Physics-Based Prediction of Residual Stresses for Fatigue and Crack Growth Life Assessment

An explicit computational framework for mechanical surface treatments; such as laser shock peening (LSP), low plasticity burnishing (LPB), and shot peening (SP) is highly sought by designers of critical commercial and defense aerospace parts in the ability to accurately capture relevant residual stress (RS) and damage states for optimal fatigue resistant.  In particular, the shockwave induced by LSP can produce residual stresses deeper into the material than other surface treatment processes, while also minimizing the deformation of the surface. However, under certain loading and geometry conditions the shockwave is prone to inducing subsurface cracking (i.e., spall fracture). A better understanding for the LSP process of the RS field through visualization of the shock-wave behavior and the resulting data in the material to critical process parameters, material properties, and geometry effects are all of interest. Corvid and Curtiss-Wright Surface Technologies proposes to explicitly capture the LSP process with an experimentally validated high-fidelity computational physics (HFCP) numerical framework following the laser shock peening process of complex geometries. Within the Phase I effort, Corvid demonstrated success in developing an HFCP framework for predicting outcomes of interest in LSP within 80% confidence for planar material coupons of Ti-6Al-4V as a function of varying LSP processing effects (irradiance, pulse width, overlap, peening patterns). In the Phase II, Corvid proposes to i) advance the physics based modeling approach to account for complex geometry (i.e. turbine blades/bulkheads), ii) investigate and define the damage limits needed for the LSP process, and iii) complete the feedback loop from the modeling approach to component design in order to optimize service life. The scale up of from material coupons with predominately planar features to more representative complex parts accounting for a component’s fillets, thin sections, and leading edges will be explicitly modeled and experimentally validated. Validation of the LSP process will be achieved by advancing from a one-dimensional RS characterization to multi-dimensional RS characterization and damage investigation through the inclusion of Scanning Electron Microscopy (SEM) and fatigue cycling. Finally, these computationally intensive high-fidelity multi-physics LSP simulations for a range of geometries, laser processing variables, and material sensitives will be used as training data for transitioning to reduced order models (ROM) to further inform crack growth assessments and increase confidence in component lifetime predictions as well as capture any inadvertent damage created through the LSP process.