Automotive component product development enhancement through multi-attribute system design optimization in an integrated concurrent engineering framework

Abstract

Automotive industry is facing a tough period. Production overcapacity and high fixed costs constrain companies' profits and challenge the very same existence of some corporations. Strangulated by the reduced cash availability and petrified by the organizational and products' complexity, companies find themselves more and more inadequate to stay in synch with the pace and the rate of change of consumers' and regulations' demands. To boost profits, nearly everyone pursue cost cutting. However, aggressive cost cutting as the sole approach to fattening margins results invariably in a reduction of operational capabilities which is likely to result in a decline in sales volume that leads to further cost reductions in a continuous death spiral. Long-term profitable growth requires, instead, a continuous flow of innovative products and processes. The focus should be, therefore, shifted from cost reduction to increased throughput. Automotive companies need to change their business model, morphing into new organizational entities based on systems thinking and change, which are agile and can swiftly adapt to the new business environment. The advancement of technology and the relentless increase in computing power will provide the necessary means for this radical transformation. This transformation cannot happen if the Product Development Process (PDP) does not break the iron gate of cycle time-product cost-development expenses-reduced product performance that constrains it. A new approach to PD should be applied to the early phases, where the leverage is higher, and should be targeted to dramatic reduction of the time taken to perform design iterations, which, by taking 50-70% of the total development time, are a burden of today's practice. Multi-disciplinary Design

(cont.) Analysis and Optimization, enabled by an Integrated Concurrent Engineering virtual product development framework has the required characteristics and the potential to respond to today's and tomorrow's automotive challenges. In this new framework, the product or system is not defined by a rigid CAD model which is then manipulated by product team engineers, but by a parametric flexible architecture handled by optimization and analysis software, with limited user interaction. In this environment, design engineers govern computer programs, which automatically select appropriately combinations of geometry parameters and drive seamlessly the analyses software programs (structural, fluid dynamic, costing, etc) to compute the system's performance attributes. Optimization algorithms explore the design space, identifying the Pareto optimal set of designs that satisfy the multiple simultaneous objectives they are given and at the same time the problem's constraints. Examples of application of the MDO approach to automotive systems are multiplying. However, the number of disciplines and engineering aspects considered is still limited to few (two or three) thus not exploiting the full potential the approach deriving from multi-disciplinarity. In the present work, a prototype of an Enhanced Development Framework has been set up for a particular automotive subsystem: a maniverter (a combination of exhaust manifold and catalytic converter) for internal combustion engines ...