Jae Hyuk Lim
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Jae Hyuk Lim has a total of 81 co-authors. Here are the top co-authors by number of shared publications:
Hong-Kyun Noh
6 shared publicationsJeong-Hoon Park
4 shared publicationsMyeong-Seok Go
Jeonbuk Natl Univ, Dept Mech Engn, 567 Baekje Daero, Jeonju Si 54896, Jeollabuk Do, South Korea
4 shared publications- Seunghwa Ryu
Department of Mechanical Engineering Korea Advanced Institute of Science and Technology (KAIST) Daejeon Republic of Korea
4 shared publications Do-Won Kim
Jeonbuk Natl Univ, Dept Mech Engn, 567 Baekje Daero, Jeonju Si 54896, Jeollabuk Do, South Korea
3 shared publicationsDonggeun Park
3 shared publications
Latest Publications
- Open Access
- 5 October 2026
- Published by Elsevier in Materials Today Communications
- Vol. 57
RVE-derived Poisson-coupled equivalent fabric modeling of metallic wire meshes
- 5 October 2026
- Published by Elsevier in Materials Today Communications
- Vol. 57
Open AccessRVE-derived Poisson-coupled equivalent fabric modeling of metallic wire meshes
- 29 September 2026
- Published by Elsevier in Aerospace Science and Technology
- Vol. 180
Design and flight demonstration of an ornithopter: Enhancing endurance and payload via neural network surrogate-based optimization
- 29 September 2026
- Published by Elsevier in Aerospace Science and Technology
- Vol. 180
Design and flight demonstration of an ornithopter: Enhancing endurance and payload via neural network surrogate-based optimization
- Open Access
- 30 April 2026
- Published by Computational Structural Engineering Institute of Korea in Journal of the Computational Structural Engineering Institute of Korea
- Vol. 39 (2) , 123-135
Development of an LLM-Driven Abaqus MCP-Integrated Finite Element Analysis Automation System with Hallucination Suppression and Performance Improvement
- 30 April 2026
- Published by Computational Structural Engineering Institute of Korea in Journal of the Computational Structural Engineering Institute of Korea
- Vol. 39 (2) , 123-135
Open AccessDevelopment of an LLM-Driven Abaqus MCP-Integrated Finite Element Analysis Automation System with Hallucination Suppression and Performance Improvement
- Top Cited Papers
- 13 August 2025
- Published by Elsevier in Computer Methods in Applied Mechanics and Engineering
- Vol. 446
Physics-informed neural network-based discovery of hyperelastic constitutive models from extremely scarce data
- 13 August 2025
- Published by Elsevier in Computer Methods in Applied Mechanics and Engineering
- Vol. 446
Top Cited PapersPhysics-informed neural network-based discovery of hyperelastic constitutive models from extremely scarce data
- Open Access
Real-time physics-informed reconstruction of transient fields using sensor guidance and higher-order time differentiation
This study proposes FTI-PBSM (Fixed-Time-Increment Physics-informed neural network-Based Surrogate Model), a novel physics-informed surrogate modeling framework designed for real-time reconstruction of transient responses in time-dependent Partial Differential Equations (PDEs) using only sparse, time-dependent sensor measurements. Unlike conventional Physics-Informed Neural Network (PINN)-based models that rely on Automatic Differentiation (AD) over both spatial and temporal domains and require dedicated causal network architectures to impose temporal causality, the proposed approach entirely removes AD in the time direction. Instead, it leverages higher-order numerical differentiation methods, such as the Central Difference, Adams-Bashforth, and Backward Differentiation Formula, to explicitly impose temporal causality. This leads to a simplified model architecture with improved training stability, computational efficiency, and extrapolation capability. Furthermore, FTI-PBSM is trained on sparse sensor measurements from multiple PDE cases generated by varying PDE coefficients, with the sensor data serving as model input. This enables the model to learn a parametric PDE family and generalize to unseen physical cases, accurately reconstructing full-field transient solutions in real time. The proposed model is validated on four representative PDE problems-the convection equation, diffusion-reaction dynamics, Korteweg-de Vries (KdV) equation, and Allen-Cahn equation-and demonstrates superior prediction accuracy and generalization performance compared to a causal PBSM, which is used as the baseline model, in both interpolation and extrapolation tasks. It also shows strong robustness to sensor noise and variations in training data size, while significantly reducing training time.
