Advertisement

Automotive Innovation

, Volume 1, Issue 4, pp 320–330 | Cite as

Cost-efficient Thermal Management for a 48V Li-ion Battery in a Mild Hybrid Electric Vehicle

  • Chao YuEmail author
  • Guangji Ji
  • Chao Zhang
  • John Abbott
  • Mingshen Xu
  • Pieter Ramaekers
  • Jianxiang Lu
Article
  • 707 Downloads

Abstract

The 48V mild hybrid system is a cost-efficient solution for original equipment manufacturers to meet increasingly stringent fuel consumption requirements. However, hybrid functions such as auto-stop/start and brake regeneration are unavailable when a 48V battery is at very low temperature because of its limited charge and discharge capability. Therefore, it is important to develop cost-efficient thermal management to warm-up the battery of a 48V mild hybrid electric vehicle (HEV) to recover hybrid functions quickly in cold climate. Following the model-based “V” process, we first define the requirements and then design different mechanisms to heat a 48V battery. Afterward, we build a 48V battery model in LMS AMESim and conduct co-simulation with simplified battery management system and hybrid control unit algorithms in MATLAB Simulink for analysis. Finally, we carry out a series of vehicle experiments at low temperature and observe the effect of heating to validate the design. Both simulation results and experimental data show that a cold 48V battery placed in a cabin with hot air can be heated effectively in the developed “Enhanced Generator Mode with 48V Battery” mode. The entire design is in a newly developed software that cyclically charges and discharges a 48V battery for quick warm-up in cold temperature without needing any additional hardware such as a heater, making it a cost-efficient solution for HEVs.

Keywords

48V Li-ion battery Thermal management Mild hybrid electric vehicle Battery modeling 

1 Introduction

After long deliberation, the official version of the “Measures for the Parallel Administration of the Average Fuel Consumption and New Energy Vehicle Credits of Passenger Vehicle Enterprises” was finally issued on September 27, 2017, and came into effect on April 1, 2018. It requires all original equipment manufacturers (OEMs) to meet targets for corporate average fuel consumption credits as well as NEV credits. The 48V mild hybrid system is a cost-efficient fuel-saving technology for OEMs because of easier integration with the existing ICE-based powertrain and lower cost compared with the high-voltage powertrain system [1, 2, 3, 4]. The 48V Li-ion battery in a mild hybrid electric vehicle (HEV) is important in supporting hybrid functions such as automatic engine stop/start, torque assist and brake regeneration. However, the performance of the battery is strongly dependent on its temperature [5, 6, 7]. Under cold conditions, the charge and discharge capability of the battery is very limited, which puts constraints on the hybrid functions or even makes them unavailable [8, 9]. To maintain the temperature of a Li-ion battery within the ideal range (typically 20–40 °C), a thermal management system is necessary [10, 11, 12]. There are mainly two types of thermal management systems: active and passive [13]. The fundamental difference between active and passive thermal management systems is whether power is required to achieve thermal energy flow. Active management requires a component that consumes energy input to achieve heat transfer to or from the battery. Passive systems rely solely on natural convection of the cooling or heating fluid (liquid or air) moving slowly because of temperature gradients to achieve heat transfer [14]. For the battery in a 48V mild HEV, passive thermal management is preferred because of the lower cost and effort in system integration [15].

This study explores a cost-efficient solution for 48V battery thermal management in mild HEV applications. The remainder of this paper is organized as follows. Section 2 presents the 48V mild HEV architecture. Sections 3 and 4 define the requirements and functions. Section 5 presents the battery model and its confirmation. Sections 6 and 7 present the simulation, experimental results and analysis. Finally, Sect. 8 draws the conclusion and outlines future work.

