Using the example of a drive inverter, the following typical test requirements can be realized very efficiently with an e-motor emulator.
Take a look at our test cases.
Setting electric motor controls
A well-adjusted motor control depends heavily on the current control of the inverter. The torque at the e-motor shaft is proportional to the current and correlates with the current angle. With an e-motor emulator this challenge can be met much easier than with a real motor. The inductance values in the d- and q- axis as well as the phase inductance can be described via adjustable parameters and without any temperature dependencies.
Unlike a real e-motor, there is no need to pay attention to the heating of the motor. So, all adjustments can be done under constant conditions. Whether saturation behavior is to be represented or not is left to the user. The given virtual rotor position is always numerically constant and is not subject to any control accuracy by the loading machine. If torque is generated during energization, this has – unlike in a mechanical setup – no effect on the stability of the rotor position.
Testing inverter overload protection
Drive inverters usually have an integrated overload protection, which ensures that the circuit breakers are not overloaded. On the one hand, this software-based monitoring must ensure that the circuit breakers are not overloaded in any operating state, and on the other hand, this overload protection must not leave too much “reserves”, as this reduces the performance of the inverter (competitive feature).
The self-protection is complex and must react very quickly due to the low thermal inertia of the semiconductors – meaning, it cannot be based on temperature measurement only. With the help of an emulated e-motor, operating conditions can be represented in an absolutely constant and reproducible manner, which greatly simplifies the verification of self-protection.
Testing recuperation behavior
By means of recuperation (energy recovery), the traction inverter can brake the drivetrain. This braking function has a strong influence on the driving behavior and the driving experience in an electric vehicle and must be designed correctly, according to the desired behavior – and of course it must also be tested. Extreme braking maneuvers, which can bring the inverter to its performance limit, are particularly demanding.
During such tests, it is important that the test bench does not have a “life of its own”. With a mechanical system, this can become critical due to the dynamometer control. With an emulated e-motor, the dynamics of the original system can be replicated correctly and without influences from the dynamometer control.
The efficiency of the inverter in an electric powertrain has a direct influence on the vehicle range or the required battery capacity. Today, the efficiency of drive inverters is usually still determined on a test bench together with a real e-motor. The measurement process is demanding and time-consuming (up to 30 hours per measurement plot!), which is due to the fact that each individual operating point must as a first step be tuned mechanically and then the e-motor must also be thermally stabilized in order to obtain reproducible results.
In addition, there are inaccuracies due to the adaptation of the e-motor and possible influences due to motor tolerances, if the tests are not always carried out with the same motor. By means of precise e-motor emulation the efficiency map of an inverter can be determined with the same resolution and the same accuracy in only a few minutes. Exact reproducibility is guaranteed, unwanted influences of the real e-motor are being omitted.
Inverter characteristics in the WLTP cycle
For a drive inverter test based on the standardized WLTP cycle, the e-motor emulation is connected to a vehicle simulation via its serial interface. In that way, the energy consumption of the inverter (power loss) can be easily determined an emulated environment – without rotating components. It is important to note that, by using an e-motor emulator in this test cycle, the influence of the e-motor is “neutralized”, and inverter behavior is evaluated in isolation.
In combination with a real e-motor, such an analysis of the inverter is not possible, as the result is always influenced be the physical e-motor. In the emulated environment, the inverter can be optimized for itself and so enables the possibility to compare different inverters in an easy and “standardized” way (inverter benchmarking).
Dynamic tests - ABS and slip control
Tests on a rotating test rig have limits in terms of the achievable dynamics. This is due to the physically existing rotating parts (rotational inertia), the torsion of the flange and the control dynamics of the dynamometer. Ultimately, a dynamometer is only a “second drive train”.
Depending on the application, this can lead to the load system behaving like a low-pass filter during dynamic tests. Therefore, dynamic tests such as those required for ABS control or slip control do not provide a representative result. There is even a risk that the drive inverter control functions will be tuned to the control characteristics of the dynamometer and the real inverter control behavior will later show a completely different result in the vehicle.
These physical limitations do not exist with an e-motor emulator, since there are no rotating parts in the test arrangement. Thus, the emulation environment ensures that the real vehicle characteristics are being correctly imprinted without compromise. The dynamic controller settings of the inverter are therefore also valid later in the vehicle, given that a correct vehicle model has been used on the test bench.
Fault behavior and safety tests
So-called “fault stimulation tests” on a drive inverter are required to verify the required behavior of the inverter in fault conditions. Especially with regards to the ISO 26262 standard, this is necessary to validate the functional safety of an electric drive train. Depending on the vehicle topology, the vehicle inverter can have a very high criticality (e.g. ASIL-D in the case of torque vectoring), which must be demonstrated by appropriate tests on the UUT. Typical fault scenarios include, for example, loss of rotor position, phase short circuits or an active short circuit of the inverter as a “fail-safe” strategy.
In addition, there are numerous other important fault cases such as overheated magnets in the e-motor, bearing imbalance or de-magnetization, to name just a few. With a real e-motor these fault stimulation tests would be difficult or even impossible to perform, for example repeat tests as part of an automated test are not technically feasible. In a test system with an e-motor emulator, faults in the e-motor can be represented in a very simple, reproducible and automated manner, since in this case the e-motor is merely modeled by means of software. Another advantage of this test setup is the extended protection for the drive inverter, as the emulator limits the phase current to a predefined maximum value, so the UUT is not damaged in case the inverter shows unwanted behavior during fault tests.
Testing high-speed drive inverters
Testing high-speed drive inverters together with a real electric machine requires highly sophisticated mechanical setups. If, in applications such as turbochargers, speeds of well over 50krpm are to be realized, the smallest imbalances or errors can lead to the destruction of the test rig. Accordingly, protective devices are required for such tests. Another complicating factor is that reduction gears must be used for very high rotational speeds in order to adapt the rotation speed to the limits of normal load systems.
This use of reduction gears causes further deficiencies such as gear backlash, friction, vibration and additional mass inertia into the test setup. When using an e-motor emulator, all of these risks and limitations are eliminated, as there are no rotating parts included in the test setup. Electric field frequencies of up to 5kHz can be emulated by default.
Testing interaction of MGUH and MGUK in Formula 1
As stipulated by FIA regulations in 2014, formula 1 powertrain systems provide for two electrical systems in the powertrain. The KERS (kinetic energy recovering system) and the HEAT (electric turbocharger) or the “MGUK” and the “MGUH”. Both drive units operate at very high rotational speeds and the design of these two systems with real electric motors has many shortcomings. Because of the high rotating speeds, reduction gears must be used to match the load systems. High-performance motors are extremely expensive and subject to wear.
In addition, testing times are limited by forementioned FIA rules and regulations, implicating, that a test setup using real e-motor is heavily affected and therefore strictly limited. This means that inverter tests in such a test bench topology are subject to tight testing times. When using e-motor emulators, these restrictions do not apply. The test setup becomes much simpler because no dynamometers, reduction gears or expensive original engines are required. What is more, there are then no longer any limits on the test time in accordance with the FIA regulations.