The deviation vector at arbitrary location of large thin-walled structure caused by manufacturing process is different and has the characteristic of field distribution, which has great influence on the assemble quality. The deviation of each point on the part is not independent, and the final assembly deviation is difficult to be controlled. In this paper, the deviation field of large thin-walled structure is described by the linear combination of a series of basic deviation patterns. The deviation propagation model is established to quantify the contribution of basic deviation patterns between parts and assembly. A new two-step optimization method based on the adjustment of key control points of the part is proposed for the deviation control of large thin-walled structures. Firstly, the effective independent method is employed to obtain the optimal measurement points, which may characterize all basic deviation patterns of the part accurately. Then a new optimization model is developed to determine the key control points for special basic deviation pattern, which have little influence on the other basic deviation patterns. Based on the genetic optimization algorithm, the optimal key control points and the adjusted quantities for special basic deviation pattern are obtained, simultaneously. A case study on the assembly process of two cylindrical thin-walled parts with initial deviations measured by the Laser Scan Device is conducted. The basic deviation pattern with great influence on the deviation of assembly is determined firstly. The key control points and the corresponding adjusted quantities for this basic deviation pattern are calculated. The results indicate that the deviation of the assembled structure may be suppressed by the adjusted deformation of the key control points of parts. It is useful on the deviation control for the assembly process of large thin-walled structures.
Non-parametric shape optimization method for thin-walled structures under strength criterion
