Abstract
The design of vehicle structural components can be tentatively accomplished by a preliminary virtual experimentation as well as the dynamic analysis. Two levels of computation can be performed: simplified mathematical models developed within general purpose mathematical environments, like MATLAB and SIMULINK, can support at least the initial steps of the product development and road test activity for preliminary and fairly fast analyses, while a detailed design is usually performed through multi-body dynamics and FEM codes, including control and flexible bodies effects. The analysis of the dynamic response of the vehicle in time and frequency domains can provide a load distribution on the main structural parts, which can be analyzed in terms of strength of the material and fatigue. Large deformation in flexible elements causes a relevant interaction between the vibration modes of the structure and the dynamic behavior of the system, which affects the vehicle performance.
The same bottom-up approach described in the first part of the paper for the dynamic behavior prediction of the vehicle can be applied to the preliminary stress analysis of the main structural components and durability. A first approximation analysis is at least aimed to identify the most critical stresses and load concentrations, occurring in the mechanical components. The interaction of several physical phenomena including thermal effects, fatigue, electromechanical coupling, required by active control devices if present, needs for the development of a multi-physics desktop, to perform an interdisciplinary design of the whole system. The feasibility of a multidisciplinary structural desktop is herewith investigated through a preliminary construction and validation on test cases of a light version, suitable for first approximation analysis. It looks like an integrated computation environment, where several professional codes, already available on market, each one focused on a specific task of the analysis (structural, multi-body kinematics and dynamics, fatigue, control systems, electromechanical coupling) interact or co-simulate.
The goal of the paper is to froze the state of art of the integration of CAE products for multidisciplinary design approach, by means of a review of the relevant tasks required to the developed desktop, and possibly the bottle-necks and the main limits of the approach, in case of a simplified and quick preliminary analysis useful for a qualitative and rough perception of problems either during a conceptual screening or a road test.
Under the above assumptions the paper should be intended to discuss the main vantages and disadvantages of the approach, instead of proposing a final solution based on a particular selection of commercial codes currently available.
The need for a documented review of the subject motivated the authors to use the software already present at the University and Fiat GM, Chassis and Suspension Design Group, although alternative choices are possible and encouraged to check the possibility of solving some operational problems encountered in the present study. The analysis was aimed to define suitable procedures, interfaces and operations to build a structural desktop for the vehicle structural design and materials selection, and to investigate difficulties in linking and co-simulating codes, time consuming and limits. For this preliminary investigation MATLAB and SIMULINK environments have been preferred to connect the structural desktop quite easily to the simplified vehicle model developed by the authors and deeply presented in the first part of the paper.