Modeling and simulation analysis on parallel hybrid air-fuel vehicle

Abstract

Based on the vehicle simulation software ADVISOR, the model of a parallel air-fuel hybrid vehicle was established, and the modeling of an air powered engine (APE), heat exchanger, braking air tank and control strategy were discussed in detail. Using the vehicle model, a hybrid vehicle refitted from a traditional diesel car was analyzed. The results show that for the New European Driving Cycle (NEDC), the Urban Dynamometer Driving Schedule (UDDS) and the Highway Fuel Economy Test (HWFET) driving cycle, the total reductions in fossil fuel consumption of the hybrid vehicle were 48.29%, 48.51% and 22.07%, respectively, and the emissions could be decreased greatly compared with the traditional diesel car, while the compressed air consumptions of the hybrid vehicle were 97.366, 85.292 and 56.358 kg/100 km, respectively. Using the diesel equivalent as the indicator of fuel economy, the hybrid vehicle could improve the fuel economy by 14.71% and 16.75% for the NEDC and the UDDS driving cycles and decrease by 5.04% for the HWFET driving cycle compared with the traditional car. The simulation model and analysis in this paper could act as the theoretical basis and research platform in optimizing the key components and control strategy of hybrid air-fuel vehicles.

Keywords

air powered engine / hybrid vehicle / internal combustion engine / simulation

Introduction

Research on the air powered vehicle (APV) is becoming hot because it does not have emissions, nor does it need fossil fuel [1-3]. However, due to the low efficiency of the air powered engine (APE) and low energy density of compressed air and liquid nitrogen, APV has the deficiency of short continual mileage [4,5]. Although the vehicle using traditional internal combustion engine (ICE) has long continual mileage, the ICE often operates in its low efficiency working area, and about 70%-80% of the energy from the combustion of the fuel cannot be converted to mechanical energy, instead, it is directed to other parts of the engine in the form of heat or emitted as exhaust. To counter this deficiency, the hybrid air-fuel vehicle (HAV) is proposed. One way to achieve air-fuel hybrid is to compress the air in the ICE cylinders into an air tank when slowing down (compression braking), and then use the high-pressure air in the air tank for subsequent acceleration [6-8]. Simulation results showed that this kind of HAV could improve the fuel economy by as much as 64% for urban driving and 12% for highway driving [9]. Guy Negre of the MDI company designed a bi-energy car incorporating gasoline and compressed air which could run at 96 km/h for more than 200 km with a 300-L air tank<FootNote>

Moteur Development International (MDI). http://www.theaircar.com/ [Accessed 13/03/07], 2007

</FootNote>. In order to optimize the running condition and recycle the exhaust flow of the ICE to propel the vehicle, a hybrid pneumatic power-system (HPPS) was developed by Huang and Tzeng [10]. The Scuderi Group introduced a split-cycle air hybrid engine with separate cylinders for compression and power [11].

A parallel HAV was introduced in this paper, which not only reduced fossil fuel consumption and emissions through optimizing the running condition of the ICE, shutting down the ICE at idle, achieving regenerative braking as the hybrid electric vehicle (HEV) does [12,13], but also improved the performances of the APE by heating the inlet compressed air or working ambient of the APE with the waste heat of the ICE [14]. Compared with HEV, this kind of HAV is cheap and does not require a battery, which eliminates the many disadvantages such as the excessively high costs, complex maintenance, long charging time and so on [15]. The model of the HAV is established first. Then using the model, the effects of fuel saving and emission reduction as well as the influence of recycling the exhaust heat of the ICE on the APE’s consumption of compressed air are analyzed.

System scheme

The system scheme of the parallel HAV, shown in Fig.1, has two power systems. One is the pneumatic power system, and the other is the ICE power system. The two power systems are coupled through a torque coupler. The function of the heat exchanger is to recycle the exhaust heat of the ICE to heat the inlet compressed air of the APE. A three-way valve is used to switch the air passages. For normal mode, the three-way valve connects the inlet valve of the APE with the stable pressure box. When the regenerative braking is activated, the three-way valve switches to connect the inlet valve of the APE with the braking air tank, and the compressed air would be pushed into the braking air tank through the inlet valve of the APE. Once the compressed air in the braking air tank is sufficient enough to drive the car, the inlet valve of the APE will again be connected with the braking air tank.

