Caractérisation du comportement micromécanique de l’acier inoxydable martensitique UNS S41500 dans l’état métallurgique brut de trempe

Abstract

This thesis presents an experimental investigation of the micromechanical behavior of martensitic stainless steel UNS S41500. The study addresses two main objectives. 1) The determination of the mechanical properties of a martensitic lath block through micro-tensile testing. 2) The analysis of the nonlinear behavior observed at low load during quasi-static tensile loading of the as-quenched condition. This research is driven by Hydro-Québec’s objective to model the micromechanical behavior of alloys used in the fabrication of water turbine runners, particularly CA6NM. This initiative stems from uncertainties surrounding fatigue damage mechanisms affecting these components in service. To tackle this challenge, the Crystal Plasticity Finite Element Method (CPFEM) has been identified as the preferred modeling approach. However, the successful application of CPFEM requires accurate knowledge of the intrinsic mechanical properties of the alloy and its deformation behavior. UNS S41500 was selected for this study because it is a wrought alloy with the same chemical composition as CA6NM, while offering superior inclusion cleanliness and being free from casting defects, making it an ideal candidate for this investigation. The morphology of lath martensite in UNS S41500 and CA6NM is well documented. Nevertheless, modeling this complex microstructure using CPFEM reveals significant gaps in the micromechanical properties required for constitutive laws. Specifically, the literature does not provide experimentally measured values for the critical resolved shear stress (CRSS) or the saturation resolved shear stress of a martensitic lath block. Existing data are derived from extrapolation of tests involving block boundaries or from inverse optimization methods. For the past 70 years, a non-linearity observed at low load during tensile testing of as-quenched martensite has been the subject of debate. This non-linearity, observed in microstructures without reformed austenite, has been attributed to various mechanisms, including the presence of mobile dislocations, residual stresses at the microstructural scale (type II), or heterogeneity in the mechanical properties of martensite. Interestingly, tempering below the onset of the austenitic transformation (AC1) (500 °C, 2 h) produces a microstructure free of reformed austenite and exhibiting a fully linear stress–strain response. Conversely, the same microstructure tempered above AC1, thus containing reformed austenite, promotes the development of this non-linearity. Therefore, to decouple the role of reformed austenite from the previously mentioned sources, the as-quenched condition will be used to investigate the origin of non-linearity in a single-phase microstructure. Future work will aim to determine whether the same mechanism accounts for non-linearity in the presence of reverted austenite Consequently, to accurately model the elastoplastic behavior of this alloy using CPFEM, it is essential to determine the micromechanical properties of its microstructure and to identify the mechanism responsible for the observed nonlinearity. The first objective was addressed by extracting micromechanical properties from micro-tensile specimens machined directly from the microstructure. Electron backscatter diffraction (EBSD) characterization enabled the selection of five candidates meeting dimensional and crystallographic orientation criteria. Two specimens were oriented along (110), one along (111), and one along (112) to cover most of the pole figure triangle. The two (110) specimens were used to assess measurement repeatability. Results for CRSS on activated slip systems {110}, {211}, and {123} exhibit good repeatability and align with extrapolated literature values. CRSS was established at 291 MPa across all planes with a standard deviation of 16 MPa. Saturation resolved shear stress was measured at 364 MPa with the same standard deviation. The (111)-oriented specimen, containing a block boundary, revealed a significant effect: CRSS and saturation stress increased by 29% and 58%, respectively. Finally, a possible non-Schmid effect was observed, though not confirmed. The second objective was investigated using specimens of varying dimensions, from macro to micro-scale, all extracted from the same material batch after austenitization at 1050 °C for one hour. This heat treatment was selected to produce martensite free of reformed austenite, thereby eliminating it as a potential mechanism. Plasticity onset was identified using the Onset of Plasticity Relaxation Analysis (OPRA) method. The experimental protocol examined the influence of reducing type-II residual stresses by significantly decreasing specimen volume, as well as the anisotropy of mechanical properties at the microscale. A macroscopic specimen tempered at 500 °C for two hours was also tested to assess stress relaxation by an alternative approach. Results convincingly demonstrate that type-II residual stresses are the dominant mechanism governing microplasticity at low load in as-quenched UNS S41500. Specifically, these stresses rationalize the discrepancy between CRSS estimated in a macroscopic specimen and that measured experimentally in a microscopic specimen under in-lath slip configuration. The residual stress tensor used in this work originates from quenched Fe-0.27C-2.3Mn-0.1Cr 1.47Si steel. Calculations of other mechanisms proposed in the literature do not support their significant contribution to low-load nonlinearity. In summary, this thesis advances knowledge of lath martensite micromechanical behavior in two key ways. First, it provides new, directly measured CRSS and saturation stress values for a lath block, enabling improved constitutive modeling in CPFEM. In second, it demonstrates that type-II residual stresses are the primary mechanism driving low-load nonlinearity in as-quenched lath martensite—a phenomenon first observed nearly 70 years ago.

Date	16 Jan 2026
Original language	French
Awarding Institution	École de technologie supérieure
Supervisor	Henri Champliaud (Supervisor), Daniel Paquet (Co-supervisor) & Jacques Lanteigne (Co-supervisor)