Past, present, and future stress intensity factor solutions for crack at holes

icaf2023 Tracking Number 37

icaf2023 presentation

This presentation summarizes advances in stress intensity factor (SIF) solutions for cracks at holes, set in the historical context of the past forty years and current problems that demand new, novel solutions to support damage tolerance (DT) assessments. Special attention is given to corner cracks, since they are usually the initial crack state and often dominate total DT lifetime. Classic solutions by Raju and Newman (1980s) and Fawaz and Andersson (2000s) modelled wide plates, requiring separate finite-width correction factors for practical application. While the Newman-Raju correction factors were state of the art for their time, they have significant limitations, and we developed new equations with improved accuracy for tension, bend, and pin-loading for single or dissimilar double cracks. Post-transition scenarios led us to develop novel compounding solutions for corner-through or dissimilar through-through crack combinations, building on formulations from NRC-Canada. Weight function (WF) solutions address the practical challenge of finite geometry effects in a different way. Our WF formulation employs analytical basis functions coupled with large matrices of reference solutions over the range of finite widths and offsets. More importantly, the WF approach handles additional stress states besides uniform remote loading and permits explicit treatment of residual stress (RS), including shakedown RS from local plasticity. Most WF solutions are based on stress gradients in a single direction, but because stresses at the corner of a hole are inherently bivariant, we have also developed a WF solution that accommodates in-plane and out-of-plane stresses for a corner (or a surface) crack. Current DT challenges involve physical issues that may not be tractable using traditional approaches. These challenges include cold expansion, interference/clearance fits, manufacturing-induced RS, nonlinear material response, out-of-plane bending, hole interaction, and crack interaction. Our recent efforts exploit advances in curved through-crack formulations, principal component analysis, automatic generation of crack fronts, and machine learning via Gaussian Process models. These tools are leading to incremental advances to support cracks at a row of holes, interference fit for through cracks, and tractable calculations of bivariant stresses near stress concentrations under remote loading.