Fatigue-Focused Manufacturing Optimization for Welded Structures

Abstract

Welded structures have an inherent vulnerability of fatigue failure at joint locations due to change in geometry and material properties. The U.S. Army, in an effort to to increase overall readiness, has taken interest in understanding fatigue behavior and failure prevention specific to their ground combat vehicles. Fatigue testing of armor-grade joints was conducted in order to understand the material behavior, and this work incorporates the information developed from these tests with structural load cases to create a proactive assembly plan. By acknowledging the higher stress regions from typical (i.e., high frequency) loading scenarios and avoiding weld placement in these areas, reduced vehicle vulnerability can be achieved.

Current fatigue life optimization research centers around topographical analysis of structural components, primarily focusing on shape and weight with fatigue limits. These optimizations rarely, if ever, consider the manufacturability and cost of the component or structure. This work develops optimization algorithms, tailored to meet manufacturing limitations, with the primary objective of minimizing weld stress exposure and a secondary consideration of cost.

The final algorithm takes direction from bin-packing optimizations by filling in a representative vehicle side panel with a set of designated plate sizes, and evaluating the stress exposure and cost based on the weld placements and weld types. The stress exposure is determined by the position of the weld and scaled by the weld type. The cost of the assembly considers the overall length of welds and wasted material. The panel assembly designs are then ranked by these evaluations. Results provide a comprehensive, feasible assembly plan that extends the fatigue life of a structure.