Modeling of HAV

Two main approaches, backward-facing approach and forward-facing approach, are usually used in vehicle simulation. Based on the software ADVISOR [16], a combined backward/forward approach is used, as shown in Fig. 2. Since the models of components such as vehicle, clutch, ICE, gearbox and so on have been established in ADVISOR, the modeling of the APE, the heat exchanger, the braking air tank and the control strategy will be detailed.

Modeling of APE

Because the APE has the characteristic of flexible control, it provides peak power or transient power while the ICE provides average power in the hybrid air-fuel power system. To satisfy the requirement of fast response to load adjustment, an electro-pneumatic inlet valve of the APE is developed. As shown in Fig. 3, the load of the APE could be easily regulated by adjusting the time of the electrical pulse applied to solenoid valve A and solenoid valve B.

It is becoming popular to use the variable mass thermodynamics theory to establish the model of the APE (thermodynamics model) [17,18]. Regarding the cylinder of the APE as a thermodynamic system, the modeling of the cylinder consists of three equations:

(1){\displaystyle \frac{\mathrm{d}{T}_{\mathrm{c}}}{\mathrm{d}\varphi }}={\displaystyle \frac{\mathrm{1}}{m_{\mathrm{c}}{C}_{\mathrm{v}}}}\left({\displaystyle \frac{\mathrm{d}{Q}_{\mathrm{w}}}{\mathrm{d}\varphi }}+p_{\mathrm{c}}{\displaystyle \frac{\mathrm{d}{V}_{\mathrm{c}}}{\mathrm{d}\varphi }}+h_{\mathrm{i}\mathrm{n}}{\displaystyle \frac{\mathrm{d}{m}_{\mathrm{E}}}{\mathrm{d}\varphi }}+h_{\mathrm{c}}{\displaystyle \frac{\mathrm{d}{m}_{\mathrm{A}}}{\mathrm{d}\varphi }}-u{\displaystyle \frac{\mathrm{d}{m}_{\mathrm{c}}}{\mathrm{d}\varphi }}\right),

(2){\displaystyle \frac{\mathrm{d}{m}_{\mathrm{c}}}{\mathrm{d}\varphi }}={\displaystyle \frac{\mathrm{d}{m}_{\mathrm{E}}}{\mathrm{d}\varphi }}+{\displaystyle \frac{\mathrm{d}{m}_{\mathrm{A}}}{\mathrm{d}\varphi }},

(3)p_{\mathrm{c}}{V}_{\mathrm{c}}=m_{\mathrm{c}}R{T}_{\mathrm{c}},

where, T_c, m_c, p_c, V_c, h_c and u_c are the temperature, mass, pressure, volume, specific enthalpy and specific internal energy of air enclosed by the APE cylinder, respectively; Q_w is the heat exchange between the air within the cylinder and the environment; h_in is the specific enthalpy of the air flows into the APE cylinder; R is the gas constant; C_v is the specific heat at constant volume; Φ is the crank angle; and m_E, and m_A are the mass of the air flows into the APE cylinder and flows out of the APE cylinder respectively, which can be expressed as

(4){\displaystyle \frac{\mathrm{d}{m}_{\mathrm{E},\mathrm{A}}}{\mathrm{d}\varphi }}={\mu }_{\mathrm{E},\mathrm{A}}{S}_{\mathrm{E},\mathrm{A}}{\psi }_{\mathrm{E},\mathrm{A}}\sqrt{p_{\mathrm{I}}{\rho }_{\mathrm{I}}},

where, variables with subscripts “E” or “A” indicate the inlet or outlet parameters of the APE cylinder, and it is noted that the three-way valve (Fig. 1) and the inlet valve of the APE (Fig. 3) are regarded as a single valve in calculating m_E and m_A; p_I and ρ_I are the pressure and density of the air on the upstream of the valves; μ is the coefficient of charge; S is the orifice area of the inlet and outlet valves, and y is the stream function associated with the pressure difference between the upstream and downstream valves.

