Combat Hybrid Power System Component Technologies: Technical Challenges and Research Priorities (2002)

The first step toward development of any hybrid electric powertrain is a detailed analysis of the drive requirements. These requirements include the maximum and continuous operational torque-speed envelopes and the time duration for these operations. A host of vehicle operational modes need to be taken into account to get these requirements. Peak torque is driven by acceleration (0.5g) and tow loads through soft soils. Typical design points for a military vehicle involve a tractive effort-to-gross vehicle weight (TE/GVW) ratio of ~1.1. The continuous torque requirement is determined by load up the steepest gradient (power is determined by the speed up the gradient) and the different drive cycles.

Once these data are available, the next step in the design process is the selection of the right type of drive; this includes the selection of the motor, the gear ratio, and the power electronics. For purposes of discussion, a hypothetical maximum torque-speed envelope is shown in Figure 2-1. The “base” speed is defined as the speed to which maximum torque is required (point A); beyond that, a constant horsepower is required. During the operation with high torque, the torque-to-ampere ratio, also known as torque constant, kt, should be maximized in order to minimize the current handled by the power electronics. This reduces the packaging and cooling of the power electronics and, incidentally, reduces the cost of the drive system also. Generally, this requirement translates into operating the motor with maximum flux until the base speed is reached, followed by flux weakening in the constant horsepower region.

For battery-powered vehicles, it is important to design the drive system to meet the high-speed operating point, because the voltage and the impedance of the motor limit the current and hence the torque produced at high speeds. The proper design of a traction drive system goes beyond optimizing for efficiency over a given drive cycle; in addition, several constraints must be met. A good optimization program incorporating all of these constraints is a necessary tool for this purpose.

Distributed-computer-controlled concepts and systems to integrate the management and control of all the critical elements of the powertrain are also important. The engine, electric motor, transmission, and battery must be coordinated and controlled in an optimal manner at every moment while the vehicle operates. This task can only be accomplished with distributed-computer-control system concepts being developed in current and future hybrid vehicles. This technology is needed to allow completely automatic operation of the powertrain. This chapter does not address such overall system

technologies, including continuously variable transmissions, but instead addresses only certain component technologies.

FIGURE 2-1 Motor torque-speed relationship.

There are several electric motors that can be considered for traction:

1. Induction motor,

2. Surface-mounted permanent magnet (PM) motor,

3. Internal PM motor,

4. PM synchronous reluctance motor, and

5. Switched reluctance motor.

Some of these motors are suitable for variable-speed traction drive, and some need to be modified to suit the requirements. For example, for a wide constant horsepower (CHP) region of operation, the induction motor is ideal, but it weighs more and has a lower full-load efficiency than other alternatives. The switched reluctance motor has a similar capability, but suffers some high torque pulsations and noise. Induction motors lend themselves to flux weakening more easily than permanent magnet motors, but they are less efficient than PM motors for full-load operation. For induction motor drives, the maximum voltage point can coincide with the base speed (point A in Figure 2-1), giving the maximum torque per ampere. For internal PM motors, the

maximum voltage (point B in Figure 2-1) is closer to the maximum speed point, making the torque per ampere lower than for induction motors.

PM motors have high full-load efficiency and they are lighter, but the flux-weakening region is no more than two times the high torque region. PM motors generally require higher currents to make them adaptable for this application having a wide constant horse power region. Consequently, the inverter must be larger, and cost and packaging become bigger issues. Surface-mounted and axial field PM machines also have excessive iron losses at high speeds, and their part-load efficiency at high speeds is lower. The axial field machines with arigap windings have very low armature reactance and are not suitable for traction drives. They are good for applications having little or no constant horse power region of operation.

PM synchronous reluctance motors with weak permanent magnets that depend on high reluctance torque offer wider CHP range but exhibit larger torque pulsations that are detrimental to the operation of the drive system. Thus, internal PM motors seem to be the most suitable of all the PM machines. In general, PM machines are better suited to multispeed gear boxes similar to automotive transmissions, while induction motors can be operated with a single speed reducer.

Wheel-hub motors have been proposed for military vehicles and they require careful assessment of their impact on ride quality as they increase the unsprung mass of the wheel and lower the wheel-hop natural frequency. PM machines with higher torque/inertia ratio are more suitable for this application. Since they directly drive the wheels one should consider all the performance issues cited above.

To aid research in this area, a tool is needed to quickly design and compare the performance of various motors suitable for a specific application. In other words, given a certain packaging constraint and performance requirements, a tool is needed to give an indication as to which of the motors is most suitable. Current motor design programs rely heavily on empirical experience factors; however, they should be able to link with power electronics and drive simulation programs to determine the performance of the whole system. The results of parameter changes in the motor should be directly quantifiable in terms of the energy consumption over a certain drive cycle for each type of motor. A preliminary effort in this direction has been put in place in the program developed under the Department of Energy contract ADVISOR.¹ More work needs to be done to validate the models and to add the motor design programs to quantify the performance with parameter changes.

