Basis functions for residual stresses

TL;DR

This paper develops a mathematical basis for representing arbitrary residual stress fields in bodies, enabling approximation and analysis of complex stress states based solely on geometry.

Contribution

It introduces a novel eigenfunction-based basis for residual stresses, extending the approach to 3D bodies and providing convergence guarantees.

Findings

Eigenfunctions form an orthonormal basis for residual stress states.

Numerical examples demonstrate convergence in $L^2$ norm.

Method extends to three-dimensional stress states.

Abstract

We consider arbitrary preexisting residual stress states in arbitrarily shaped, unloaded bodies. These stresses must be self-equilibrating and traction free. Common treatments of the topic tend to focus on either the mechanical origins of the stress, or methods of stress measurement at certain locations. Here we take the stress field as given and consider the problem of approximating any such stress field, in a given body, as a linear combination of predetermined fields which can serve as a basis. We consider planar stress states in detail, and introduce an extremization problem that leads to a linear eigenvalue problem. Eigenfunctions of that problem form an orthonormal basis for all possible residual stress states of sufficient smoothness. In numerical examples, convergence of the approximating stress fields is demonstrated in the $L^2$ norm for continuous stress fields as well as for a stress field with a simple discontinuity. Finally, we outline the extension of our theory to three dimensional bodies and states of stress. Our approach can be used to describe arbitrary preexisting residual stress states in arbitrarily shaped bodies using basis functions that are determined by the body geometry alone.