Target setting and structural design of an EPS-in-the-Loop test bench for steering feeling simulation 2016-01-1559

The adoption of Electrical Power Steering (EPS) systems has greatly opened up the possibilities to control the steering wheel torque, which is a critical parameter in the subjective and objective evaluation of a new vehicle. Therefore, the tuning of the EPS controller is not only becoming increasing complicated, containing dozens of parameters and maps, but it is crucial in defining the basic DNA of the steering feeling characteristics. The largely subjective nature of the steering feeling assessment means that EPS tuning consists primarily of subjective tests on running prototypes. On account of that, this paper presents an alternative test bench for steering feeling simulation and evaluation. It combines a static driving simulator with a physical EPS assisted steering rack. The end goal is to more accurately reproduce the tactile feedback to the driver by including a physical hardware in lieu of complicated and difficult to obtain software models. The focus of the activity in this paper was to define the specifications for this test bench and to entirely design it.

The test bench contains some actuators which mimic the vehicle’s loading of the steering rack. The specifications of these actuators have been derived from an experimental campaign performed with some high performance GT sport cars due to their high force and speed demands for the steering rack. The force spectra were obtained via strain gauge measurements of the track rod of the tested vehicles: a set of maneuvers encompassing different driving conditions were performed and the measured data were matched to the results of dynamic simulations and condensed to build a reliable specification framework. This paper presents the measures performed and the review of the acquired data. Additionally, the whole design of the EPS-in-the-Loop test bench is presented, including both the static and dynamic analyses carried out to validate it through multibody analyses with flexible bodies.