Bhagyashree Prabhune
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
Research Subjects
Bhagyashree Prabhune has a total of 52 co-authors. Here are the top co-authors by number of shared publications:
- Krishnan Suresh
Department of Mechanical Engineering, University of Wisconsin-Madison, Madison, USA
6 shared publications Oak Ridge Oak Ridge National Laboratory (Ornl)
3 shared publications- Yousub Lee
Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
3 shared publications - Brian Jordan
Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
2 shared publications Madhura Limaye
Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
2 shared publications- Saketh Sridhara
University of Wisconsin–Madison Department of Mechanical Engineering
2 shared publications
Latest Publications
- 1 March 2027
- Published by Office of Scientific and Technical Information (OSTI)
Residual Stress Development and Deformation of Wire-Arc Direct Energy Deposition Parts
Wire-arc direct energy deposition (Wire-arc DED) enables efficient fabrication of large metallic components because of its high deposition rate and high material efficiency. However conventional Wirearc DED relies on a build plate to constrain part deformation during deposition, which requires postprocessing removal and can increase material waste and residual strain redistribution. This project investigated the feasibility of shifting Wire-arc DED toward build plate free, freeform fabrication by examining residual strain and distortion development in three walls: (1) as-deposited, (b) base plate removal, and (3) freeform wall printed without a base plate using a small anchor support. Residual strains were measured using neutron diffraction method at ORNL’s High Flux Isotope Reactor (HFIR). A sequentially coupled thermo-mechanical finite element framework was developed to simulate deposition, cooling, boundary condition changes associated with base plate removal and freeform deposition. The model captured characteristic strain bands, edge tension, mid-height compression, and strain re-distribution after base plate removal. Statistical comparisons using standard deviation, RMSE, normalized RMSE, and bias showed better quantitative agreement for the base-plate and base plate removal cases than for the freeform wall although the freeform simulation still reasonably captured the overall strain magnitude and trend. Overall, the predicted and measured results demonstrate the capability of a physics-based, 5-axis tool path aware modeling framework for residual strain prediction in Wire-arc DED. Also, the work identified key modeling gaps including free surface deformation, bottom layer sagging, material hardening behavior, 5-axis associated heat distribution, and geometric uncertainty and discussed for improved residual strain prediction.
- 1 March 2027
- Published by Office of Scientific and Technical Information (OSTI)
Residual Stress Development and Deformation of Wire-Arc Direct Energy Deposition Parts
Wire-arc direct energy deposition (Wire-arc DED) enables efficient fabrication of large metallic components because of its high deposition rate and high material efficiency. However conventional Wirearc DED relies on a build plate to constrain part deformation during deposition, which requires postprocessing removal and can increase material waste and residual strain redistribution. This project investigated the feasibility of shifting Wire-arc DED toward build plate free, freeform fabrication by examining residual strain and distortion development in three walls: (1) as-deposited, (b) base plate removal, and (3) freeform wall printed without a base plate using a small anchor support. Residual strains were measured using neutron diffraction method at ORNL’s High Flux Isotope Reactor (HFIR). A sequentially coupled thermo-mechanical finite element framework was developed to simulate deposition, cooling, boundary condition changes associated with base plate removal and freeform deposition. The model captured characteristic strain bands, edge tension, mid-height compression, and strain re-distribution after base plate removal. Statistical comparisons using standard deviation, RMSE, normalized RMSE, and bias showed better quantitative agreement for the base-plate and base plate removal cases than for the freeform wall although the freeform simulation still reasonably captured the overall strain magnitude and trend. Overall, the predicted and measured results demonstrate the capability of a physics-based, 5-axis tool path aware modeling framework for residual strain prediction in Wire-arc DED. Also, the work identified key modeling gaps including free surface deformation, bottom layer sagging, material hardening behavior, 5-axis associated heat distribution, and geometric uncertainty and discussed for improved residual strain prediction.
- 3 November 2026
- Published by Elsevier in Composites Part B: Engineering
- Vol. 330
AI-enabled large-format additive manufacturing of thermoplastics: Recent advances and future directions
- 3 November 2026
- Published by Elsevier in Composites Part B: Engineering
- Vol. 330
AI-enabled large-format additive manufacturing of thermoplastics: Recent advances and future directions
- Open Access
- 17 June 2026
- Published by Elsevier in SSRN Electronic Journal
Physics-informed Tandem Neural Network for Rapid and Accurate Temperature History Prediction in Polymer Composite Extrusion-based Large-Format Additive Manufacturing
Large-format additive manufacturing (LFAM) using polymer composites enables the fabrication of massive, load-bearing structures without the need for traditional tooling. However, the continuous deposition of large thermal masses introduces severe temperature management challenges. Improper layer times lead to inadequate temperature histories, resulting in poor interlayer bonding or structural collapse. Real-time temperature history prediction is critical for thermal management, yet high-fidelity finite element analysis is too computationally expensive, and purely data-driven models lack physical consistency and generalizability without massive datasets. To overcome these limitations, this study proposes a novel Physics-Informed Tandem Neural Network (PITNN) framework to rapidly and accurately predict the temperature history during the LFAM using polymer composites. The PITNN architecture couples an Inverse Model, which identifies latent boundary conditions from an initial 30-second temperature observation, with a Forward PINN that embeds physical loss to predict the entire temperature history. To resolve the simulation-to-reality gap, the network is fine-tuned using experimental measurements. When compared to experimental data, the baseline 1D reduced-order model exhibited mean absolute errors (MAE) between 7.2°C and 12.2°C, while the un-tuned PITNN showed an MAE of 11.7°C to 13.0°C. Following experimental fine-tuning, the PITNN successfully captured the complex real-world thermal dynamics, drastically reducing the MAE to 2.2–2.5°C with an R² exceeding 0.98. Crucially, the model performs these predictions with an inference time of just 2.3 ms. This speed and accuracy demonstrate that the PITNN is an effective digital twin framework, fully capable of enabling real-time layer time optimization and closed-loop thermal control in extrusion-based LFAM.