According to the modeling approach of HAV, the torque should be the input of the APE model. However, the torque would be the output if the thermodynamics model of the APE were used. Besides, the thermodynamics model consists of several differential equations which are time consuming to solve. To deal with these problems, interpolation model of the APE is established in this paper using the calculation results and input data of the thermodynamics model above. As shown in Fig. 4, the SIMULINK model of the engrgy use of APE accepts the inputs of torque requested Trq_req and speed requested n_req by the driveline or the vehicle controller, the air temperature T_sta and air pressure p_sta of the stable pressure box, and the air temperature T_bra and air pressure p_bra of the braking air tank. The resulting air consumption flow q_con, the recycling air flow q_rec and the air temperature T_rec can be calculated via interpolation of the input air consumption data at the specified speed, specified load, specified air pressure and specified air temperature. The two 3D interpolation blocks are used to calculate the maximum driving torque and the maximum braking torque of the APE for the current inputs respectively.

Modeling of braking air tank

Regarding the air in the braking air tank as the ideal gas, the following equations can be used to model the braking air tank:

(5)\mathrm{d}{m}_{\mathrm{f}}=m_{\mathrm{i}\mathrm{n}\mathrm{i}}-m_{\mathrm{e}\mathrm{n}\mathrm{d}},

(6)\mathrm{d}{m}_{\mathrm{f}}{C}_{\mathrm{p}}{T}_{\mathrm{f}}=m_{\mathrm{i}\mathrm{n}\mathrm{i}}{C}_{\mathrm{v}}{T}_{\mathrm{i}\mathrm{n}\mathrm{i}}-m_{\mathrm{e}\mathrm{n}\mathrm{d}}{C}_{\mathrm{v}}{T}_{\mathrm{e}\mathrm{n}\mathrm{d}},

(7)p_{\mathrm{e}\mathrm{n}\mathrm{d}}{V}_{\mathrm{b}\mathrm{r}\mathrm{a}}=m_{\mathrm{e}\mathrm{n}\mathrm{d}}R{T}_{\mathrm{e}\mathrm{n}\mathrm{d}},

(8)p_{\mathrm{i}\mathrm{n}\mathrm{i}}{V}_{\mathrm{b}\mathrm{r}\mathrm{a}}=m_{\mathrm{i}\mathrm{n}\mathrm{i}}R{T}_{\mathrm{i}\mathrm{n}\mathrm{i}},

where, m_f is the mass of air flow in or out of the braking air tank; V_bra is the volume of the braking air tank; m_ini, T_ini and p_ini are the initial air mass, the temperature and the pressure of the braking air tank; m_end, T_end and p_end are the air mass, the temperature and the pressure of the braking air tank after charging or discharging; and C_p is the specific heat at constant pressure. Regarding t as the time, dm_f is equal to q_recdt and T_f is equal to T_rec during the charging process of the braking air tank. In the discharging process, dm_f is equal to q_condt and T_f is supposed to be equal to T_ini.

Modeling of heat exchanger

The method to model the heat exchanger is also interpolation. As shown in Fig. 5, the 3D interpolation model of the heat exchanger accepts the inputs of the exhaust gas flow and the temperature from the ICE model and the air consumption flow q_con from the APE model. The resulting inlet air temperature of the APE T_sta and the heat transfer of the heat exchanger can be calculated via interpolation of the inlet air temperature data, the heat transfer data and the exhaust gas temperature after the heat exchanger data at the specified exhaust gas flow, the specified exhaust temperature and the specified air consumption flow, where the resulting heat transfer of the heat exchanger is used for calculating the heat transfer efficiency of the heat exchanger.

