Lightweight optimization of the side structure of automobile body using combined grey relational and principal component analysis

In this paper, the side structure of the automobile body, as the main assembly to withstand the impact force in side collision, is taken as the research object for multi-objective lightweight optimization. First, finite element analysis (FEA) models of the basic NVH (Noise, Vibration and Harshness) performance of the automobile body and the crashworthiness performance of the vehicle are respectively constructed and validated by actual experiments, through which lightweight controlling quotas are extracted. Second, contribution analysis is employed to determine the final parts for lightweight optimization considering both discrete material variables and continuous thickness variables. Third, design of experiment (DoE) based on Optimal Latin Hypercube Sampling (OLHS) method is performed, considering the total mass, the bending stiffness and the torsional stiffness of the automobile body, the maximum impact acceleration at lower end of the B-pillar and the total material cost of the selected optimization parts as five objective functions. On this basis, the combination of thickness-material parameters of the optimization parts is optimized based on Grey Relational Analysis (GRA), and the Principal Component Analysis (PCA) is applied to evaluate the weighting values corresponding to various objective functions. Meanwhile, a comparison between GRA and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) is conducted to illustrate the unique merits of GRA in multi-objective lightweight optimization of the side structure of the automobile body. Finally, the effectiveness of the lightweight optimization is demonstrated by the comparison between the initial design and the optimal design. The results indicate that the total mass of the automobile body is reduced by 4.54 Kg while other mechanical performance of the automobile body are basically well guaranteed. Hence, the combined grey relational and principal component analysis is a very powerful method used for multi-objective lightweight optimization of the automobile body.