- 17 June 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessPhysics-informed Tandem Neural Network for Rapid and Accurate Temperature History Prediction in Polymer Composite Extrusion-based Large-Format Additive Manufacturing
Large-format additive manufacturing (LFAM) using polymer composites enables the fabrication of massive, load-bearing structures without the need for traditional tooling. However, the continuous deposition of large thermal masses introduces severe temperature management challenges. Improper layer times lead to inadequate temperature histories, resulting in poor interlayer bonding or structural collapse. Real-time temperature history prediction is critical for thermal management, yet high-fidelity finite element analysis is too computationally expensive, and purely data-driven models lack physical consistency and generalizability without massive datasets. To overcome these limitations, this study proposes a novel Physics-Informed Tandem Neural Network (PITNN) framework to rapidly and accurately predict the temperature history during the LFAM using polymer composites. The PITNN architecture couples an Inverse Model, which identifies latent boundary conditions from an initial 30-second temperature observation, with a Forward PINN that embeds physical loss to predict the entire temperature history. To resolve the simulation-to-reality gap, the network is fine-tuned using experimental measurements. When compared to experimental data, the baseline 1D reduced-order model exhibited mean absolute errors (MAE) between 7.2°C and 12.2°C, while the un-tuned PITNN showed an MAE of 11.7°C to 13.0°C. Following experimental fine-tuning, the PITNN successfully captured the complex real-world thermal dynamics, drastically reducing the MAE to 2.2–2.5°C with an R² exceeding 0.98. Crucially, the model performs these predictions with an inference time of just 2.3 ms. This speed and accuracy demonstrate that the PITNN is an effective digital twin framework, fully capable of enabling real-time layer time optimization and closed-loop thermal control in extrusion-based LFAM.
- Open Access
- 1 February 2024
- Published by Office of Scientific and Technical Information (OSTI)
Integrated Process and Materials Modeling for Development of Additive Manufacturing of Refractory Materials for Critical Applications
The U.S. Department of Energy's Office of Scientific and Technical Information
- 1 February 2024
- Published by Office of Scientific and Technical Information (OSTI)
Open AccessIntegrated Process and Materials Modeling for Development of Additive Manufacturing of Refractory Materials for Critical Applications
The U.S. Department of Energy's Office of Scientific and Technical Information
- 28 October 2023
- Published by Springer Nature in Engineering with Computers
- Vol. 40 (3) , 1357-1374
On why mesh untangling may not be required
- 28 October 2023
- Published by Springer Nature in Engineering with Computers
- Vol. 40 (3) , 1357-1374
On why mesh untangling may not be required
- 5 July 2023
- Published by Springer Nature in Computational Mechanics
- Vol. 73 (1) , 159-176
An isoparametric tangled finite element method for handling higher-order elements with negative Jacobian
- 5 July 2023
- Published by Springer Nature in Computational Mechanics
- Vol. 73 (1) , 159-176
An isoparametric tangled finite element method for handling higher-order elements with negative Jacobian
- Open Access
Isoparametric Tangled Finite Element Method for Nonlinear Elasticity
An important requirement in the standard finite element method (FEM) is that all elements in the underlying mesh must be tangle-free i.e., the Jacobian must be positive throughout each element. To relax this requirement, an isoparametric tangled finite element method (i-TFEM) was recently proposed for linear elasticity problems. It was demonstrated that i-TFEM leads to optimal convergence even for severely tangled meshes. In this paper, i-TFEM is generalized to nonlinear elasticity. Specifically, a variational formulation is proposed that leads to local modification in the tangent stiffness matrix associated with tangled elements, and an additional piece-wise compatibility constraint. i-TFEM reduces to standard FEM for tangle-free meshes. The effectiveness and convergence characteristics of i-TFEM are demonstrated through a series of numerical experiments, involving both compressible and in-compressible problems.
Open AccessIsoparametric Tangled Finite Element Method for Nonlinear Elasticity
An important requirement in the standard finite element method (FEM) is that all elements in the underlying mesh must be tangle-free i.e., the Jacobian must be positive throughout each element. To relax this requirement, an isoparametric tangled finite element method (i-TFEM) was recently proposed for linear elasticity problems. It was demonstrated that i-TFEM leads to optimal convergence even for severely tangled meshes. In this paper, i-TFEM is generalized to nonlinear elasticity. Specifically, a variational formulation is proposed that leads to local modification in the tangent stiffness matrix associated with tangled elements, and an additional piece-wise compatibility constraint. i-TFEM reduces to standard FEM for tangle-free meshes. The effectiveness and convergence characteristics of i-TFEM are demonstrated through a series of numerical experiments, involving both compressible and in-compressible problems.