Abstract

Weight optimization of aluminium alloy automobile parts reduces their weight while maintaining their natural frequency away from the frequency range of the power spectral density (PSD) that describes the roadway profile. We present our algorithm developed to optimize the weight of an aluminium alloy sample relative to its fatigue life. This new method reduces calculation time; It takes into account the multipoint excitation signal shifted in time, giving a tangle of the constraint signals of the material mesh elements; It also reduces programming costs. We model an aluminium alloy lower vehicle suspension arm under real conditions. The natural frequencies of the part are inversely proportional to the mass and proportional to flexural stiffness, and assumed to be invariable during the process of optimization. The objective function developed in this study is linked directly to the notion of fatigue. The method identifies elements that have less than 10% of the fatigue life of the part's critical element. We achieved a weight loss of 5 to 11% by removing the identified elements following the first iteration.