In all electrical machines the majority of losses come from either the copper loss in windings or iron losses in the magnetic materials. Special magnetic materials such as Metglass, which have very low iron loss even at high frequencies, will be ideal for stators

of electrical machines, but these materials are not easy to manufacture into a laminated form. Development of low-loss materials that can be easily manufactured is one of the challenges.

Interior PM motors have an excellent potential to be used in electric traction. Neodium-iron-boron magnets like Magnequench are ideally suitable for motors as they have high enough residual flux density. It is preferable to design these motors with high reactance (1 per unit) to avoid accidental demagnetization due to terminal short circuits. Making these buried PM motors for large machines is a challenge. Techniques of injecting magnetic materials into the motors, and curing and magnetizing them on-site are needed. These materials need to be developed without sacrificing their characteristics of remnant flux density and energy product.

The efficiency of power electronics has improved considerably in the last 5 to 10 years and is not much of a concern, particularly for insulated gate bipolar transistor (IGBT) inverters used in electric motors, since they tend to operate at reasonably low switching frequencies so as to limit iron loss in the motors. Further improvement in efficiency can be obtained by the use of wideband gap materials (e.g., SiC; see Chapter 4) for the anti-parallel diodes. SiC diodes used with IGBT power devices in inverters have shown considerable reduction in reverse recovery losses and electromagnetic interference (EMI) comparable to soft switching techniques for motor drives. SiC power devices also offer high-temperature operation and high thermal conductivity but are still in the development stage for high-current, high-voltage application. This is an area of prime research for inverters.

Packaging of power inverters to meet the cooling EMI and electromagnetic compliance (EMC) requirements is also a challenge. The most important development needed in inverters is in the area of packaging and thermal management. There are two levels of cooling requirements to be met: (1) to take heat of the silicon wafer to the substrate on which it is mounted and (2) to take heat from the substrate to the external cooling system. Phase transition liquids and carbon foam materials have excellent promise for meeting these cooling requirements. There is a need for some fundamental research in this area to come up with materials that are most suitable.

All variable-speed drive systems require large bus capacitors for exchanging energy between the windings of the motors and the direct current (DC) bus in order to keep the voltage ripple within limits and increase the life of the batteries by limiting the ripple current. The single most common cause of failure in power electronics is the failure of these capacitors. The electrolytic capacitors for this purpose become very bulky and cannot operate at elevated temperatures. Some of the polyfilm capacitors can

withstand higher temperature and carry relatively large ripple currents, but they are also bulky. Ceramic capacitors offer the advantages of both smaller size and good high-temperature capabilities, but have complex failure modes that are not well understood. Fundamental work is needed in developing materials for bus capacitors to reduce their size, improve efficiency, operate at higher temperature, and increase ripple current carrying capacity.

Since a high voltage is produced in the majority of hybrid power systems on vehicles, it becomes imperative to have a DC/DC converter to supply all the auxiliary loads on the vehicle. Although the technology for this is well developed for low-power converters (e.g., a few watts), further work needs to be done for high-power applications. It is a big challenge to meet all of the vehicle standards for EMI and EMC as well as specifications of efficiency and packaging. The soft switching technologies are most suitable to these converters. Since the ratio of voltage conversion is going to be high (e.g., 320:14 V), it is necessary to have a transformer interface and use a combination of devices. Several topologies are possible, and evaluation and development of the optimized converter are a challenge. For example, an IGBT device may be used for the front end and metal oxide semiconductor field effect transistor (MOSFET) in the output. Under this architecture, choice of switching frequency and soft switching become critical.

Both the electric motors and power electronics need cooling systems in order to operate at their full capacity. Combining the two systems is one of the major challenges. One way to do this is to incorporate the power electronics into the motors with the cooling system made common to both of them. This is a real packaging challenge, but leads to the most desirable solution. New materials such as graphite foam and silicon carbide, which have higher thermal conductivity, need to be developed for this application. Use of these materials, in conjunction with phase transition fluids (vapor to liquid and vice versa), will optimize the performance of the drive systems. Problems of handling these materials need to be addressed, and their physical characteristics (e.g., resistance to corrosion) need to be improved.

Table 2-1 summarizes the major points of this chapter.

Elasser, A., M. Kheraluwala, M. Ghezzo, R. Steigerwald, N. Krishnamurthy, and J. Kretchmer. 1999. A Comparative Evaluation of New Silicon Carbide Diodes and State-of-art Silicon Diodes for Power Electronics Applications. IEEE Industry Applications Society Conference Record: 341-345.

Kennedy, Joseph. 2000. Reconnaissance, Surveillance and Targeting Vehicle (RST - V). Online. Available at www.dtic.mil/ndia/ewc/Kennedy.pdf. Accessed December 2002.

Klett, James and Bret Conway. 2000. Thermal Management Solutions Utilizing High Thermal Conductivity Graphite Foams. online at <http://www.ms.ornl.gov/researchgroups/cmt/foam/Graphite_Foam_Heat_Exchangers.pdf> .

Skellenger, Gerald D., Michael G. Reynolds, and LO Hewko. 1993. Freedom: an Electric Hybrid for Automobiles. Proceedings of the International Conference on Hybrid Power Plants, Zurich, Switzerland, November 9.