Computational & Technology Resources
an online resource for computational,
engineering & technology publications
PROCEEDINGS OF THE FIFTH INTERNATIONAL CONFERENCE ON ENGINEERING COMPUTATIONAL TECHNOLOGY
Edited by: B.H.V. Topping, G. Montero and R. Montenegro
Pedestrian Hood Panel Robust Design Using First Design Methodology
L. Jézéquel1, G. Lavaud12 and Y. Tourbier2
1Mechanical Department (LTDS), Ecole Centrale de Lyon, Ecully, France
L. Jézéquel, G. Lavaud, Y. Tourbier, "Pedestrian Hood Panel Robust Design Using First Design Methodology", in B.H.V. Topping, G. Montero, R. Montenegro, (Editors), "Proceedings of the Fifth International Conference on Engineering Computational Technology", Civil-Comp Press, Stirlingshire, UK, Paper 134, 2006. doi:10.4203/ccp.84.134
Keywords: robust optimization, robust design methodology, hood panel, first design.
This paper deals with uncertainty management during the design cycle of complex products. We will present a new design methodology named first design aiming at finding a solution at first time. Indeed uncertainty problems may lead designers to late and expensive design iterations. This methodology uses simulation models which complexity follows the progressive refinement of the product. This prevents from simulating performance with unknown parts of the product at the beginning of the design cycle. Functional models seem quite relevant for this purpose. We will detail the construction of such a model for a pedestrian hood panel optimization.
Designers used to split the design of a complex product into several sub-problems. They solved each sub problem independently from the others. Therefore, they are likely to face compatibility problems during validation of the whole product. That is why a number of multidisciplinary optimization methods developed [1,2,3,4,8]. These strategies consist of using an optimizer at the component level to improve the sub-system performance and another optimizer at the system level to enforce compatibility between the sub systems. However, late changes in specifications are still likely to occur.
Detailed simulation models may also be responsible for late changes. They enable designers to accelerate the design cycle, but they lead to uncontrolled design choices. In the early design step, nothing is known about the product except its performance targets. Nevertheless, some data is required to define specifications of the sub systems. Simulations used for these choices require much information that is not yet available. False values used in the early design steps will possibly cause erroneous sub system specifications and late changes in the design cycle.
To prevent from such problems, we recommend using the first design methodology [5,6]. This multi-level design strategy uses simulation models the complexity of which follows the progressive refinement of the product. Moreover, if any optimization target changes, it is easy to propagate changes to all component specifications and avoid total redevelopment of the product. We illustrate the first design method with the design of a pedestrian hood panel respecting new legislation requirements. We propose a low-complexity crashworthiness model to simulate the head injury criterion and the displacement of the hood. As detailed in the article, the Craig and Bampton method  helped us to build such a functional mass and stiffness model.
As a conclusion, we stress the multiple advantages of the design method. First, it prevents an unexpected redesign of the product at the end of the design process. Secondly, functional models used at the beginning of the design cycle reduce the design space and select important design variables. Finally, the low evaluation cost of the functional models allows the consideration of heavy optimization techniques such as topology optimization.
purchase the full-text of this paper (price £20)