Abstract:This contribution focuses on the development of Model Order Reduction (MOR) for one-way coupled steady state linear thermomechanical problems in a finite element setting. We apply Proper Orthogonal Decomposition (POD) for the computation of reduced basis space. On the other hand, for the evaluation of the modal coefficients, we use two different methodologies: the one based on the Galerkin projection (G) and the other one based on Artificial Neural Network (ANN). We aim at comparing POD-G and POD-ANN in terms of relevant features including errors and computational efficiency. In this context, both physical and geometrical parametrization are considered. We also carry out a validation of the Full Order Model (FOM) based on customized benchmarks in order to provide a complete computational pipeline. The framework proposed is applied to a relevant industrial problem related to the investigation of thermomechanical phenomena arising in blast furnace hearth walls. Keywords: Thermomechanical problems, Finite element method, Proper orthogonal decomposition, Galerkin projection, Artificial neural network, Geometric and physical parametrization, Blast furnace.

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is to accelerate the solution of parameterized one - dimensional coupled steady - state linear thermo - mechanical problems through Model Order Reduction (MOR) while maintaining acceptable accuracy. Specifically, the research aims to develop a MOR framework based on the Finite Element Method (FEM) for efficient calculation of thermo - mechanical phenomena under different physical and geometric parameters. ### Problem Background 1. **Requirement for High - Fidelity Numerical Computation**: In recent years, the development of technology has made it necessary to perform high - fidelity numerical computations for many configurations, which has put great pressure on computing resources. 2. **Requirement for Model Order Reduction**: In order to accelerate the computation while ensuring accuracy, the Model Order Reduction (MOR) method is introduced to reduce the computational complexity and improve efficiency. ### Research Objectives - **Develop MOR Framework**: Develop a MOR framework in a finite - element environment for one - dimensional coupled steady - state linear thermo - mechanical problems. - **Compare Different Methods**: Use two different methods to calculate modal coefficients - Galerkin projection (POD - G) and artificial neural network (POD - ANN), and compare them in terms of error and computational efficiency. - **Verify in Industrial Applications**: Apply this framework to model the thermo - mechanical phenomena of the blast furnace bottom wall to verify its effectiveness and practicality. ### Specific Contents 1. **Selection of Model Order Reduction Methods**: - Use Proper Orthogonal Decomposition (POD) to construct the reduced - order basis space. - For the calculation of modal coefficients, use Galerkin projection (POD - G) and artificial neural network (POD - ANN) respectively. 2. **Handling of Parameterized Problems**: - Consider the changes of physical parameters (such as the thermal conductivity and thermal expansion coefficient of materials, etc.) and geometric parameters (such as the diameters and thicknesses of various parts of the furnace bottom wall). 3. **Performance Evaluation**: - Compare the performance of POD - G and POD - ANN in terms of error and computational efficiency. - Verify the effectiveness of the Full Order Model (FOM) through customized benchmark tests. ### Application Scenarios This research particularly focuses on modeling the thermo - mechanical phenomena in the blast furnace bottom wall, aiming to provide an efficient computational tool for industrial applications. By introducing deep - learning techniques (such as artificial neural networks), the research hopes to open up new ways for the modeling of multi - physical - field problems. ### Summary The core problem of the paper is how to accelerate the solution of complex thermo - mechanical problems through the MOR method while maintaining sufficient accuracy, and verify the effectiveness of this method through actual industrial cases.

Finite element based model order reduction for parametrized one-way coupled steady state linear thermomechanical problems

Model reduction applied to square to rectangular martensitic transformations using proper orthogonal decomposition

Fast Model Order Reduction Via Nonlinear Optimization

Low‐dimensional Modeling of Linear Heat Transfer Systems Using Incremental the Proper Orthogonal Decomposition Method

Model Order Reduction for High Dimensional Linear Systems Based on Rank-1 Incremental Proper Orthogonal Decomposition.

Model order reduction of thermo-mechanical models with parametric convective boundary conditions: focus on machine tool

A local basis approximation approach for nonlinear parametric model order reduction

A multi-field decomposed model order reduction approach for thermo-mechanically coupled gradient-extended damage simulations

Parametric POD-Galerkin Model Order Reduction for Unsteady-State Heat Transfer Problems

Thermomechanical modelling for industrial applications

A reduced-order modeling for thermo-mechanical coupling analyses by using radial integration boundary element method

POD-based reduced order methods for optimal control problems governed by parametric partial differential equation with varying boundary control

A manifold learning approach to nonlinear model order reduction of quasi-static problems in solid mechanics

Parametric reduced-order modeling for component-oriented treatment and localized nonlinear feature inclusion

Nonlinear Reduced Order Modeling of Heated Structures with Temperature-Dependent Properties

Structured Optimization-Based Model Order Reduction for Parametric Systems

Neural-Network-Augmented Projection-Based Model Order Reduction for Mitigating the Kolmogorov Barrier to Reducibility of CFD Models

Nonlinear model order reduction for problems with microstructure using mesh informed neural networks

An optimisation-based domain-decomposition reduced order model for parameter-dependent non-stationary fluid dynamics problems

Component based model order reduction with mortar tied contact for nonlinear quasi-static mechanical problems

Comparison of Reduction Methods for Finite Element Geometrically Nonlinear Beam Structures