When it comes to validating and refining Hardware in the Loop (HiL) or Driver in the Loop (DiL) test systems, early in the development process, real-time simulation is an essential part of the process. Real-time simulations provide an opportunity to connect physical components and virtual models in a virtual test environment. High performance computing and advances in engineering simulation techniques are enablers for real-time simulations.
There are numerous advantages related to virtualization. This allows validation of many Advanced Driver Assistance Systems (ADAS). One can explore off-normal conditions that might damage test equipment or hurt test personnel. Systems can be tested before “cutting metal” and creating prototypes. Testing costs can be reduced, enabling broader and more intensive test plans.
A Real-Time simulation mandates that the virtual model delivers a response/solution output at specific time increments in the same or less than the physical system. This way the state of the virtual system can be kept in sync with the physical components to provide an accurate testing environment.
Adams-Real Time represents MSC’s expansion into the world of real-time simulations. Adams-Real Time was initially released in 2017 and continues to advance in every subsequent release. What differentiates Adams Real Time from many other real time solutions is the topology and parametrics of the system is preserved. So elements such as hardpoints, joints, springs, dampers and bushings are maintained and can be modified. The benefit of this is that the model can capture higher frequency characteristics in the system responses and different configurations can be explored in a far shorter turnaround time.
A workflow that manages the interactions between the virtual and real entities in the system also needs to be established. Let’s look at how Adams-RT facilitates these Real-time requirements.
During real-time simulation, the “accuracy” of the simulation not only depends on the correctness of the solution but also when the solution becomes available. Variable time stepping would not work in this scenario. Adams-Real Time provides a fixed time-step integrator where the user can control the number of integrator time steps and the number of iterations within each, to meet prescribed solution deadlines (tx). These levers allow the analysts to capture critical system responses under the imposed Real Time constraints.
Since the underlying intent in real time simulations is to capture system dynamics within specific time constraints, a full-fidelity modeling approach might not be viable. For vehicles, Adams-Real Time provides utilities to simplify some common model details, accelerate solution times, and evaluate performance metrics. Utilities exist within the Adams environment to generate simplified representations of anti-roll bars, tires and flex bodies. The Adams-RT model is exported out as a Functional Mock-up Unit (FMU) and then ported onto a HIL platform for execution. Modeling parameters such as hard points, bushing characteristics etc. can also be exposed in the FMU for run time tuning. This allows consumers of the FMU to make changes to the model parameters without requiring access to the original Adams model.
Adams-RT is currently compatible with the Concurrent Simulation Workbench and dSPACE SCALEXIO platforms. These platforms provide the communication channels to connect model I/O to physical hardware, execute it in real time and visualize the results on the platform. Adams Real Time models are exported as a Functional Mock-Up Unit (FMU) using the tool independent FMI standard and then ported onto the Real Time platform for execution.
HiL is particularly helpful for testing the interaction of un-modeled complex systems and control modules, with a larger system that includes models. This may be ABS, VSC, engines, dampers, flight controls, or autonomous algorithms. A recent implementation of an Adams-enabled HiL simulation allows assessing transmission shift quality at a large automotive OEM. Longitudinal vehicle acceleration associated with a gear change impacts a rider’s perception of vehicle ride comfort. This acceleration can be managed by tuning the way the transmission shifts during a gear change. The engine, transmission and dynamometer in the test cell are connected to a full vehicle Adams model via a HiL platform.
At a conceptual level, the physical torque at the output of the transmission is made available to the Adams model as an input. Given this input torque the Adams model reacts and provides a driveshaft speed response to the dynamometer which then enforces this speed on the engine and transmission. The model also provides the longitudinal vehicle acceleration as the vehicle changes gear, which can be related to a ride comfort index. The ability to use a combination of vehicle dynamics models and hardware to simulate vehicle behavior provides an opportunity for both more efficient and more comprehensive testing programs as well as reducing vehicle prototype demands.
What are your thoughts on real time vehicle dynamics models? Drop us a line, we would love to know.