Open AccessReal-time physics-informed reconstruction of transient fields using sensor guidance and higher-order time differentiation
This study proposes FTI-PBSM (Fixed-Time-Increment Physics-informed neural network-Based Surrogate Model), a novel physics-informed surrogate modeling framework designed for real-time reconstruction of transient responses in time-dependent Partial Differential Equations (PDEs) using only sparse, time-dependent sensor measurements. Unlike conventional Physics-Informed Neural Network (PINN)-based models that rely on Automatic Differentiation (AD) over both spatial and temporal domains and require dedicated causal network architectures to impose temporal causality, the proposed approach entirely removes AD in the time direction. Instead, it leverages higher-order numerical differentiation methods, such as the Central Difference, Adams-Bashforth, and Backward Differentiation Formula, to explicitly impose temporal causality. This leads to a simplified model architecture with improved training stability, computational efficiency, and extrapolation capability. Furthermore, FTI-PBSM is trained on sparse sensor measurements from multiple PDE cases generated by varying PDE coefficients, with the sensor data serving as model input. This enables the model to learn a parametric PDE family and generalize to unseen physical cases, accurately reconstructing full-field transient solutions in real time. The proposed model is validated on four representative PDE problems-the convection equation, diffusion-reaction dynamics, Korteweg-de Vries (KdV) equation, and Allen-Cahn equation-and demonstrates superior prediction accuracy and generalization performance compared to a causal PBSM, which is used as the baseline model, in both interpolation and extrapolation tasks. It also shows strong robustness to sensor noise and variations in training data size, while significantly reducing training time.
- Open Access
- 2 June 2025
- Published by Springer Nature in JMST Advances
- Vol. 7 (3) , 217
Correction: Physics-informed machine learning: technological trends in turbulence flow, topology optimization, and convergence enhancement with temporal causality consideration
- 2 June 2025
- Published by Springer Nature in JMST Advances
- Vol. 7 (3) , 217
Open AccessCorrection: Physics-informed machine learning: technological trends in turbulence flow, topology optimization, and convergence enhancement with temporal causality consideration
- 15 May 2025
- Published by Springer Nature in JMST Advances
- Vol. 7 (1-2) , 113-124
Physics-informed machine learning: technological trends in turbulence flow, topology optimization, and convergence enhancement with temporal causality consideration
- 15 May 2025
- Published by Springer Nature in JMST Advances
- Vol. 7 (1-2) , 113-124
Physics-informed machine learning: technological trends in turbulence flow, topology optimization, and convergence enhancement with temporal causality consideration
- Open Access
Physics-Informed Neural Network-Based Discovery of Hyperelastic Constitutive Models from Extremely Scarce Data
The discovery of constitutive models for hyperelastic materials is essential yet challenging due to their nonlinear behavior and the limited availability of experimental data. Traditional methods typically require extensive stress-strain or full-field measurements, which are often difficult to obtain in practical settings. To overcome these challenges, we propose a physics-informed neural network (PINN)-based framework that enables the discovery of constitutive models using only sparse measurement data - such as displacement and reaction force - that can be acquired from a single material test. By integrating PINNs with finite element discretization, the framework reconstructs full-field displacement and identifies the underlying strain energy density from predefined candidates, while ensuring consistency with physical laws. A two-stage training process is employed: the Adam optimizer jointly updates neural network parameters and model coefficients to obtain an initial solution, followed by L-BFGS refinement and sparse regression with l_p regularization to extract a parsimonious constitutive model. Validation on benchmark hyperelastic models demonstrates that the proposed method can accurately recover constitutive laws and displacement fields, even when the input data are limited and noisy. These findings highlight the applicability of the proposed framework to experimental scenarios where measurement data are both scarce and noisy.
