The research group "Co-Simulation" at ViF concentrates on the research and development of methods and tools for the co-simulation of complex dynamic systems. The following sections give an overview of the core topics.
Contact: Dr. Martin Benedikt
In today's development processes, the simulation of dynamic systems allows predictions and concept decisions related to the final product to be made at an early stage. This not only involves the modeling, simulation and testing of individual structural components or modules, but also requires the interplay of a large number of functions (with simulation models and also hardware components from various domains) to build up the full system (right up to the full vehicle).
The term "co-simulation" refers to a simulation approach where the components of modern mechatronic systems are interconnected in a suitable way. Therefore the complex interactions between these sub-systems from different development areas are taken into account. The coupling of existing (specific) simulation programs (and the models implemented therein) from different areas of expertise, represents a promising approach for the simulation of the overall system.
Co-simulation platform ICOS
The research findings are integrated in the independent co-simulation platform ICOS, developed at ViF. ICOS takes the complex interactions in a suitable and correct way into account. The platform enables the precise co-working of different simulation tools (right up to the real-time-capable "Hardware-in-the-Loop" systems) via a so-called "co-simulation framework". Only the verified interaction of numerous models (and therefore also simulated components) enables a realistic virtual concept design and validation of the overall system consisting of vehicle, driver and environment.
One important challenge in the development of co-simulation is the coupling of models with strong differing dynamic behavior (time constants). In this case, extremely small time steps must be set for the data exchange between the models. Otherwise, the additional quantization caused by the co-simulation can lead to errors, which can force the co-simulation to diverge. This means that incorrect results are generated, which might even go unnoticed.
ViF concentrates on the development of new and highly innovative coupling algorithms to successfully overcome this problem. Methods such as "Waveform Relaxation", multi-rate approaches, adaptive time step control or so-called energy-preserving coupling methods guarantee the correct interaction of models, even with large step sizes. With ICOS, it is possible to achieve a very high accuracy, a good convergence behavior and a short calculation time compared to conventional methods.
A co-simulation framework has to interconnect the various simulation tools. Additionally, numerous tasks have to be fulfilled:
- communication setup
- control of the co-simulation procedure
- synchronization data exchange
- step size control and extrapolation strategies
A further focus of research at ViF is the development of a decentralized, fully self-organizing network (framework) for the interconnection of the simulation tools. This framework handles the mentioned tasks in an autonomous and correct way.
Nowadays, co-simulation methods typically are limited to offline, non-real-time applications which is a strong limitation regarding re-use of existing simulation models. In the case of so-called "hard-real-time" simulation, all participating simulation models have to deliver their specific calculation results within pre-defined time slots. If this rule is violated an error occurs and the real-time-system has to be stopped. The "Hardware-in-the-loop systems" (HiL) which are used in the modern development process belong to this group (e.g. component or ECU testrigs).
The topics real-time systems, real-time models and real-time co-simulation will gain more and more importance in the future. Therefore they are core research topics at ViF.
Fields of application for co-simulation at ViF
ICOS is used in many projects at ViF. Here is a selection of topics:
- Integrated Safety: coupling of FEM for crash simulation with vehicle dynamics simulation and controller development
- Vehicle electrical system simulation: coupling of mechanical/electrical sub-systems
- Battery simulation: coupling of electrochemical & thermal models
- Thermal simulation: coupling of thermal, electrical and mechanical models for the development and optimization of energy management systems
- Mechanical Simulation: coupling of multi-body-simulation/FEM and controller development
- Vehicle dynamics simulation: development and calibration of vehicle dynamics controller based on multi-body-simulation