The quasi-static model of the heat exchanger is established to acquire the 3D interpolation data. The heat exchanger can be divided into the hot side and the cold side, and the heat transfer of each side can be expressed as

(9)Q_{\mathrm{h}\mathrm{x}}={\displaystyle \frac{Nu{k}_{\mathrm{h}\mathrm{x}}}{{\mathrm{L}}_{\mathrm{h}\mathrm{x}}}}{A}_{\mathrm{h}\mathrm{x}}(T_{\mathrm{f}}-T_w),

where, Q_hx is the heat transfer; A_hx is the heat transfer area; k_hx is the heat transfer coefficient; L_hx is the reference length; T_f is the temperature of fluid; T_w is temperature of the heat exchanger wall; and Nu is the Nusselt number, which can be expressed as

(10)Nu=C_1·{\mathrm{Re}}_{\square }^{C_2}·{\mathrm{Pr}}_{\square }^{\frac{1}{3}},

where, Re and Pr are the Reynolds number and the Prandtl number, respectively; and C₁ and C₂ are the empirical constants respectively, which can be generated from the experimental data.

Modeling of control strategy

It is assumed that the HAV is ICE dominated, which means that the ICE is considerably larger in power rating than the APE in this study. This allows one to focus most of the control to move on the ICE. The baseline static control strategy is used, which shuts down the ICE when its torque request or speed request falls below a limit. In order to use the exhaust heat of the ICE to improve the efficiency of the pneumatic power system, the APE will assist with the torque once the ICE is on. When the ICE and the APE are both on at the same time, their output power is allotted as

IF (P_{veh_r}×R_a)<P_{ape_max}

P_{IC_r}＝P_{veh_r}×(1-R_a);

ELSE

P_{IC_r}＝P_{veh_r}-P_{ape_max};

END

where, P_{veh_r} is the total power request for vehicle driving; P_{ape_max} is the maximum power the APE can output; P_{IC_r} is the power request of the ICE; and R_a is the rate of power request of the APE to the total power request.

Simulation setting

The HAV used in the simulation is refitted from a traditional diesel car equipped with a Volkswagen 67 kW 1.9 L turbo diesel engine and a 5-speed manual transmission, whose transmission ratios are 3.78, 2.12, 1.36, 0.97 and 0.76, respectively. Other parameters are listed in Table 1.

The total mass of the HAV is set to 1362 kg. In order to keep the total power of the HAV to be approximately the same as the traditional car, the HAV is equipped with a 56 kW diesel engine (Scaled from the Volkswagen 67 kW 1.9 L turbo diesel engine by torque). The APE is refitted from a 195 diesel ICE whose main parameters are listed in Table 2. The APE continues to use poppet outlet valve driven by the camshaft as traditional ICE does. Figure 6 shows the instantaneous effective areas of the inlet and outlet valve. Since the inlet valve is an electro-pneumatic valve, the curve of its instantaneous effective area varies in different intake durations.

The parameters of the heat exchanger are referred to an air-to-air intercooler of the ICE. Since the maximum exhaust air flow rate q_e of the ICE used in the HAV simulation analysis is 89.2 g/s, the experimental hot-side pressure drop Δp of the intercooler is less than 10 kPa, as shown in Fig. 7. In order to enhance the heat transfer, the parameters of the heat exchanger used in the simulation (Table 3) has been doubled compared with the parameters of the intercooler referred. Because the increasing of parameters would cause the pressure drop of the heat exchanger to become even smaller, the influence of the pressure drop of the heat exchanger on the performances of the ICE is ignored in the simulation.

The air pressure of the stable pressure box p_sta is set to 1.5 MPa and kept as constant. The ambient temperature and pressure are 293 K and 0.101 MPa, respectively. The initial air pressure, the initial air temperature and the volume of the braking air tank are set to 0.6 MPa, 400 K and 0.015 m³, respectively. To simplify the analysis, it is assumed that the air in the high-pressure air tank is sufficient enough to keep the stable pressure as 1.5 MPa throughout the simulation; and that the pressure reducing process of the pressure reducing valve shown in Fig. 1 is regarded as isothermal.

