Design and Analysis of Hybrid Power Systems with Variable Inertia Flywheel

Transcription

1 World Electric Vehicle Journal Vol. 4 - ISSN WEVA Page452 EVS25 Shenzhen, China, Nov 5-9, 21 Design and Analysis o Hybrid Power Systems with Variable Inertia Flywheel Hung-Kuo Su 1, Tyng Liu 2 1, 2 Department o Mechanical Engineering, National Taiwan University, No. 1, Sec. 4, Roosevelt Road, Taipei, 1617 Taiwan (R.O.C) 1 Graduate Student, R @ntu.edu.tw 2 Associate Proessor, tliu@ntu.edu.tw Abstract The purpose o this study is to analyze and to design or the hybrid power systems with variable inertia lywheel while the hybrid power system consists o 1 power plant, 1 variable inertia lywheel which is used as the energy storage device, a planetary gear set, and a set o actuators. First, the kinematic and kinetic equations o the hybrid power system are developed in order to establish the relationships o the speed and torque o all elements, and the speciic speed and torque o the output needed or the vehicle can be ound by using the sotware ADVISOR in various operation modes. Then, with the prescribed mechanical brake energy recovery system model and a control model, a comprehensive analysis can be achieved. Finally, various driving modes, such as ECE, ECE+EUDC and New York Bus driving mode are investigated in order to demonstrate the characteristics o the hybrid power system. The numerical results show and conclude the eectiveness o the variable inertia lywheel, and the improvement on the eiciency o hybrid power systems. Copyright Form o EVS25. Keywords: brake energy regeneration, hybrid power system, variable inertia lywheel 1 Introduction The researches on hybrid electric vehicles (HEV) are becoming more important in recent years, because the advantages o HEV are the signiicant reduction in uel consumption and low emissions. Today, the cost o state-o-the-art batteries suitable or vehicular application remains high, and the use o a motor generator and battery to transmit and retrieve the mechanical energy needs complicate systems or energy conversions [1]. In the search or a simple and practical hybrid system, the usage o a lywheel to retrieve energy in a hybrid system could be a possible design concept. The use o lywheels in hybrid vehicles has been proposed by many researchers [1-6]. Flybrid Systems Company, in 29, proposed a hybrid lywheel system, connected with drive shat on the power train by continuous variable transmission, and then connected with lywheel by a clutch [2]. Diego-Ayala used only a planetary gear set to connect the output with the lywheel [1] and R.M. van Drute used a planetary gear set and continuous variable transmission to connect with a lywheel [3-6]. In all these concepts, the inertia EVS25 World Battery,Hybrid and Fuel Cell ElectricVehicle Symposium 1

2 World Electric Vehicle Journal Vol. 4 - ISSN WEVA Page453 o the lywheel is constant. It is possible that i the variable inertia lywheel is used, the energy storage and retrieve might become more eective and eicient. The purpose o this study is to investigate the eectiveness and improvement o using a variable inertia lywheel in the hybrid power system or either electric vehicles or traditional internal combustion engine vehicles. 1.1 Hybrid Power Systems The motor and transmission are the general vehicle components. In the hybrid power system, a variable moment o inertia lywheel, a planetary gear set and a set o actuator are added on to the system, as shown in Figure 1, [1]. The angular velocity and torque o the system can be expressed as the ollowing equations: ( N31 1) ω Out ω Ring N31ω w = +, (1) 1.2 Equations o Hybrid Power Systems The perormance o the Hybrid Power System can be calculated by using equations (1) and (2). However, the losses at the gearbox should also be included by accounting or its eiciency ηgb. Depending on the direction o the energy low, the torque loss T w_loss at the sun and the torque loss T Ring_loss at the ring can be determined as the ollowing. ω > ω ω T < ω ω T T > ω ω ω ω < ω τring : τ w : τ Out = 1: N31:( N31 1), (2) where N 31 is -N 3 / N 1 and N 3 is gear number o sun gear, and N 1 is gear number o ring gear. In the equations, there are three variables, either or the angular velocities, or or the torques. Which means two variables should be assigned, or so called, controlled, and the system is said to be a two degree-o-reedom system. Figure 2: Operating logic or the System 1. I T O ut > and ω Out <(1-A)ω w, where A= -1 / (N 31-1), the vehicle will be accelerated by the lywheel. With respect to the require torque o vehicle, T Out, the torque o the lywheel and the torque o the ring gear can be expressed as: T = (1 A) T η, (3) w Out GB TRing = ATOutηGB. (4) The losses o torque in the transerring are: ( ) T = (1 Aτ ) η 1, (5) w _ loss Out GB ( ) T = AT η. (6) Ring _ loss Out GB 1 Figure 1: The Hybrid Power System Structure In the system, there are actuators used to make the ring gear o planetary gear decelerate, so that the brake kinetic energy will be transerred to lywheel. When lywheel speed is high enough to drive the vehicle, the actuators will make the ring gear o planetary gear decelerate, and the lywheel kinetic energy will be transerred to the wheel to drive the vehicle. The operating logic is shown as Figure I T Out < and ω Out >(1-A)ω w, the vehicle will be decelerated by the lywheel. The torque o the lywheel and the torque o the ring gear can be expressed as: T = (1 A) T η, (7) w Out GB TRing = ATOut ηgb. (8) The losses o the torques in the transerring are: ( ) Tw _ loss = (1 A) TOut 1 / ηgb 1, (9) EVS25 World Battery,Hybrid and Fuel Cell ElectricVehicle Symposium 2

