Abstract:In this work, we have developed a variational Bayesian inference theory of elasticity, which is accomplished by using a mixed Variational Bayesian inference Finite Element Method (VBI-FEM) that can be used to solve the inverse deformation problems of continua. In the proposed variational Bayesian inference theory of continuum mechanics, the elastic strain energy is used as a prior in a Bayesian inference network, which can intelligently recover the detailed continuum deformation mappings with only given the information on the deformed and undeformed continuum body shapes without knowing the interior deformation and the precise actual boundary conditions, both traction as well as displacement boundary conditions, and the actual material constitutive relation. Moreover, we have implemented the related finite element formulation in a computational probabilistic mechanics framework. To numerically solve mixed variational problem, we developed an operator splitting or staggered algorithm that consists of the finite element (FE) step and the Bayesian learning (BL) step as an analogue of the well-known the Expectation-Maximization (EM) algorithm. By solving the mixed probabilistic Galerkin variational problem, we demonstrated that the proposed method is able to inversely predict continuum deformation mappings with strong discontinuity or fracture without knowing the external load conditions. The proposed method provides a robust machine intelligent solution for the long-sought-after inverse problem solution, which has been a major challenge in structure failure forensic pattern analysis in past several decades. The proposed method may become a promising artificial intelligence-based inverse method for solving general partial differential equations.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is, given the shapes of a continuum before and after deformation, how to recover the three - dimensional deformation mapping of the continuum without fully knowing the boundary conditions (such as load conditions and displacement boundary conditions) and the material constitutive relations. Specifically, the goal of the paper is to solve the inverse deformation problem in continuum mechanics through variational Bayesian inference theory and the hybrid probabilistic finite element method (VBI - FEM). ### Problem Background In continuum mechanics, especially in elasticity theory, an object deforms from an initial configuration $\Omega_X$ to a current configuration $\Omega_x$ under the action of external forces. The deformation process can be described by a continuous function $\psi: \Omega_X\rightarrow\Omega_x$, which is called the deformation mapping or mapping. The traditional forward problem can be solved for the deformation mapping by known boundary conditions and material properties. However, in many engineering applications, we do not know the exact boundary conditions or material properties, but only have the shape information before and after deformation. ### Challenges of the Inverse Problem The core of the inverse problem is: based only on the shape information before and after deformation, how to recover the detailed three - dimensional deformation field without knowing the internal deformation, the actual boundary conditions (including traction and displacement boundary conditions), and the specific material constitutive relations. This problem is of great significance in fields such as structural failure analysis and material damage assessment, but it has been a major challenge in the past few decades. ### Solutions in the Paper To solve this problem, the author proposes an elasticity mechanics method based on variational Bayesian inference theory and combines it with the hybrid probabilistic finite element method (VBI - FEM). The main features of this method are as follows: 1. **Variational Bayesian Inference**: The elastic strain energy is introduced as prior information into the Bayesian inference network, thereby intelligently recovering the deformation mapping of the continuum. 2. **Hybrid Probabilistic Finite Element Method**: An operator splitting algorithm is developed, which includes a finite element step (FE - step) and a Bayesian learning step (BL - step), similar to the classical expectation - maximization (EM) algorithm. 3. **Handling Complex Deformations**: This method can handle deformation problems with strong discontinuities or fractures without knowing the external load conditions. ### Mathematical Expression This method is achieved by minimizing the probabilistic elastic entropy potential energy. The specific formula is as follows: \[ (u^\star, \Sigma^\star)=\arg\min_{\tilde{u}, \Sigma\in V\oplus\mathbb{R}^+}\tilde{\Pi}(\tilde{u}, \Sigma) \] where the probabilistic potential energy $\tilde{\Pi}(\tilde{u}, \Sigma)$ is defined as: \[ \tilde{\Pi}(\tilde{u}, \Sigma)=-\int_{\Omega_x}\log\left(\int_{\Omega_X}p(\tilde{X}|\tilde{u}, \Sigma)p(x|\tilde{X}; \tilde{u}, \Sigma)d\Omega_X\right)d\Omega_x+\gamma\int_{\Omega_X}\phi(\tilde{u})d\Omega_X \] Here, $\phi(\tilde{u}) = \frac{1}{2}\tilde{u}^2$ is a regularization function for stabilizing the calculation; $\tilde{X}=X + \tilde{u}$, and $\gamma$ is a hyperparameter. ### Conclusion This method provides a powerful machine - intelligence solution that can solve the inverse problems that have long plagued fields such as structural failure analysis, especially when only the shape information before and after deformation is known.

A Variational Bayesian Inference Theory of Elasticity and Its Mixed Probabilistic Finite Element Method for Inverse Deformation Solutions in Any Dimension

An Inverse Model and Mathematical Solution for Inferring Viscoelastic Properties and Dynamic Deformations of Heterogeneous Structures

A Bayesian framework for constitutive model identification via use of full field measurements, with application to heterogeneous materials

Evolutionary-variational method in mathematical plasticity

Inverse Modeling of Viscoelasticity Materials using Physics Constrained Learning

Mixed Fem Formulation Within the Toupin–Mindlin Gradient Elasticity Theory: Validity Analysis

A variational minimization formulation for hydraulically induced fracturing in elastic-plastic solids

Finite Element Approximation of Finite Deformation Dislocation Mechanics

Error Estimates of a Fully Discrete Multiphysics Finite Element Method for a Nonlinear Poroelasticity Model

A TPAA-SBFEM based full-scale forward and inverse interval analysis for dynamic viscoelastic problems

Towards a Generalized Approach to Nonlocal Elasticity via Fractional-Order Mechanics

A Unified Determinant-Preserving Formulation for Compressible/Incompressible Finite Viscoelasticity

Inferring Displacement Fields from Sparse Measurements Using the Statistical Finite Element Method

Reduced basis approximation and a posteriori error estimation: applications to elasticity problems in several parametric settings

Inverse deformation analysis: an experimental and numerical assessment using the FEniCS Project

On Boundary Condition Implementation Via Variational Principles in Elasticity-Based Homogenization

On the use of constrained reactive mixtures of solids to model finite deformation isothermal elastoplasticity and elastoplastic damage mechanics

A finite element implementation of finite deformation surface and bulk poroelasticity

A mechanics-informed deep learning framework for data-driven nonlinear viscoelasticity

Large deformations in terms of stretch and rotation and global solution to the quasi-stationary problem

Well-posedness of evolutionary differential variational–hemivariational inequalities and applications to frictional contact mechanics