This is the authors' abstract. We don't add key points for this paper.
Femtosecond Laser Shock Peening (FLSP) is an ultrafast laser surface treatment that has received increasing attention due to its minimal thermal damage and high controllability. The fatigue life and corrosion resistance of materials have been significantly enhanced following FLSP treatment. These performance improvements are mainly attributed to the introduction of residual stress. However, effective methods to predict the distribution of residual stress after FLSP remain lacking, which has limited its further application. In this paper, a predictive framework has been developed to link FLSP process parameters, microstructure, and residual stress using multimodal data mining. Microstructural information was obtained using electron backscatter diffraction and then fed into a convolutional neural network. Through multimodal input, complex structural features were effectively extracted, allowing accurate prediction of residual stress distributions under various FLSP conditions, even with a limited number of samples. In addition, the influence of different microstructural inputs on the prediction accuracy of the model was also analyzed. The results demonstrated that the model was able to achieve reliable prediction accuracy in regions beyond the trained range of laser parameters. Overall, this model is expected to facilitate new opportunities for optimizing FLSP processes by addressing existing limitations, thereby enabling improved material performance across diverse advanced manufacturing applications.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Jia et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: