Abstract:X-ray diffraction is ideal for probing sub-surface state during complex or rapid thermomechanical loading of crystalline materials. However, challenges arise as the size of diffraction volumes increases due to spatial broadening and inability to deconvolute the effects of different lattice deformation mechanisms. Here, we present a novel approach to use combinations of physics-based modeling and machine learning to deconvolve thermal and mechanical elastic strains for diffraction data analysis. The method builds on a previous effort to extract thermal strain distribution information from diffraction data. The new approach is applied to extract the evolution of thermomechanical state during laser melting of an Inconel 625 wall specimen which produces significant residual stress upon cooling. A combination of heat transfer and fluid flow, elasto-plasticity, and X-ray diffraction simulations are used to generate training data for machine-learning (Gaussian Process Regression, GPR) models that map diffracted intensity distributions to underlying thermomechanical strain fields. First-principles density functional theory is used to determine accurate temperature-dependent thermal expansion and elastic stiffness used for elasto-plasticity modeling. The trained GPR models are found to be capable of deconvoluting the effects of thermal and mechanical strains, in addition to providing information about underlying strain distributions, even from complex diffraction patterns with irregularly shaped peaks.

What problem does this paper attempt to address?

### Problems the Paper Aims to Solve The paper aims to address the challenges encountered when probing the subsurface state of crystalline materials using X-ray diffraction (XRD) under complex or rapid thermomechanical loading. Specifically, as the diffraction volume increases, the analysis becomes difficult due to spatial broadening and the inability to distinguish between different lattice deformation mechanisms. The paper proposes a new method that combines physical modeling and machine learning to separate thermoelastic and mechanical elastic strains from diffraction data. ### Main Challenges 1. **Spatial Broadening**: As the diffraction volume increases, the effect of spatial broadening makes it more difficult to resolve local strain distributions. 2. **Strain Mechanism Confusion**: Both thermal stress and mechanical stress alter the lattice state. Particularly in cubic crystal systems, these two stresses have the same effect on the volumetric part of the lattice strain tensor, making them hard to distinguish. 3. **Complex Strain Distribution**: When using high-energy X-rays in thick samples, the distribution of thermal and mechanical strains can be very broad, and a single diffraction image cannot directly reconstruct these distributions. ### Solutions 1. **Machine Learning Model**: Utilize a Gaussian Process Regression (GPR) model to map the underlying thermomechanical strain field from the diffraction intensity distribution. 2. **Physical Modeling**: Combine heat transfer, fluid flow, elastoplastic models, and X-ray diffraction simulations to generate training data. 3. **First-Principles Calculations**: Use Density Functional Theory (DFT) to determine temperature-dependent thermal expansion coefficients and elastic moduli for elastoplastic modeling. 4. **Transfer Learning**: Improve the model based on previous work, including using DFT to calculate temperature-dependent thermomechanical properties, modeling stress distribution in elastoplastic models, updating the X-ray diffraction modeling framework, and enhancing GPR model training. ### Application Case The method is applied to the laser melting process of an Inconel 625 wall sample to extract the evolution information of the thermomechanical state. Through this method, researchers can extract the distribution information of thermal and mechanical strains from complex diffraction patterns, even if these patterns contain irregularly shaped peaks. ### Keywords Synchrotron X-ray diffraction, stress, superalloy, machine learning, physical modeling, first-principles calculations, Gaussian Process Regression ### Summary The paper proposes a new method that combines physical modeling and machine learning to separate thermal and mechanical strains from X-ray diffraction data, addressing the challenges of analyzing lattice deformation mechanisms under complex thermomechanical loading conditions. This method performs well in handling high temperatures and complex strain distributions, providing a powerful tool for understanding material behavior during processes such as laser melting.

Deconvoluting Thermomechanical Effects in X-ray Diffraction Data using Machine Learning

Deconvoluting Thermomechanical Effects in X-ray Diffraction Data using Machine Learning

Combining synchrotron X‐ray diffraction, mechanistic modeling and machine learning for in situ subsurface temperature quantification during laser melting

Harnessing interpretable and unsupervised machine learning to address big data from modern X-ray diffraction

Combining Synchrotron X-ray Diffraction, Mechanistic Modeling, and Machine Learning for In Situ Subsurface Temperature Quantification during Laser Melting

Machine learning atomic dynamics to unfold the origin of plasticity in metallic glasses: From thermo- to acousto-plastic flow

Data-driven approach for synchrotron X-ray Laue microdiffraction scan analysis

Artifact Identification in X-ray Diffraction Data using Machine Learning Methods

Machine learning-enabled identification of micromechanical stress and strain hotspots predicted via dislocation density-based crystal plasticity simulations

Rapid detection of rare events from in situ X-ray diffraction data using machine learning

Automated Analysis Framework of Strain Partitioning and Deformation Mechanisms Via Multimodal Fusion and Computer Vision

Visualizing Three-Dimensional Micromechanical Response in Nanomaterials

Uncovering Material Deformations via Machine Learning Combined with Four-Dimensional Scanning Transmission Electron Microscopy

Deep Learning of Structural Morphology Imaged by Scanning X-ray Diffraction Microscopy

Adaptively driven X-ray diffraction guided by machine learning for autonomous phase identification

Computing virtual dark-field X-ray microscopy images of complex discrete dislocation structures from large-scale molecular dynamics simulations

Artifact identification in X‐ray diffraction data using machine learning methods

Formalism to Image the Dynamics of Coherent and Incoherent Phonon with Dark-Field X-ray Microscopy using Kinematic Diffraction Theory

X. Advanced Technologies of Materials Characterization

Diffraction-Based Multiscale Residual Strain Measurements

Study on hot deformation behavior and recrystallization mechanism of an Al-6.3Zn-2.5Mg-2.6Cu-0.11Zr alloy based on machine learning