2 Powertrain System Architecture

The mild hybrid system in this study consists of a 48V belt starter generator (BSG), a 48V Li-ion battery pack and an accessory power module (APM), shown in Fig. 1. The 48V inductive BSG is located at the P0 position, which has a peak motoring power of around 9 kW and peak generating power of around 12 kW. The 48V LiFePo4 battery pack is placed under the front passenger seat to use the cabin air for heating and cooling. It has 14 serial cells and provides a nominal voltage between 42 and 56 V with a capacity of 6 Ah. The APM is a bi-directional converter that first pre-charges the 48V DC bus using the 12V battery after key crank and then switches to buck mode to charge the battery and support the 12 V network in the vehicle.
Fig. 1

48V mild hybrid powertrain system architecture

The main functions of the 48V mild hybrid system include engine stop/start, brake regeneration and torque assist. In this architecture, a 12V starter controlled by an engine control module (ECM) is necessary and is always used for initial or cold start considering start performance consistency between cold and normal temperature. To achieve better start performance, the 48V battery and 48V BSG are used for auto-stop/start if available. The reason not to use the 48V BSG for initial start is the low capability of the 48V Li-ion battery compared with a 12V battery (such as an AGM) at low temperature. For example, under − 30 °C the engine cannot be started by a 48V Li-ion battery because the minimum required power for cold cranking is typically higher than 2 kW, but the 48V Li-ion battery discharge capability is only around 1 kW at the same temperature. Figure 2 shows the 48V battery charging and discharging power limits versus battery temperature at 50% state of charge (SOC). We see that the 48V battery does not charge or discharge below − 30 °C, then provides more than 7 kW of discharge power and 3 kW of charging power as the temperature rises to 0 °C, and finally reaches its maximum capability at above 30 °C.
Fig. 2

48V battery charge/discharge power limits versus battery temperature

3 Requirements for the Definition of 48V Battery Warm-Up

The first phase of system development is to define requirements. Actually, there are lots of requirements for the 48V system to meet, which are grouped and managed on the basis of use cases. A use case describes a specific operation of the system to achieve a goal, which is to heat the 48V battery at cold temperature in this study. There are many use cases in the 48V hybrid system, such as “start the engine” and “regenerate brake energy” which trace to corresponding requirements. Table 1 lists the main requirements for 48V battery warm-up. To meet the requirements in this use case, functions were designed that will be discussed in the next section.
Table 1

Requirements for 48V battery warm-up

Requirement ID

Requirement content

01

The 48V battery shall be protected from being overcharged or over discharged during vehicle operation

02

48V battery contactor shall keep open state at initial start if the 48V battery cannot be charged or discharged and shall be closed when its capability is recovered during vehicle operation

03

The 48V battery SOC shall be maintained within the designed range (e.g., 20–80%) during vehicle operation to enlarge 48V battery lifespan

04

The actual power of the 48V battery shall be maintained within its charging and discharging power limits based on SOC, voltage and temperature

05

48V mild HEV shall be able to operate at low temperature (down to − 30 °C) and 12V battery SOC shall be maintained within the designed range during vehicle operation

06

The 48V battery shall be heated to recover its capability in order to support hybrid functionalities (e.g., BSG stop start, brake regeneration) within 1 h at low temperature (down to − 30 °C)

4 Function and Logic Design

To heat the 48V battery, we designed three functions and their control logic, shown in Figs. 3 and 4, respectively. In the function “Generator Mode without 48V Battery” shown in Fig. 3 case 1 and Fig. 4a, the battery is unavailable at very low temperature (such as − 35 °C) with no charging or discharging capability, and the battery contactor is not allowed to close. The BSG is controlled as a generator to maintain 48V bus voltage and support APM power consumption. In this case, the 48V battery is heated by hot cabin air after heating is turned on by the user. In the function “Generator Mode” shown in Fig. 3 case 2 and Fig. 4b, the 48V battery is at very low temperature (such as − 25 °C) with very limited charging and discharging capability, and the battery contactor is allowed to close to stabilize the 48V bus voltage. The BSG is also controlled as a generator to support APM power consumption and maintain the battery near zero charging and discharging power. In this case, the battery is mainly heated by hot cabin air and maybe heat generation due to a small current going through the battery in dynamic conditions because of BSG and APM control accuracy. In the function “Enhanced Generator Mode” shown in Fig. 3 case 3 and Fig. 4c, the battery is at low temperature (such as − 20 °C) with limited charging and discharging capability, and the battery contactor also closes. If the SOC is lower than a threshold value, then the BSG is controlled to support APM power consumption and charge the 48V battery at its maximum charging power. If the SOC is higher than a threshold value and the APM power is within the battery discharging power limit, then the BSG is controlled to zero torque and the APM discharges the battery. If the APM power exceeds the battery discharging power limit, then the BSG is controlled to compensate the gap and avoid over-discharging of the battery by the APM. The battery is discharged to its low SOC threshold and then charged again. This process runs cyclically to make use of battery self-heating to quickly warm-up the battery while still keeping its SOC within the proper window and not violating the power limits.
Fig. 3