Simulation analysis

The NEDC, the UDDS and the HWFET driving cycles are selected to analyze the fuel economy and the effects of emission reduction of the HAV. Tables 4 and 5 show the comparison of diesel oil consumption (L_e) and emissions between the refitted HAV and the traditional diesel car. It can be seen that the L_e of the hybrid vehicle decreases by 48.29%, 48.51% and 22.07%, respectively, in comparison with the traditional car for the NEDC, the UDDS and the HWFET driving cycle. Because the baseline strategy in this paper is focused on fuel saving, the engine may work in its high NO_x emission area as shown in the torque (M) and speed (n) map (Fig. 8). For the NEDC and the UDDS driving cycles, the NO_x and PM emission still decrease greatly because of the shutting down of the ICE when its torque request or speed request falls below a limit. For the HWFET driving cycle, the NO_x emission only decreases by 4.17% for its short idle time and high average vehicle speed. It can also be seen in Fig. 8 (b) that the ICE is controlled in its high efficiency area.

Table 6 shows the simulation results of the HAV including the air consumption of the APE (g_a), the heat transfer and the heat transfer efficiency of the heat exchanger. It can be seen that the heat transfer efficiencies of the three driving cycles are all less than 30%, and the compressed air consumptions of the three driving cycles are somewhat high since the total efficiency of the pneumatic power system in this paper is relatively low. Especially for the NEDC and the UDDS driving cycle, because the vehicle needs frequent starts and stops, the APE should work alone frequently and cannot make good use of the exhaust heat of the ICE since the ICE is shut down or working in the state of low exhaust heat between whiles. Assuming the air pressure in the high pressure air tank is 30 MPa, the volumes of the air could be calculated as 252, 220 and 146 L/100 km, respectively, for the NEDC, the UDDS and the HWFET driving cycle.

The diesel equivalent (DE) is used to analyze the fuel economy. Since the volume of the braking air tank is only 0.015 m³, the energy changes of the air in the braking air tank between the beginning and the end of each driving cylces are ignored and DE is calculated as

(11)\mathrm{D}\mathrm{E}={\displaystyle \frac{g_{\mathrm{a}}R{T}_{\mathrm{0}}\ln (p_{\tan }/p_{\mathrm{0}})}{{\eta }_{\mathrm{c}}{L}_{\mathrm{v}}{\rho }_{\mathrm{d}}}}+L_{\mathrm{e}},

where, g_a is the air consumption as shown in Table 6; T₀ and p₀ are the ambient temperature and pressure respectively; p_tan is the air pressure in the high-pressure air tank; L_e is the diesel oil consumption as shown in Tables 4 and 5; η_c is the efficiency of air compressor; L_v and ρ_d are the lower heating value and density of diesel oil, respectively. Setting the value of T₀, p₀, p_tan, η_c, L_v and ρ_d to 293 K, 0.101 MPa, 30 MPa, 0.7, 850 g/L and 43 kJ/g, the diesel equivalent (DE) of the HAV can be calculated as 4.623, 4.527 and 4.159 L, which decreases by 14.71% and 16.75% for the NEDC and the UDDS driving cycle and increases by 5.04% for the HWFET driving cycle compared with the traditional car.

To analyze the effect of the exhaust heat of the ICE on the efficiency of the APE, two different cases come into being through retaining or removing the sub-model of the heat exchanger in the HAV model. For the case of removing the sub-model of the heat exchanger, the inlet air temperature of the APE is set to an ambient temperature of 293 K and kept as constant. Figure 9 shows the comparison of the instantaneous air consumption flow rate q_con between the two cases. According to the control strategy, the exhaust heat of the ICE can be recycled by the APE only when the ICE is on. The higher the vehicle speed is, the more exhaust heat of the ICE can be recycled, thus causing a larger difference between the two cases. The air consumption per 100 km for the case of removing the sub-model of the heat exchanger can be calculated by numerical integration, and the calculation results are 107.31 kg, 98.22 kg and 69.03 kg, respectively, for the NEDC, the UDDS and the HWFET driving cycle, which increase by 9.27%, 13.16% and 18.35%, respectively, compared with the case of retaining the sub-model of the heat exchanger.

