Tolerance Optimization of Patch Parameters for Locally Reinforced Composite Structures

Abstract

A rising number of applications and increasing volume of composite structures production lead to a high relevance of variation management during their design. Structural optimization for lightweight purposes often results in designs consisting of a base laminate with local reinforcement patches. Nominally, these optimized designs offer a thorough exploitation of lightweight potential. Yet, they suffer from variations of the reinforcements resulting in a worsened manufacturing behavior and reduced structural performance. To ensure the quality, tolerances should be allocated for the parameters of the local reinforcement patches. Therefore, in the current contribution a tolerance optimization method is presented identifying optimal tolerance values for the design parameters of the reinforcements with respect to the structural behavior. This includes the discussion of challenges regarding the suitable parametrization and modeling of local reinforcement patches for variation simulation based on Finite Element Analysis (FEA), the usage of surrogate modeling to reduce the computational effort of structural analyses, as well as an approach to penalize tight tolerances of different parameter types. The proposed tolerance optimization is applied to a use case. Tolerances for the patch parameters are optimized, meeting the structural quality constraints of the composite structure.

Continuous simulation of variations during the design of endless fiber reinforced composite structure assemblies

“…The fiber angles of the plies are an important design parameter because they significantly influence the strength and stiffness of the FRP part due to the highly anisotropic material properties. Additionally, local reinforcement patches are used to increase the strength and stiffness of highly stressed areas while keeping the lightweight design [11]. The optimal location, shape and direction of the reinforcement patches can be calculated by laminate optimization approaches [12].…”