Functions to warm-up 48V battery

Fig. 4

Control logic to warm-up 48V battery in three cases. a Control logic for case 1 generator mode without 48V battery, b control logic for case 2 generator mode with 48V battery, c control logic for case 3 enhanced generator mode with 48V battery

5 Modeling and Confirmation

5.1 48V Battery Modeling

We built physical models of a 48V battery to predict its electrical and thermal behavior under given conditions. The governing equation is derived from the energy and charge balance [16].

The electric model was developed by using the first-order RC equivalent circuit model [17], shown in Fig. 5. The equation for the electrical behavior of the battery is
$$U_{\text{t}} = U_{\text{ocv}} + U_{\text{ohm}} + U_{\text{p}}$$
(1)
where \(U_{\text{t}}\) is the terminal voltage, \(U_{\text{ocv}}\) is the open-circuit voltage, \(U_{\text{ohm}}\) is the voltage drop on ohmic internal resistance, and \(U_{\text{p}}\) is the polarization voltage drop.
Fig. 5

48V battery equivalent circuit

The thermal model was developed by using the heat transfer balance [18, 19]. The three contributions to the heat generated are given by the equation
$${\text{d}}h_{\text{gen}} = {\text{d}}h_{\text{omh}} + {\text{d}}h_{\text{p}} + {\text{d}}h_{s}$$
(2)
where \({\text{d}}h_{\text{omh}}\) is the heat generated by ohmic internal resistance, \({\text{d}}h_{\text{p}}\) is the heat generated by the polarization resistance, and \({\text{d}}h_{s}\) is the reversible entropic heat.
The heat generation from ohmic internal resistance is
$${\text{d}}h_{\text{omh}} = I^{2} \times R_{\text{omh}} \times {\text{d}}t$$
(3)
where \(I^{{}}\) is current, \(R_{\text{omh}}\) is ohmic internal resistance, and \(t\) is time.
The heat generation from polarization resistance is
$${\text{d}}h_{\text{p}} = I^{2} \times R_{\text{p}} \times {\text{d}}t$$
(4)
where \(R_{\text{p}}\) is polarization resistance.
The reversible entropic heat is
$${\text{d}}h_{s} = \left( {\frac{{{\text{d}}U_{\text{ocv}} }}{{{\text{d}}T}}} \right)_{\text{SOC}} \times I \times T$$
(5)
where \(U_{\text{ocv}}\) is the open-circuit voltage, \(T\) is the temperature (K), and \({\text{SOC}}\) is the state of charge.
The heat exchange is
$${\text{d}}h_{\text{exch}} = {\text{d}}h_{\text{conv}} + {\text{d}}h_{\text{rad}} + {\text{d}}h_{\text{cond}}$$
(6)
where \({\text{d}}h_{\text{conv}}\) is the heat convection, \({\text{d}}h_{\text{rad}}\) is the heat radiation, and \({\text{d}}h_{\text{cond}}\) is the heat conduction.
The cell temperature is calculated by the equation
$$\frac{{{\text{d}}T}}{{{\text{d}}t}} = \frac{{{\text{d}}h_{\text{exch}} + {\text{d}}h_{\text{gen}} }}{{m_{\text{cell}} \times C_{p} }}$$
(7)
where \(m_{\text{cell}}\) is the cell mass and \(C_{p}\) is the heat capacity of the cell.
Figure 6 shows the integrated electrical and thermal model of the 48V battery in AMESim. In the model, cells 4 and 11 are temperature measuring points, which are at the same locations as the real temperature sensors in the 48V battery. The model has three inputs, which are cabin air temperature, cabin air flow rate and input current. The main outputs of the model are cell temperature, cell voltage and heat generation.
Fig. 6

48V battery model in AMESim

5.2 48V Battery Model Confirmation

We confirmed both the electric and thermal battery models with real vehicle experimental data under standardized driving cycles and real-world driving conditions.

