Promoting excellence in mobility engineering

  1. FISITA Store
  2. Technical Papers

Optimization of the Steering Characteristics in The Virtual Whole Vehicle - Considering Real-time-capable Multi-body Suspensions, Steering, Control Systems and Tires
FISITA2014/F2014-IVC-068

Authors

Miquet, Charles; Henning, Josef; Schmidt, Eberhard; Schick, Waldemar; - IPG Automotive GmbH

Abstract

Research and /or Engineering Questions/Objective

So far, the domains of multi-body system suspensions (MBS) and real-time simulation have been completely separate entities. On the one hand the current MBS techniques are handicapped by very slow computation performance, productivity and global vehicle evaluation of integrated systems. Furthermore, MBS cannot be used for real-time applications for controller development and XiL methods, where real-world systems such as the ECU, the steering system, ESC and the powertrain system are integrated in a whole virtual vehicle. On the other hand, the real-time models are often not detailed enough to modify directly design parameters such as hard points and bushings or represent dynamic chassis characteristics, which in some applications are essential for gaining knowledge about the interaction. Now the long-held dream of applying a full-vehicle model with MBS (>100 DOF) in real-time has become true for the first time worldwide. This milestone achievement makes it possible to efficiently evaluate vehicle dynamics for ride and handling in the whole vehicle. The method will be illustrated by the example “optimization of the vehicle steering behaviour”. The optimization was evaluated with a maneuver catalogue and evaluation criteria - in the same manner as with real road tests.

Methodology

IPG has developed real-time MBS axles as well as a new axle interface for its vehicle model. Different model types can now be switched: MBS axles as well as the given multidimensional 3D K&C models. The MBS axle models were developed by means of MESA VERDE. The benefit is that this formalism represents the full non-linearity of all effects and is able to generate the differential equations of motion automatically. This has a symbolical and alphanumerical structure for dynamic allocation as well as intelligent substitution algorithms for minimal differential equations, which leads to a reduction of matrix operations. C-code can be generated directly and automatically. In addition to the service-based software architecture with multi-threading the comprehensive interface structure with the model management was key factor of success to meet the real-time target.

Results

Studies are possible in the interaction of many subsystems and components, the optimization can be conducted on complete vehicle system level. The simulation speed is boosted by a factor in the range of up to 1000 resulting in a corresponding increase in parameter studies and test scenarios. New fields of application are now made possible (e.g. fully integrated MIL, SIL and HIL tool chain) to validate the interactions with multiple controllers and their variants. All of the remaining time-tested tools and methods, such as automation, analysis and test bench communication, can be combined in this process. The large number of variants is managed by DOE methods with optimization algorithms.

Conclusion

New optimization potential can be harnessed. A vast array of chassis design parameters such as geometries, kinematics, hard-points, etc. can be approved in combination with multiple systems such as air suspensions, steering, powertrain or in the whole vehicle.

KEYWORDS – vehicle dynamics, intelligent vehicle control

Add to basket