3 World Electric Vehicle Journal Vol. 4 - ISSN WEVA Page454 ( ) TRing _ loss = ATOut 1/ ηgb 1. (1) For a given torque T w acting on the lywheel, the change o the angular velocity o the lywheel can be determined by the rotational equation o motion, which is: ( w w _ loss ) T T Δt Δ ω w =, (11) I w where ω w, I w, and T w_loss are the change in speed, the inertia, and the torque losses o the lywheel respectively. The kinetic energy o the lywheel and the ring gear stored or released are: E = T ω Δ t, (12) w w w E = T ω Δ t. Ring Ring Ring (13) The operating logic or the hybrid power system combined with ADVISOR which is numerical simulation sotware [ 7]. They can be arranged to become a modular numerical simulation mode. The ADVISOR give the required torque and necessary angular velocity to the operating logic or the hybrid power system, and the operating logic or the hybrid power system will then determine the motor to be On or O. The process is shown as in Figure 3. y y x z Figure 4: Variable Inertia Flywheel structure To adjust or to change the inertia o the lywheel, can be in many dierent ways. It will certainly aect the ability o the lywheel in storing or releasing kinetic energy. In this case, the inertia o variable inertia lywheel is expressed by the ollowing equation: I = rρ l ( l + 4r ) l m l s ( + ) x l r l k spring 2 ( kspring mω ) l x r I 2 (14) The mass o variable inertia lywheel is expressed by the ollowing equation: 2 ( ρ ) m = 4 2r l. (15) Ater calculation, the stored energy o a variable inertia lywheel is show n as Figure 5. ADVISOR Motor Transmission Motor(ON/OFF) T, Hybrid system Fly wheel Figure 3: modular numerical simulation mode 1.3 Variable Inertia Flywheel Structure In this study, the inerti a o the lywheel can be various with respect to the speed o the lywheel [8]. The changing o the position o the lumped-mass will change the inertia o the lywheel as Figure 4. Figure 5: variable inertia lywheel o stored energy map In Figure 5, the X-axis is the speed o the lywheel, the Y-axis is the inertia, and Z-axis shows how much energy could be restored. As shown, the relationship o the inertia and the changing o the inertia o the lywheel with the overall eiciency is implicit, and need some detailed numerical simulations to evaluate all design parameters and their interaction. Now i projection o Figure 5 is on X-Y lat and with a dierent unction, which the inertia is varied with the angular velocity, is selected, the way o changing inertia o the lywheel is shown in Figure 6. EVS25 World Battery,Hybrid and Fuel Cell ElectricVehicle Symposium 3