5.2.1 Electrical Model Confirmation

A charging and discharging test was conducted on a 48V mild HEV. We recorded the maximum and minimum cell voltage profiles under the battery current measured by the battery management system (BMS) and then compared them with the simulation data from the electrical battery model using the same current load as the input. The results in Fig. 7 show a good agreement between the simulation and real test data with acceptable average error.
Fig. 7

48V electrical battery model confirmation

5.2.2 Thermal Model Confirmation

We confirmed the thermal battery model by testing a 48V mild HEV under free-drive conditions in the climate chamber. Figure 8a and b shows that the battery temperature calculated by the physical model follows the trend of real measurements under free-drive conditions in the cold chamber at − 20 °C and in the hot chamber at 40 °C. The difference between the model prediction and real measurement is less than 3 °C and acceptable.
Fig. 8

Confirmation of thermal model of 48V battery for two cases. a Test on dynamometer in cold chamber (− 20 °C) and b test on dynamometer in hot chamber (40 °C)

6 Simulation

We integrated the 48V battery plant model in AMESim with the simplified hybrid control unit (HCU) and BMS models in Simulink, as shown in Fig. 9. Using the integrated model, the functions designed in Sect. 4 were simulated with inputs and outputs shown in Fig. 10. One of the simulation inputs was the cabin air temperature profile measured in air heating mode shown in the top diagram in Fig. 10.
Fig. 9

Integrated system model in Simulink

Fig. 10

Simulation results for the 48V battery warm-up

The current in case 1, “Generator Mode without 48V battery,” is zero since the battery contactor is open. The current for case 2, “Generator Mode with 48V Battery,” is nearly but not zero because of control accuracy and delay with the contactor closed. For case 3, “Enhanced Generator Mode with 48V Battery,” the battery is charged (positive current) and then discharged (negative current) cyclically to heat it and maintain the SOC within the proper range. The predicted temperature rises of the 48V battery for the three cases are shown in the bottom diagram of Fig. 10, and the comparison can be seen in Table 2. The results show that the 48V battery warm-up for “Enhanced Generator Mode with 48V Battery” is much more significant than in the other two cases.
Table 2

Simulation results comparison for the 48V battery warm-up in 1 h

Case

Initial temp. (°C)

Final temp. (°C)

Warm-up rate (°C/min)

Generator mode without 48V battery

− 25

− 18

0.1

Generator mode with 48V battery

− 25

− 17

0.1

Enhanced generator mode with 48V battery

− 25

12

0.6

7 Experimental Validation

To validate the warm-up effect, a series of tests were conducted in the vehicle under cold temperature. The results are shown in Fig. 11.
Fig. 11

Experimental results for the 48V battery warm-up. a Generator mode without 48V battery, cabin heating active, b generator mode with 48V battery, cabin heating inactive, c enhanced generator mode with 48V battery, cabin heating inactive, d enhanced generator mode with 48V battery, cabin heating active

In the first test, the case “Generator Mode without 48V Battery,” the battery contactor was controlled open and the cabin heating was turned on by the user. As shown in Fig. 11a, the battery temperature increased from − 26 to − 20 °C in 60 min. In the second test, the case “Generator Mode with 48V Battery,” the contactor was controlled closed and the cabin heating was turned off by the user. As shown in Fig. 11b, the battery temperature increased from − 25 to − 20 °C in 60 min. In the third test, the case “Enhanced Generator Mode with 48V Battery,” the contactor was controlled closed and the cabin heating was turned off by the user. As shown in Fig. 10c, the battery temperature increased from − 25 to − 9 °C in 60 min. In the last test, the case “Enhanced Generator Mode with 48V Battery,” the contactor was controlled closed while the cabin heating was turned on by the user. As shown in Fig. 11d, the battery temperature increased from − 28 to − 11 °C in 33 min.

