Improved Near Surface Residual Stress Characterisation through Development of an Uncertainty Estimator for the Eigenstrain Reconstruction Method

Veck, Samantha (2024). Improved Near Surface Residual Stress Characterisation through Development of an Uncertainty Estimator for the Eigenstrain Reconstruction Method. PhD thesis The Open University.

DOI: https://doi.org/10.21954/ou.ro.000172a2

Abstract

Near surface tensile residual stresses contribute to structural degradation processes including fatigue initiation and crack growth and stress corrosion cracking. Characterisation of residual stress within 250 microns of the surface is challenging, with high uncertainty and limited spatial resolution provided by laboratory techniques such as Incremental Centre Hole Drilling (ICHD) and X-Ray Diffraction (XRD). Near surface characterisation improvement is crucial to reducing conservatism of structural integrity assessments for safety critical components.

Synchrotron X-ray diffraction provides high spatial resolution and low uncertainty measurements, however it requires more resources than laboratory-based measurements, with access, time and cost restrictions. The Eigenstrain Reconstruction Method (ERM) uses limited measurement data to reconstruct a continuous description of stress over a part domain and is an attractive option for spatial resolution improvement. Unfortunately, it is unknown if ERM is suitable for use with laboratory-based data sets, if reconstruction error can be estimated and if uncertainty reduction occurs.

The aim of this thesis is to determine if near surface residual stress characterisation improvement is possible using ERM with laboratory-based residual stress data. Systematic numerical and experimental studies are conducted to determine ERM feasibility for characterisation improvement and to develop an uncertainty estimator for the method. Experimental results are validated against synchrotron X-ray diffraction measurements.

The first original contribution to knowledge is the successful validation of ERM using laboratory-based residual stress data to reduce near surface uncertainty. A 49% reduction in average near surface uncertainty has been demonstrated at ICHD measurement positions. The second original contribution is the development and validation of an ERM uncertainty estimator. For the first time, ERM reconstructions are attributed an uncertainty value and can now be used quantitatively to support integrity assessments. The methods developed can be adopted without residual stress measurement technique advancement, providing immediate practical application opportunity.

Viewing alternatives

Metrics

Item Actions

Export

You can export this page using these formats Digital Object Identifier - DOI