Open AccessPhysics-Informed Neural Network-Based Discovery of Hyperelastic Constitutive Models from Extremely Scarce Data
The discovery of constitutive models for hyperelastic materials is essential yet challenging due to their nonlinear behavior and the limited availability of experimental data. Traditional methods typically require extensive stress-strain or full-field measurements, which are often difficult to obtain in practical settings. To overcome these challenges, we propose a physics-informed neural network (PINN)-based framework that enables the discovery of constitutive models using only sparse measurement data - such as displacement and reaction force - that can be acquired from a single material test. By integrating PINNs with finite element discretization, the framework reconstructs full-field displacement and identifies the underlying strain energy density from predefined candidates, while ensuring consistency with physical laws. A two-stage training process is employed: the Adam optimizer jointly updates neural network parameters and model coefficients to obtain an initial solution, followed by L-BFGS refinement and sparse regression with l_p regularization to extract a parsimonious constitutive model. Validation on benchmark hyperelastic models demonstrates that the proposed method can accurately recover constitutive laws and displacement fields, even when the input data are limited and noisy. These findings highlight the applicability of the proposed framework to experimental scenarios where measurement data are both scarce and noisy.
- 7 January 2025
- Published by Wiley in Advanced Theory and Simulations
- Vol. 8 (4)
Enhancing Prediction Performance and Generalizing for Transverse Behavior of Unidirectional Composites via Strategic Input Feature Augmentation
This study proposes an efficient method for analyzing complex fracture patterns in the cross‐sections of unidirectional (UD) composites, influenced by the volume fraction (VF) and fiber arrangement, and for predicting the corresponding transverse mechanical responses. Traditional finite element (FE) analysis incurs high computational costs when evaluating responses for every configuration. To address this, deep learning (DL), particularly convolutional neural networks (CNNs), have been applied, but these approaches have typically focused on limited VF spaces, leading to large data requirements and reduced prediction accuracy for new configurations. In this research, a novel DL approach is introduced that can be effectively extended to broader VF spaces by integrating low‐cost, physically insightful auxiliary features as multi‐modal inputs. Specifically, the Mori–Tanaka (MT) feature and stress concentration factor (SCF) are selected as auxiliary inputs and incorporated them into the conventional CNN model for comparative analysis. The results showed that the model incorporating the MT feature significantly improved extrapolation performance in unseen VF spaces while maintaining robust predictive performance as the training dataset size increased. In contrast, the SCF feature does not demonstrate similar benefits. These findings illustrate that integrating advanced features like the MT feature into the DL model can offer more effective and versatile solutions for predicting material properties.
- 7 January 2025
- Published by Wiley in Advanced Theory and Simulations
- Vol. 8 (4)
Enhancing Prediction Performance and Generalizing for Transverse Behavior of Unidirectional Composites via Strategic Input Feature Augmentation
This study proposes an efficient method for analyzing complex fracture patterns in the cross‐sections of unidirectional (UD) composites, influenced by the volume fraction (VF) and fiber arrangement, and for predicting the corresponding transverse mechanical responses. Traditional finite element (FE) analysis incurs high computational costs when evaluating responses for every configuration. To address this, deep learning (DL), particularly convolutional neural networks (CNNs), have been applied, but these approaches have typically focused on limited VF spaces, leading to large data requirements and reduced prediction accuracy for new configurations. In this research, a novel DL approach is introduced that can be effectively extended to broader VF spaces by integrating low‐cost, physically insightful auxiliary features as multi‐modal inputs. Specifically, the Mori–Tanaka (MT) feature and stress concentration factor (SCF) are selected as auxiliary inputs and incorporated them into the conventional CNN model for comparative analysis. The results showed that the model incorporating the MT feature significantly improved extrapolation performance in unseen VF spaces while maintaining robust predictive performance as the training dataset size increased. In contrast, the SCF feature does not demonstrate similar benefits. These findings illustrate that integrating advanced features like the MT feature into the DL model can offer more effective and versatile solutions for predicting material properties.