Multi-scale Residual Stress Assessment with Uncertainty Quantification in a W-Cu Nuclear-reactor Component

Due to its multiscale nature, residual stress (RS) cannot be uniquely assessed without defining the reference length scale. This issue poses significant challenges when evaluating the structural integrity of materials and structures. Additionally, probing RS at different length scales requires fundamentally different experimental techniques, each introducing distinct sources of uncertainty. As a result, direct cross-comparison of RS across different length scales is difficult to achieve.

This study presents a multiscale evaluation of RS in a dissimilar jointed tungsten-copper (W-Cu) monoblock geometry component designed for nuclear fusion reactor applications. The focus is on RS induced by severe thermal cycles and gradients occurring during in-service conditions rather than those arising from manufacturing processes.

A key aspect of this research is the evaluation of RS using two experimental methods based on the relaxation phenomenon, while accounting for various uncertainties throughout the assessment process. At the macroscopic scale, the Contour Method (CM) is employed to provide a full-field evaluation of at least one RS component. In addition, the Plasma Focused Ion Beam Digital Image Correlation (PFIB-DIC) technique is used to investigate RS distribution at the intergranular (Type II) and intragranular (Type III) levels, achieved through the appropriate selection of the gauge volume (i.e., ring core size). For both methods, regularisation algorithms are applied to post-process the experimentally obtained data while evaluating the propagated errors.

The findings contribute to a better understanding of the challenges involved in probing RS across different length scales from an experimental perspective. Ultimately, further insights into the practical application of multiscale RS evaluation in the context of fatigue will be discussed.