This thesis answers the question how to derive subsystem requirements with respect to given targets at vehicle level and how to evaluate the performance of subsystems and modules at early stages of design and prototyping.
It is shown that based on identified subsystem properties the expectable performance of the vehicle is predictable much earlier than vehicle prototypes are available for test drives.
In a first stage, the vehicle is divided into steering and front-axle subsystems. Based on desired properties of the vehicle corresponding requirements for subsystem dynamics are derived. To this end, methods of dynamic substructuring in the frequency domain are applied to the problem.
During the thesis both virtual and experimental testing methods are developed, which are capable to determine the actual dynamics of the steering and the front axle. In case of the front-axle subsystem virtual testing is essential, since each subsystem has to be investigated apart from the vehicle and the other subsystem, which in case of the front axle is not very practical at experimental testing. Therefore, a purpose-built virtual roller test-rig for multi-body simulation is developed in this work, which allows for assessing the front axle including compliant parts and rolling wheels with tires under realistic load conditions.
By assembling the thus identified or calculated subsystem dynamics of the steering and the front axle in the frequency domain, the vehicle performance becomes predictable. This allows for assessing each subsystem by its own contribution to the vehicle performance very effectively and furthermore helps to separate different causes from each other.
Although methods of dynamic substructuring in the frequency domain are generally limited to linear subsystems, it is shown that even subsystems with nonlinearities can be investigated as well, if the nonlinearities are dependent of the input force level. To this end, for each level of input force the assembling and / or disassembling of the subsystems is done iteratively until the frequency-dependent interface forces are identified with satisfying accuracy. As a consequence, both the required and identified subsystem dynamics vary not only over frequency but also over the input-force level.
In a second stage, the beforehand derived requirements to subsystem dynamics are used for designing the involved modules. For this purpose, many design parameters of the modules of each subsystem are varied stochastically using design-of-experiment methods. Subsystem designs evolving from the thus modified modules are each assessed with respect to the corresponding subsystem requirements. By this means, ranges of valid values for each parameter are identified, which will be called as “parameter corridors”.
Additionally, the results of the DoE studies reveal the most sensitive parameters of the modules of each subsystem with respect to the given vehicle target. It is demonstrated how to adjust or extend the permissible ranges of values for some chosen parameters by properly tuning others.
«
This thesis answers the question how to derive subsystem requirements with respect to given targets at vehicle level and how to evaluate the performance of subsystems and modules at early stages of design and prototyping.
It is shown that based on identified subsystem properties the expectable performance of the vehicle is predictable much earlier than vehicle prototypes are available for test drives.
In a first stage, the vehicle is divided into steering and front-axle subsystems. Based on d...
»