Plastic Deformation Generative Modeling: A Permanent-Deformation-Driven Framework for 3D Shape Generation
Abstract
Most three-dimensional (3D) generative models construct objects by sampling latent representations, denoising geometric structures, predicting implicit surfaces, or generating collections of spatial primitives. Although these approaches can produce complex geometry, the generation process does not necessarily correspond to a physically meaningful material-forming process. This paper proposes Plastic Deformation Generative Modeling (PDGM), a framework in which 3D geometry is generated primarily through the controlled accumulation of irreversible material deformation. Inspired by forging, stamping, extrusion, drawing, and compression, PDGM represents an object through its geometry, material parameters, deformation state, and plastic history. A sequence of parameterized forming operations is applied to an initial material configuration, and the resulting geometry is updated using an elastoplastic constitutive model or a learned approximation of its response. Unlike purely elastic deformation, the proposed framework preserves permanent changes after the external load is removed. The formulation combines a generative operation policy, a plastic-state transition model, geometric validity constraints, and a residual-shape evaluation objective. It can be implemented using finite element methods, material point methods, differentiable simulation, or learned deformation surrogates. This paper formalizes the framework, presents its mathematical components, proposes a training strategy, and specifies an experimental protocol for comparison with conventional 3D generation and deformation methods. Since no experiments are conducted in this manuscript, the proposed advantages are presented as testable hypotheses rather than established empirical findings. PDGM offers a research direction for generating 3D objects through material transformation instead of relying exclusively on direct geometric synthesis.
Related
- Fold-and-Deploy Generative Modeling: A Material-Centric Framework for Three-Dimensional Shape Generation Through Folding, Bending, and Deployment
- Subtractive Trajectory Generative Modeling: A Material-Removal-Centric Framework for Three-Dimensional Shape Generation
- Layerwise Deposition-Field Generative Modeling: A Process-Conditioned Framework for Three-Dimensional Generation Through Thin-Film Deposition and Surface Growth
- Deposition-Based 3D Generative Modeling: A Framework for Three-Dimensional Object Generation Through Gas-Phase Particle Deposition, Local Adsorption, and Branching Growth
- Granular Flow and Deposition Generative Modeling: A Physics-Driven Framework for Three-Dimensional Shape Generation Through Discrete Particle Dynamic
BibTeX
@misc{zhang2026plastic,
title = {{Plastic Deformation Generative Modeling: A Permanent-Deformation-Driven Framework for 3D Shape Generation}},
author = {Jincheng Zhang},
howpublished = {Zenodo},
year = {2026},
month = oct,
doi = {10.5281/zenodo.23636320},
url = {https://zenodo.org/records/23636320},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Plastic Deformation Generative Modeling: A Permanent-Deformation-Driven Framework for 3D Shape Generation
AU - Jincheng Zhang
PY - 2026
DA - 2026/10/25
PB - Zenodo
DO - 10.5281/zenodo.23636320
UR - https://zenodo.org/records/23636320
AB - Most three-dimensional (3D) generative models construct objects by sampling latent representations, denoising geometric structures, predicting implicit surfaces, or generating collections of spatial primitives. Although these approaches can produce complex geometry, the generation process does not necessarily correspond to a physically meaningful material-forming process. This paper proposes Plastic Deformation Generative Modeling (PDGM), a framework in which 3D geometry is generated primarily through the controlled accumulation of irreversible material deformation. Inspired by forging, stamping, extrusion, drawing, and compression, PDGM represents an object through its geometry, material parameters, deformation state, and plastic history. A sequence of parameterized forming operations is applied to an initial material configuration, and the resulting geometry is updated using an elastoplastic constitutive model or a learned approximation of its response. Unlike purely elastic deformation, the proposed framework preserves permanent changes after the external load is removed. The formulation combines a generative operation policy, a plastic-state transition model, geometric validity constraints, and a residual-shape evaluation objective. It can be implemented using finite element methods, material point methods, differentiable simulation, or learned deformation surrogates. This paper formalizes the framework, presents its mathematical components, proposes a training strategy, and specifies an experimental protocol for comparison with conventional 3D generation and deformation methods. Since no experiments are conducted in this manuscript, the proposed advantages are presented as testable hypotheses rather than established empirical findings. PDGM offers a research direction for generating 3D objects through material transformation instead of relying exclusively on direct geometric synthesis.
ER -