výkon Točivý (kw) moment (Nm) DIESEL HVO D50H30B Palivo Tabulka 5: Maximální hodnoty výkonových parametrů Obrázek 2: Výkonové parametry motoru pro jednotlivá paliva Výsledek opacity výfukových plynů je ukázán na obr. 3. Nejvyšší opacita byla detekována čistým naftovým palivem 36,2 %. HVO vykazuje nepatrnou opacitu Pod 26,8% a nejnižší opacita byla měřena na D50H30B20 - pouhých 15,9 %. Výsledky studie (Pexa, 2016) rovněž ukazují, že biopaliva jsou významně ovlivněna znečišťujícími látkami spalovacími motory. Zvláště biopaliva obsahující HVO nebo butanol vykazují nižší kouřivost a až o 40% nižší produkci NOx. 248

250 Obrázek 3: Kouřivost testovaných paliv 4. ZÁVĚR Cílem experimentu bylo analyzovat palivovou směs D50H30B20, která sestává z biopaliv o 50%. Vybraná biopaliva jsou v případě HVO a butanolu charakterizována jako biopaliva II. generace. Také v případě kombinace HVO a biobutanolu je možné použít materiály z různých obnovitelných zdrojů. Studie ukázala, že HVO vykazuje o 26% méně opacity ve srovnání s motorovou naftou. Míchané palivo D50H30B20 vykazuje dokonce nižší opacitu o 56% než čistá nafta. Biopaliva vykazovala mírně nižší výkonové parametry než čistá nafta, avšak v závislosti na produkci emisních parametrů je rozdíl výkonu minimální. Potenciál směsi je také v maximalizaci využití alkoholu, který je vhodnější pro benzínové motory, ale z pohledu environmentální udržitelnosti, produkce škodlivých emisí a biodiverzity je dlouhodobě využitelným zdrojem energie. POUŽITÁ LITERATURA [1] Aatola, H., Larmi, M., Sarjovaara, T., Mikkonen, S Hydrotreated vegetable oil (HVO) as a renewable diesel fuel: trade off between NOx, particular emission, and fuel consumption of a heavy duty engine. SAE Technical Paper, [2] da Rocha Filho, DN, Brodzki D, Djéga-Mariadassou G Formation of alkanes, alkylcykloalkanes and alkylbenzenes during the catalytic hydrocracking of vegetable oils. Fuel,72, pp [3] Directive 28/2009/CE of the European Parliament and of the Council [4] Dockery, D.W., Schwartz J, Spengler, J.D Air pollution and daily mortality: associations with particulates and acid aerosols. Environ Res, 59, pp [5] Duckhan, K.,Seonghwan, K., Sehun, O. No, S.-Y Engine performance and emission characteristics of hydrotreated vegetable oil in light duty diesel engines. Fuel [6] European Automobile Manufacturers Association Share of Diesel in new Passenger Cars. ACEA. Available at: 249

