New simulation tools provide engineers with an advanced way to develop and test engine management systems using thermodynamics to assume the characteristics of the in-cylinder combustion process occurring inside diesel and gasoline engines.
Using hardware-in-the-loop (HIL) simulators for electronic control unit (ECU) development is commonplace today, but the design parameters are evolving. "Future engine control concepts will require more precise physical-based engine models for the HIL simulator since new actuator and sensor signals need to be simulated in closed loop on the HIL test bench," explained Application Engineer Tino Schulze, who is based at dSPACE's headquarters in Paderborn, Germany. "More precise physical-based engine models offer the possibility to do more ECU testing and development on the HIL test benches instead of using real engine test benches," added Schulze.
dSPACE's new in-cylinder simulation models provide an aid to engineers working on tomorrow's technology, such as gasoline engines fitted with variable valve lift and variable valve timing devices. Schulze explained that standard, real-time capable simulation models—so called Mean Value Engine Models—are not well suited for an HIL test bench for future engine concepts since these models do not incorporate the physical effects of different valve timing and valve lifts. "It is necessary to switch to the new in-cylinder, pressure-based simulation models to simulate the effect of valve lift and timing control on engine behavior," said Schulze.
Diesel engine management systems are also evolving. "It can be expected that in-cylinder pressure management–such as an integrated pressure sensor in-the-glow-plug—will be used to achieve more accurate control of the diesel combustion. With an HIL simulator, this new ECU sensor signal needs to be simulated as a result of the different injection patterns that can be controlled by the ECU. Thus, real-time capable diesel engine simulation models—which can generate consistent in-cylinder pressure curves as a result of the actual injection pattern—are necessary to test future diesel engine management systems," Schulze said.
New thermodynamic-based engine models will enable the investigation of future engine concept behaviors without requiring the engine to be available as real hardware, explained Schulze. Using a thermodynamic approach correlates to building "a single calculation zone inside of the simulated cylinder," according to Schulze. Thermodynamics considers aspects such as the energy flows through the intake and exhaust combustion process as well as wall heat losses. "The energy balance inside of the thermodynamics calculation zone allows the simulation of the in-cylinder temperature and pressure with high accuracy in real time," said Schulze.
The ability to replicate real-time performance across the entire engine speed and load range is crucial to HIL testing because the purpose of the HIL simulator is to establish stable closed-loop communication between the software model and the real engine ECU. Schulze said that the main challenge of this HIL simulator is to ensure that the ECU does not recognize that it is connected to a simulation model, so signals sent to the ECU from the HIL simulator are consistent with the actual actuator signals generated by the ECU (i.e., injection timing, throttle position) for all engine working points.
The diesel and gasoline in-cylinder engine simulation models, which have a common base structure, were developed with the aim for easy and precise parameterization based on specific measurements. Although standard real-time capable engine models can be parameterized with a set of conventional engine test bench measurement data, that is not the case with dSPACE's new engine simulation models. "The parameter cannot be obtained directly from standard engine test bench measurements," Schulze said, adding, "In-cylinder pressure measurements are necessary. The parameters of the model need to be calculated by comparing the simulation results with the measured in-cylinder pressure curves. For this reason, a generic optimization process was developed which calculates the model parameters automatically."
Developing and testing engine management systems on an HIL simulator offers some advances compared to conventional scenarios. For instance, the need for engine prototypes and engine test bench facilities to develop and test new engine control algorithms can be reduced significantly. And compared to non-real-time engine simulation models running on standard PCs, dSPACE's new simulation model can be coupled directly to the real ECU hardware through the HIL simulator, according to Schulze.