4 I World Electric Vehicle Journal Vol. 4 - ISSN WEVA Page rad/sec Figure 6: Dierent unction o the lywheel on the energy map 1.4 Test Conditions A 15,433 kg electric bus which the magnitudes o these weights, the speciications o test vehicle, the variable Inertia lywheel and battery are shown in 7. Appendix. The motor o 137(W) is shown as Figure 7. The X-axis is torque, the Y-axis is revolutions per minute (rpm), and Z-axis is the eiciency in transerring energy Figure 7: Motor characteristic Patent[8] linear w 2 w 3 w 4 w 1/2 w 1/3 w 1/4 w 1/5 w 1/6 I= Whereas the lywheel s losses map, including bearing riction and wind resistance losses, was developed on the basis o experimental tests perormed by Suzuki [9]. These values change dynamically as the simulation runs, whose main data is shown as Figure 8. To assess the hybrid vehicle under city driving conditions, three widely accepted driving cycles were selected: the ECE cycle [1], the ECE+EUDC cycle[1] and New York bus cycle [11], whose main data are shown in Table 1 and the driving cycles are shown as Figure 9~11. P (W) P ax P wind P wind +P ax rpm Figure 8: Energy losses at the lywheel kpha Fig. 9: ECE cycle[1] Fig. 1:ECE+EUDC cycle [1] kpha Figure 11: NewYork Bus cycle [11] EVS25 World Battery,Hybrid and Fuel Cell ElectricVehicle Symposium 4

5 World Electric Vehicle Journal Vol. 4 - ISSN WEVA Page456 Table 1: Data o driving cycle parameters ECE cycle ECE+EUDC NewYork cycle bus cycle Time 195 sec 1225 sec 6 sec Distance.99 km 1.93 km.99 km Idle time 64 sec 339 sec 44 sec No. o stops 3 times 13 times 11 times Max speed Average speed Max acceleration 1.6 m/sec m/sec m/sec 2 Max deceleration m/sec 2 m/sec 2 m/sec 2 Average acceleration.64 m/sec 2.54 m/sec m/sec 2 Average deceleration m/sec 2 m/sec 2 m/sec 2 2 Comparisons o various driving modes The SOC (State o Charge) o the vehicle beore testing the driving mode is SOC initial. When the vehicle inishes the driving mode, the SOC is SOC remained, and the SOC used, is SOC used. They are related as: SOC used =SOC initial -SOC remained. (16) For the comparison, the EI (Eiciency Improve) index is deined. Here we irst deine SOC basic = SOC used o a no-recharged EV ater complete the driving test as the baseline o comparison. The EV index is expressed as: EI= ( SOC -SOC ) basic SOC used used 1%. (17) The EI index is showing the improvement o the compared hybrid vehicle. Its physical meaning is that how arther, in percentage, the tested hybrid vehicle can go while using the same SOC as the no-recharged EV. Here we set up A 15,433 kg electric bus o the hybrid power systems with ixed inertia lywheel, and simulate to run the ECE cycle or one time. We set the ixed inertia lywheel o I=24.1. For the planetary gear set, A=.89, N 31 = - N 3 /N 1 = = - 11/89, gear number o sun gear is 11, gear number o ring gear is 89 and gear number o planetary gear is 39. We ocus on 45sec to 1 sec or the acceleration and the deceleration, as shown in Figure 12. The red line is the driving mode whose unit is, and the angular speed o the planetary gear set with the reduction ratio is the black line. The green line is the angular velocity o the lywheel. The blue line is the angular velocity o the ring gear. At 49.2 sec, the bus shall start to accelerate, and the lywheel and the ring gear will slow down. When the speed o lywheel and the speed o bus with the reduction ratio get very close, the actuator will then stop action, and the bus will be driven by original power source (here it is the motor) and so the lywheel is ree spinning now. When 85.2 sec, the bus shall decelerate, and the lywheel begins to accelerate and the ring gear begins slowly down. When the speed o lywheel and the speed o bus with the reduction ratio get very close, during the deceleration, again the actuator would stop action. Than the bus will be decelerated by its original brake system and lywheel will be ree spinning again. (rad / sec ) carrier/(1-a) ring Sun Figure 12: ECE cycle or each component in 45~1 sec For the hybrid power systems with ixed inertia lywheel, adjusting the reduction ratio and the operation speed, which is range o operation o the lywheel while driving the vehicle, will clearly aect the SOC remained. To compare the results o all variations, we have the simulation o running the ECE cycle or 585 sec, and results are shown in Figure 13. The X-axis is the operation speed o the lywheel, the Y-axis is the reduction ratio o the planetary gear set, and the Z-axis is SOC remained. We ind the reduction ratio o the planetary gear set,.89 and operation speed, 11 rad/sec will give the most SOC remained. These parameters will be set up or electric bus o the hybrid power systems with ixed inertia lywheel or urther discussion. Use the same way to search the better parameters or the hybrid power systems with variable inertia lywheel. EVS25 World Battery,Hybrid and Fuel Cell ElectricVehicle Symposium 5