The comparison of results in Table 3 shows that the warm-up for “Enhanced Generator Mode with 48V Battery” with cabin heating is the best, followed by the case “Enhanced Generator Mode with 48V Battery” without cabin heating, and then the other two cases have almost equal effect.
Table 3

Experiment results comparison for the 48V battery warm-up

Test

Initial temp. (°C)

Final temp. (°C)

Duration (min)

Warm-up rate (°C/min)

Generator mode without 48V battery, cabin heating active

− 26

− 20

60

0.1

Generator mode with 48V battery, cabin heating inactive

− 25

− 20

60

0.1

Enhanced generator mode with 48V battery, cabin heating inactive

− 25

− 9

60

0.2

Enhanced generator mode with 48V battery, cabin heating active

− 28

− 11

33

0.5

During all these tests, the 48V battery SOC was maintained within the desired range (20–80%) and the actual power was maintained within its charging and discharging limits, meeting the requirements defined in Sect. 3. However, only the case “Enhanced Generator Mode with 48V Battery” with cabin heating meets the warm-up rate requirements. In this case, the average warm-up rate is 0.5 °C/min in the 33-min test, which shows that the 48V battery temperature rises from − 30 to 0 °C in 1 h, recovering more than 7 kW of discharging power and 3 kW of charging power. This completely supports auto-stop/start, which is the most perceptible feature for the customer, and partially supports torque assist and brake regeneration for better performance and fuel economy.

We find in Fig. 11d that the actual charging power follows the limits, which is similar to the behavior in Fig. 11c, but the actual discharging power (absolute value) is higher than that in Fig. 11c. The reason is mainly that the APM load increases by around 0.3 kW because of the cabin blower after cabin heating is turned on by the user, which increases the actual discharge of the 48V battery power and therefore heat generation.

The experimental results also agree very well with the simulation in Sect. 6, showing that the integrated system model can be trusted and used for optimization in future.

8 Conclusions and Outlook

To solve the problem of restricted hybrid function and diminished HEV performance caused by the limited capability of the 48V Li-ion battery in cold temperature, we have designed a cost-efficient passive thermal management solution. The design consists of a newly developed software function to cyclically charge and discharge a 48V battery for quick warm-up in cold temperature without needing any additional hardware such as a heater. We have built a 48V battery model, confirmed it for simulation analysis and conducted a series of experiments to validate it. The simulation and experimental results show that a cold 48V battery in the cabin of a mild HEV is effectively heated up in the developed “Enhanced Generator Mode with 48V Battery” during cabin heating, which meets the system requirements initially defined.

In the future, our optimization work will mainly focus on the following:
  1. (1)

    Exploring the relationship between SOC and such features as resistance and capability to determine the proper SOC range for “Enhanced Generator Mode with 48V Battery.” The SOC range will be determined on the basis of a trade-off between heat generation at lower SOC with higher battery resistance and heat generation at higher SOC with a larger discharge power limit.

     
  2. (2)

    Exploring the possibility of increasing BSG motoring torque with engine load levering to raise the actual battery discharge power in “Enhanced Generator Mode with 48V Battery” for more heat generation. However, the impact of engine load levering on fuel economy should be analyzed.

     

Notes

Acknowledgements

We thank our colleagues at the Pan Asia Technical Automotive Center Co. for support in building and co-simulating the models and providing test data, and our colleagues at Ricardo UK Co. for giving helpful advice.

