Modeling of fast pre-joining processes optimization for skin-stringer panels

Abstract

Purpose – The purpose of this paper is to propose a new model for optimizing pre-joining processes quickly and accurately, guiding workers to standardized operations. For the automatic riveting in panel assemblies, the traditional approach of determination of pre-joining processes entirely rests on the experience of workers, which leads to the improper number, location and sequence of pre-joining, the low quality stability and the high repair rate in most cases. Design/methodology/approach – The clearances computation with the complete finite element model for every process combination is time-consuming. Therefore a fast pre-joining processes optimization model (FPPOM) is proposed. This model treats both the measured initial clearances and the stiffness matrices of key points of panels as an input; considers the permissive clearances as an evaluation criterion; regards the optimal number, location and sequence as an objective; and takes the neighborhood-search-based adaptive genetic algorithm as a solution. Findings – A comparison between the FPPOM and complete finite element model with clearances (CFEMC) was made in practice. Further, the results indicate that running the FPPOM is time-saving by >90 per cent compared with the CFEMC. Practical implications – This paper provides practical insights into realizing the pre-joining processes optimization quickly. Originality/value – This paper is the first to propose the FPPOM, which could simplify the processes, reduce the degrees of freedom of nodes and conduct the manufacturers to standardized manipulations.