6 World Electric Vehicle Journal Vol. 4 - ISSN WEVA Page457 We ind the set o parameters gives the most SOC remained is the inertia o the lywheel using the unction I(ω)= ω 1/6, as shown in Table 2. better than EV with ixed inertia lywheel (I = 24.1) and EV with brake electric energy regeneration EV EV W/RB EV W/FW EV W/VFW.6 Figure 13: Optimization o parameters map or ixed inertia lywheel Table 2 Optimization o parameters or variable inertia lywheel SOC remained SOC used EI % Patent[8] Linear ω ω ω ω 1/ ω 1/ ω 1/ ω 1/ ω 1/ I= EV W/FW I= EV W/RB EV With the best parameters, we simulate the bus to run the New York bus cycle or 1 minutes, and results are shown as Figure 14. The green line is the hybrid power systems with variable inertia lywheel, whose SOC remained is the greater than the other hybrid vehicles. The blue line is the hybrid power systems with ixed inertia lywheel, whose SOC remained is also greater than the EV with regenerative brake, which is the red line. Now we will simulate the hybrid power systems with a variable inertia lywheel to run the ECE cycle or 585 sec, ECE+EUDC or 3675 sec and the New York bus cycle or 1 minutes and compare the results, as shown in Table 3. There is an EI improvement, upto 33.7% or the EV 1/6 with the variable inertia lywheel (I = a + bω ). It is SOC Figure 14: SOC remained o vehicles using NewYork bus cycle Table 3: Simulation results EI % ECE ECE +EUDC New York bus Cycle EV without energy regeneration EV with brake electric energy regeneration EV with Flywheel (I = 24.1) EV with variable inertia Flywheel (I = a + bω 1/6 ) The increased EI is mainly bec ause o the nature o the New York bus cycle. In the New York bus cycle, there are accelerations and decelerations 11 times in 1 minutes, and during 9% o the cycle time, the vehicle speed is lower than 2 km/h, and about 67% o the cycle time the bus is stalled. Since the electric machine o EV will operate in low eiciency when the vehicle is at low speed and demands high torque, the electric machine used in EV bus or braking energy recovery will not be very suitable. It only has 3% improvement, even less that in the case o ECE cycle. For the New York bus cycle, when using the variable inertia lywheel (I = a + bω 1/6 ) to store and retrieve energy, the large amount o energy input and output can be coped with the vehicle speed, and provide better arrangement o the energy. While in low speed, the small inertia lywheel can be accelerated quickly to the operation speed and spinning into a large inertia lywheel, which is more suitable or kinetic energy EVS25 World Battery,Hybrid and Fuel Cell ElectricVehicle Symposium 6

