Determination of microscopic residual stresses using diffraction methods, EBSD maps, and evolutionary algorithms


Residual stresses, both macroscopic and microscopic, are originated during conventional metallurgical processes. Knowing their magnitude and distribution is of great importance in the structural design of applications where fatigue, stress corrosion or thermal cycling occur (e.g., in the aerospace industry). The importance of these stresses is reflected in the large number of articles published in recent years, mainly focused on studying macroscopic stresses. However, there are no experimental studies that quantify the magnitude of microscopic triaxial stresses. This lack is due in part to the limitations of diffraction techniques (neutrons and synchrotron radiation). Since the measurement volume is much higher than the variation of these microscopic stresses, its calculation is greatly complicated, because the methods used in the case of macroscopic stresses are not valid for microscopic ones. Furthermore, there is no reliable procedure to obtain the relaxed lattice parameter value, a key factor in the calculation of residual stresses. The aim of this paper is to present the main ideas oriented to develop a methodology for mapping microscopic stresses, particularly in aluminium alloys such as those commonly used in the aerospace industry. The procedure will use experimental diffraction results obtained from large European facilities, mainly by neutron diffraction. This information will be analyzed using evolutionary algorithms, computational techniques that handle a large number of variables. The procedure will be based on the analysis of the shift of the diffraction peaks and, fundamentally, their broadening. For simplicity, non-heat treatable alloys will be used as they do not experience lattice parameter variation with heat treatments.

Proceedings of the Genetic and Evolutionary Computation Conference Companion
J. Manuel Colmenar
J. Manuel Colmenar
Associate Professor

My research interests are focused on metaheuristics applied to optimization problems. I have worked on different combinatorial optimization problems applying trajectorial algorithms such us GRASP or VNS. Besides, I am very interested in applications of Grammatical Evolution, specifically in model and prediction domain, as alternative to machine learning approaches.