251 [7] Goyal, P., Jaiswal, N., Kumar, A., Dadoo, J.K., Dwarakanath, M Air quality impact assessment of NOx and PM due to diesel vehicles in Delhi. Transp Res Part D: Trans and Environ, 15, pp [8] Hönig, V., Hromádko, J Possibilities of using vegetable oil to power diesel engines as well as their impact on engine oil. Agronomy Research, 12(2), pp [9] Huber, G.W., O Connor, P., Corma, A Processing biomass in conventional oilrefineries: production of high quality diesel by hydrotreating vegetable oils in heavy vacuum oil mixtures. Appl Catal A, 329, pp, [10] Jacobson, M.Z Global direct radioactive forcing due to multicomponent anthropogenic and naturals aerosols. J of Geophys Res,106, pp [11] Koch, D Transport and direct radiative forcing of carbonaceous and sulphate aerosols in the GSISS GCM. J of Geophys Res, 106:203, pp [12] Kuronen, M, Mikkonen, S., Aakko, P., Murtonen, T Hydrotreated vegetable oil as fuel for heavy duty diesel engines. SAE Technical Paper, [13] Lapuerta, M., Villajos, M., Agudelo, J.R., Boehman, A.L Key properties and blending strategies of hydrotreated vegetable oil as biofuel for diesel engines. Fuel Process Technology, 92, pp [14] Masjuki, H.H., Kalam, M.A., Maleque, M.A., Kubo, A., Nonaka, T Performance, emissions and wear characteristics of an indirect injection diesel engine using coconut oil blended fuel. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, 215(3), pp [15] Pexa, M. & Mařík,J The Impact of biofuels and technical condition to its smoke Zetor 8641 Forterra. Agronomy Research, 12(2), pp [16] Pexa, M., Čedík, J., Pražan, R. Smoke and NOx emission of combustion engine using biofuels Agronomy Research, 14(2), pp [17] Pourazar, J, Frew, A.J., Blomberg, A., Helleday, R., Kelly, F.J., Wilson, S Diesel exhaust exposure enhances the expression of IL-13 in the bronchial epithelium of healthy subjects. Respir Med, 98, pp [18] Stumborg, M., Wong, A., Hogan, E Hydroprocessed vegetable oils for diesel fuel improvement. Bioresour Technology, 56, pp [19] Yilmaz, N. & Morton, B Effects of preheating vegetable oils on performance and emission characteristics of two diesel engines. Biomass and Bioenergy, 35(5), pp ACKNOWLEDGEMENT Paper was created with the grant support CZU 2016:31150/1312/ Analysis of the impact of biofuels on the operating parameters of internal combustion engines. 250

252 SEZNAM AUTORŮ Jiří B Baran Peter...29 Barta Dalibor...29 Bazala Jiří...86 Beroun Stanislav Bolehovsky Ondrej Bortel Ivan Brabec Pavel , 238 Brezáni Miloš...29 Č Čučo Marián D Denk Petr Dittrich Aleš... 50, 191 Doleček Vít Drápal Lubomír F Fischer Oliver...86 G Gotfrýd Ondřej... 40, 151 H Hatschbach Petr Hébert Guillaume...86 Hönig Vladimír Hrdlička Martin...92 Hujo Ľubomír Hvězda Ch Chríbik Andrej J Jablonický Juraj Janoško Ivan Jindra Petr K Kollar Luboslav Kotek Martin Kuchar Peter L Labuda Róbert Lach Ján Laurin Josef M Macek Jan... 70, 111, 131 Malý Miroslav Marčev David Mařík Jakub Miklánek Ľubomír... 40, 151 Morkus Josef , 199 N Novotný Pavel P Páv Karel... 4 Pechout Martin

253 Pešek Daniel Polóni Marián...23 S Schmidt Jiří Scholz Celestýn...50 Skácel Jan Syrovátka Zbyněk...12 Š Škara Petr...61 T Takáts Michal... 12, 180 V Vávra Jiří... 12, 180 Vítek Oldřich Vojtíšek Michal , 222 Voženílek Robert , 232 Z Zvolský Tomáš

254 Název publikace: Sborník přednášek XLVIII. mezinárodní konference kateder a pracovišť spalovacích motorů českých a slovenských vysokých škol Autor publikace: Kolektiv autorů (obsah) Sborník sestavili a redakčně upravili: Pavel Brabec Aleš Dittrich Robert Voženílek Vydavatel: Technická univerzita v Liberci Povoleno: Rektorátem TU v Liberci čj. Re 33/17 dne Vyšlo: srpen 2017 Vydaní: prvé Počet výtisků: 64 Tiskárna: Vysokoškolský podnik Liberec, s.r.o., Studentská 1402/2, Liberec Číslo publikace: Tato publikace neprošla redakční ani jazykovou úpravou vydavatele. Tisk byl uskutečněn z podkladů autorů příspěvků ISBN

255