Conclusions

The simulation model of the HAV developed in this paper can be used to conduct performance analysis, key components optimizing, control strategy designing, etc., of HAV. The simulation shows that:

1) The HAV can save fossil fuel greatly; the fuel economy could be improved for urban and suburban driving cycles; however, it could be slightly deteriorated for highway driving in comparison with the traditional vehicle.

2) Since the high pressure air can be produced using clean energy and the exhaust after-treatment for producing the high pressure air could be more effective in an air filling station than the one on a vehicle, running the HAV can reduce emission greatly compared with the traditional vehicle, especially for urban and suburban driving cycles.

3) Although the compressed air consumption of the HAV is a little higher, it can be reduced by optimizing the control strategy, achieving regenerative braking, developing an appropriate heat exchanger to recycle the exhaust of ICE using a high efficiency APE such as muti-stage APE and so on.

References

[1] Chen Ying,Liu Hao,Tao Guoliang. Simulation on the port timing of an air-powered engine. International Journal of Vehicle Design, 2005, 38(2): 259-273

[2] Liu Lin, Yu Xiaoli. Practicality study on air-powered vehicle. Frontiers of Energy and Power Engineering in China, 2008, 2(1): 14-19

[3] Pinglu CHEN, Xiaoli YU, Liu Lin. Simulation and experimental study of electro-pneumatic valve used in air-powered engine. Journal of Zhejiang University Science A, 2009, 10(3): 377-383

[4] Knowlen C, Mattick A T, Bruckner A P, Hertzberg A. High efficiency energy conversion systems for liquid nitrogen automobiles. SAE Transactions, 1998, 981898

[5] Liu H, Tao G L, Chen Y. Energy analysis on power system of air-powered vehicle. Journal of Zhejiang University (Engineering Science), 2006, 40(4): 694-698 (in Chinese)

[6] Schechter M M. Regenerative compression braking—a low cost alternative to electric hybrids. SAE Transactions, 2000, 2000-01-1025

[7] Higelin P, Vsile I, Charlet A, Chamaillard Y. Parametric optimization of a new hybrid pneumatic-combustion engine concept. International Journal of Engine Research, 2004, 5(2): 205-217

[8] Andersson M, Johansson B, Hultqvist A. An Air hybrid for high power absorption and discharge. SAE Transactions, 2005, 2005-01-2137

[9] Tai C, Tsao T, Levin M B, Barta G, Schechter M M. Using camless valvetrain for air hybrid optimization. SAE Transactions, 2003, 2003-01-0038

[10] Huang K D, Tzeng S C. Development of a hybrid pneumatic-power vehicle. Applied Energy, 2005, 80(1): 47-59

[11] Alexander D. Scuderi’s split-cycle solutions. Automotive Engineering International, 2006, 114(5): 24-26

[12] Liao G Y, Weber T R, Pfaff D P. Modelling and analysis of powertrain hybridization on all-wheel-drive sport utility vehicles. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, 2004, 218(10): 1125-1134

[13] Katrašnik T. Hybridization of powertrain and downsizing of IC engine: A way to reduce fuel consumption and pollutant emissions-Part 1. Energy Conversion and Management, 2007, 48(5): 1411-1423

[14] Yu X L, Yuan G J, Shen Y M, Liu Zhentao, Su Shichuan. Theoretical analysis of air-powered engine work cycle. Chinese Journal of Mechanical Engineering, 2002, 38(9): 118-122 (in Chinese)

[15] Kellaway M J. Hybrid buses-What their batteries really need to do. Journal of Power Sources, 2007, 168(1): 95-98

[16] Markel T, Brooker A, Hendricks T. ADVISOR: a systems analysis tool for advanced vehicle modeling. Journal of Power Sources, 2002, 110(2): 255-266

[17] Liu Hao, Tao Guoliang, Chen Ying. Research on the displacement and stroke-bore ratio of the air-powered engine. Journal of Engineering Design, 2006, 13(4): 381-384

[18] Nie X H, Yu X L, Hu J Q, Chen P L. Influence of inlet and outlet valve’s open timing on pneumatic engine and their optimal design. Journal of Engineering Design, 2009, 16(1): 16-20 (in Chinese)

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg