Multiscale Analysis of Residual Stresses with Novel Non-Destructive and Destructive Approaches using Surface Displacement Measurements
Grant
Overview
Affiliation
View All
Overview
abstract
Engineered materials and parts undergo severe mechanical and thermal loads during their development that result in the formation of internal residual stresses. Residual stresses (nominally force divided by area) are caused by internal forces that remain even after all external loads are removed. Despite the fact that residual stresses can reduce the service life considerably, comprehensive and practical methods to determine them without destroying the parts do not exist. Current non-destructive techniques allow their quantification on the surface of the parts only. Thus, development of new materials or manufacturing techniques can be greatly helped by a thorough estimation of residual stresses that might develop during the processes. For example, residual stresses are extremely prevalent in additively manufactured parts and strongly depend on process parameters in a complex way that are not understood well. This award supports fundamental research in inverse computational methods based on the mechanics of materials to estimate residual stresses in both non-destructive and destructive fashions from the measurement of forces and surface displacements. Advancement in residual stress analysis will promote both opportunistic use and mitigation in the development of new materials and manufacturing processes, which will greatly enhance automotive, aerospace, defense, and nuclear industries. The research team will also promote engineering and education through participation in the summer camp programs that will involve female and underrepresented high school students. For the development of the non-destructive approach, surface displacements will be measured cost-effectively on varying part sizes using digital camera images to trace speckle patterns. Mechanics-based mathematical algorithms derived using adjoint equations will be implemented to infer residual stress distribution from these measurements. Validation on a 3D printed rubber like material will be conducted. For the development of the destructive approach, very thin sectioned samples will be retrieved from the polycrystalline specimens to relieve all residual stresses, with the implication that the thin slices will deform as a result. Subsequent loadings of the samples will be used to determine the heterogeneous, anisotropic granular microstructure of the material using mechanics-based mathematical models. Application of reverse deformation on the now known microstructure to retrieve the original configuration can then be used to predict the residual stresses in the given section of the material. The fundamental scientific issue underlying both approaches is: What are the minimum number of experimental data sets and deformation modes, i.e., experiments, required to uniquely map residual stresses?
