Application of Digital Twins in EPS/SbW-Development

As digitalization plays in increasing role within automotive development projects the terminology “Digital Twin” becomes more and more prominent. The paper describes the Digital Twin approach that is used in ZF steering development including model build up, model validation, capabilities and potential use cases in development of robust steering systems. By using some subsystems of a steering system as an example it is shown that the tolerance chains, system wear and environmental conditions play a very dominant role especially in steering system characteristics. When developing any kind of simulation model for a steering system those influences need to be considered when validating the model. For this purpose, the concept of Digital Twins is used at ZF as a digital representation of system characteristics of an individual system including the influence of noise factors. After the description of ZF’s concept of Digital Twins, a typical 1-D model build up is shown including validation using specially designed test rigs. In addition, the statistical validation using the Monte-Carlo method is described. For this, main tolerances and the effects of environmental conditions are set randomly within the given range of tolerance bands and environmental conditions to generate a statistical distribution of model responses. This distribution is then compared with measured responses, allowing model validation against systems where not all noise factors are known. In the last part of this paper the potential and a certain use case of a Digital Twin is described on more detail. It includes a design study of a mechanical system and the validation of robust system behavior in closed loop conditions.