References

  1. 1.
    Roscoe, S., Pascal, R., Tobias, S., et al.: Optimizing the architecture of 48-V mild hybrid. MTZ Worldw. 02, 26–31 (2018)Google Scholar
  2. 2.
    Jason, K., Matthew, H., James, S., et al.: HyBoost: an intelligently electrifies optimised downsized gasoline engine concept. In: Proceeding of the FISITA 2012 World Automotive Congress, Lecture Notes on Electrical Engineering 191, 483–496 (2013)Google Scholar
  3. 3.
    Ran, B., Victor, A., James, B.: Effect of 48 V mild hybrid system layout on powertrain system efficiency and its potential of fuel economy improvement, SAE Paper 2017-01-1175Google Scholar
  4. 4.
    Zifan, L., Andrej, I., Zoran, S. F.: Impacts of real-world driving and driver aggressiveness on fuel consumption of 48 V mild hybrid vehicle. SAE Paper 2016-01-1166Google Scholar
  5. 5.
    Yazhou, G., Maji, L., Jia, Z., et al.: Temperature characteristics of ternary-material lithium-ion battery for vehicle applications. SAE Paper 2016-01-1196Google Scholar
  6. 6.
    Noboru, S.: Thermal behavior analysis of lithium-ion batteries for electric and hybrid vehicles. J. Power Sources 99(1–2), 70–77 (2001)Google Scholar
  7. 7.
    Krishnashis, C., Pradip, M., David, S., et al.: Analysis of Li-Ion battery characteristics and thermal behavior. In: Proceeding of ASME 2013 Heat transfer Summer Conference: Heat Transfer in Energy Systems, 1, pp. V001T01A024 (2013)Google Scholar
  8. 8.
    Whsan, S., Josh, L., Michael, F., et al.: Impact of temperature on the A123 li-ion battery performance and hybrid electrical vehicle range. SAE Paper 2013-01-1521Google Scholar
  9. 9.
    Yan, J., Yancheng, Z., Chao-Yang, W.: Li-ion cell operation at low temperatures. J. Electrochem. Soc. 160(4), A636–A649 (2013)CrossRefGoogle Scholar
  10. 10.
    Kai, C., Zeyu, L., Yiming, C., et al.: Design of parallel air-cooled battery thermal management system through numerical study. Energies 10(10), 1677 (2017)CrossRefGoogle Scholar
  11. 11.
    Satyam, P., Scott, M., Roydon, F., et al.: Thermal management of lithium-ion pouch cell with indirect liquid cooling using dual cold plates approach. SAE Paper 2015-01-1184Google Scholar
  12. 12.
    Hiroyasu, B., Koji, K., Hideomi, K.: Battery heating system for electric vehicles. SAE Paper 2015-01-0248Google Scholar
  13. 13.
    Rami, S., Kizilel, R., Selman, J.R., et al.: Active vs. passive thermal management of high power lithium-ion packs limitation of temperature rise and uniformity of temperature distribution. J. Power Sources 182(2), 630–638 (2008)CrossRefGoogle Scholar
  14. 14.
    SAE J3073, SAE Surface Vehicle Information Report, 2016-05Google Scholar
  15. 15.
    Aziz, A., Rahimian, S.K., Blizanac, B., et al.: Exploring the opportunity space for high-power li-ion batteries in next-generation 48 V mild hybrid electric vehicle. SAE Paper 2017-01-1197Google Scholar
  16. 16.
    Venkatasailanathan, R., Paul, W.C.N., Sumitava, D., et al.: Modeling and simulation of lithium-ion batteries from a systems engineering perspective. J. Electrochem. Soc. 159(3), R31–R45 (2012)CrossRefGoogle Scholar
  17. 17.
    Xiaoqiang, Z., Weiping, Z., Geyang, L.: A review of li-ion battery equivalent circuit models. Trans. Electr. Electron. Mater. 17(6), 311–316 (2016)CrossRefGoogle Scholar
  18. 18.
    Yue, M., Ho, T., Marina, T.: Electro-thermal modeling of a lithium-ion battery system. SAE Paper 2010-01-2204Google Scholar
  19. 19.
    Meng, G., Ralph, E.W.: Thermal model for a li-ion battery pack with mixed parallel and series configuration. J. Electrochem. Soc. 158(10), A1166–A1176 (2011)CrossRefGoogle Scholar

Copyright information

© China Society of Automotive Engineers (China SAE) 2018

Authors and Affiliations

  1. 1.Pan Asia Technical Automotive Center Co. LtdShanghaiChina
  2. 2.Ricardo UK LtdShoreham-by-SeaUK

Personalised recommendations