7 World Electric Vehicle Journal Vol. 4 - ISSN WEVA Page458 storage and release. However, or the lywheel which is ixed inertia, the lywheel with large inertia will not speed up quick enough, and with small inertia just cannot store enough energy to drive the bus. 3 Conclusions From our simulations, the results are all avored or the HEV with variable inertia lywheel. The improvement, EI index can be rom 16.1% upto 33.8%, especially in case o New York bus cycle. Here are some summarized remarks. 1. Various types o hybrid power systems are investigated and compared numerically to evaluate the eectiveness o their energy recovery perormance in dierent driving patterns. 2. On the energy recovery systems, the electric bus using variable inertia lywheel in the New York bus Cycle simulation can improve up to 33.8% eiciency better than the EV bus without energy regeneration, and also better than other hybrid power system evaluated in this study. 3. The hybrid power system with variable inertia lywheel is satiable or a heavy-vehicle in the stop- and-go with severe acceleration-anddeceleration operation. With the aid o the variable inertia lywheel, the motor, or the power plant, can be operated more eiciently and energy recovered is also increased with the help o the variable inertia lywheel i the range o operation, size, and the adjustment o the inertia o the lywheel have been selected and designed properly. Acknowledgements The support o Giant Lion Know-How Co. Ltd. to this research is grateully appreciated, and the special thanks is to Mr. Tai-Her Yang, or his helpul and inspiring discussions during the course o this research. Reerences [1] Diego-Ayala, U., Martinez-Gonzalez, P., McGlashan, N., The mechanical hybrid vehicle: an investigation o a lywheel-based vehicular regenerative energy capture system, Proceedings o the Institution o Mechanical Engineers part D- Journal o Automobile Engineering, vol. 222(28), no. 11, pp [2] Flybrid Systems Company: [3] Druten, R.M., Transmission Design o The Zero Inertia Powertrain, Ph.D. dissertation(21), Tech. Univ. Eindhoven, Eindhoven, The Netherlands. [4] Serrarens, F. A., and Veldpaus, F. E., Control o A Flywheel Assisted Driveline with Continuously Variable Transmission, ASME Dynamic Systems, Measurement and Control, Vol. 125(23), no. 3, pp [5] Shen, S., and Veldpaus. F. E., Analysis and Control o a Flywheel Vehicular Powertrain, IEEE Transaction on Control Systems Technology, Vol. 12(24), no.5 pp [6] Shen, S., Vroemen, B., and Veldpaus. F. E., IdleStop and Go: a way improve uel economy, Vehicle System Dynamics, Vol. 44(26), no.6, pp [7] Markel, T., Brooker, A., Hendricks, T., Johnson, V., Kelly, K., Kramer, B., O Keee, M., Sprik, S., Wipke, K., ADVISOR: a systems analysis tool or advanced vehicle modeling, Journal o Power Sources, vol. 11(22), no. 2, pp [8] Yang, Tai-Her, Principle and structure o actively driving or centriugal linear ollowing dynamic lywheel eect, US Patent (1993) No. 5,269,197. [9] SuZuki, Y., Koyangi, A., Kobayashi, M., Shimada, R., Novel applications o the lywheel energy storage system, Renewable Energy, Vol. 3(25), pp [1] EEC Directive 9/C81/1, Emission test cycles or the certiication o duty vehicles in Europe, EEC emission cycles (199). [11] Michael P., Emissions Results rom Hybrid-Electric and Conventional Transit Buses in New York City, West Virginia University (2). [12] Su, Hung-Kuo, Design and Analysis o Hybrid Power Systems with Variable Inertia Flywheel, Master Thesis(21), National Taiwan University, Taipei, Taiwan. Authors Hung-Kuo Su Room 58, College o Engineering building, No. 1, Sec. 4, Roosevelt Road, Taipei, 1617 Taiwan (R.O.C) Tel: Fax: R @ntu.edu.tw He received his M.S. in Mechanical Engineering rom National Taiwan University, Taiwan, in 21. Tyng Liu Room 63, College o Engineering building, No. 1, Sec. 4, Roosevelt Road, Taipei, 1617 Taiwan (R.O.C) Tel: Fax: tliu@ntu.edu.tw He received his PhD in Mechanical Engineering rom Rutgers University, U.S.A., in Dr. Liu is an Associate Proessor o Department o Mechanical Engineering, National Taiwan University. EVS25 World Battery,Hybrid and Fuel Cell ElectricVehicle Symp osium 7

8 World Electric Vehicle Journal Vol. 4 - ISSN WEVA Page459 Appendix Weight o the test vehicle Speciications o Variable Inertia Flywheel parameters weight parameters Value Vehicle kg l.13 m Motor 255 kg ρ 79 kg/m 3 (137kw) Transmission 5 kg l.1 m (1 speed) Battery (Li 1135 kg R.159 m ion) passenger 1837 kg l x.5 m Total kg k spring 9 14 N/m Speciications o the test M kg vehicle parameters Value I 5.32~ (13.89times) g ( gravity) 9.81 m/s 2 Total mass kg ρ (Air Speciications o the 1.23 kg/m 3 density) battery μ (road parameters Value.85 adhesion) R r ( Rolling resistance).1 Temperature range ~ 25 ~ 41 ( ) Max ~ Weight Ampere or ~ 7.45 distribution.635:.365 charge and (Ahr) (ront : back) discharge C D (coeicient Eiciency.968 ~.99 o o charge ~ aerodynamic and drag) discharge A (Forward Voltage 11.7(V)~6(V) projected m range area) h (Center o.7747 m Weight o (kg) gravity one cell height) L (Wheelbase) m Average heat capacity 795 (J/kgK) Wr (Wheel.5 m Temperature 35 ( ) radius) o keeping Iw(Wheel Cooling o.32 (m 2 ) inertia) the surace Gear box One speed 1:1 Air low.7/12 Passengers weight 15/ = kg (re. SAE ) thickness o in Thermal coeicient (kg/s).1 (m) 15 (W/m 2 K) EVS25 World Battery,Hybrid and Fuel Cell ElectricVehicle Symposium 8