Robert Buarque de Macedo
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
Robert Buarque de Macedo has a total of 30 co-authors. Here are the top co-authors by number of shared publications:
- Craig Hamel
Sandia National Laboratories (SNL-NM), Albuquerque, NM (United States)
5 shared publications Albuquerque Sandia National Laboratories (Snl-Nm)
4 shared publicationsKyle Johnson
4 shared publicationsJan Fuhg
Cornell University
3 shared publications- Reese Edward Jones
Sandia National Laboratories
3 shared publications Usdoe National Nuclear Security Administration (Nnsa)
3 shared publications
Latest Publications
- Open Access
- 13 April 2026
- Published by Elsevier in SSRN Electronic Journal
3D One-shot Geometry Agnostic Prediction of Non-linear Porous Microstructure Mechanics using Graph Neural Networks
Accurate prediction of material failure is critical to automotive, civil, aerospace and other industries that handle metals under high stress and/or repeated loads. Predicting where and when failure occurs is heavily dependent on the microstructure of the material such as the geometry and location of pores. Nevertheless, adding such microstructural information to finite element method (FEM) simulations requires a level of mesh fidelity that can render practical use of models infeasible, especially when considering non-linear and history-dependent materials. An alternative approach is surrogate modeling with deep learning, where one trains a neural network to quickly predict simulation behavior. In this work, we investigate graph neural networks (GNNs) as surrogate models for the mechanical evolution of 3D non-linear porous microstructures undergoing damage. Unlike convolutional neural networks, a GNN is ”mesh agnostic”: it can input and output an FEM mesh rather than being constrained to predictions on fixed sized voxelated representations of models. We illustrate how GNNs are capable of predicting the evolution of plastic strain in material microstructures with varying porosity distributions, and how results generalize to domains outside the training distribution for metals undergoing damage evolution. Further, we investigate how our techniques can be used to predict damage evolution in viscoplastic materials undergoing cyclic failure. Our work demonstrates how deep learning can be used effectively to predict mechanics in material microstructures without requiring expensive retraining when the simulation geometry is modified, a critical step towards the use of surrogate models in real-world 3D applications.
- 13 April 2026
- Published by Elsevier in SSRN Electronic Journal
Open Access3D One-shot Geometry Agnostic Prediction of Non-linear Porous Microstructure Mechanics using Graph Neural Networks
Accurate prediction of material failure is critical to automotive, civil, aerospace and other industries that handle metals under high stress and/or repeated loads. Predicting where and when failure occurs is heavily dependent on the microstructure of the material such as the geometry and location of pores. Nevertheless, adding such microstructural information to finite element method (FEM) simulations requires a level of mesh fidelity that can render practical use of models infeasible, especially when considering non-linear and history-dependent materials. An alternative approach is surrogate modeling with deep learning, where one trains a neural network to quickly predict simulation behavior. In this work, we investigate graph neural networks (GNNs) as surrogate models for the mechanical evolution of 3D non-linear porous microstructures undergoing damage. Unlike convolutional neural networks, a GNN is ”mesh agnostic”: it can input and output an FEM mesh rather than being constrained to predictions on fixed sized voxelated representations of models. We illustrate how GNNs are capable of predicting the evolution of plastic strain in material microstructures with varying porosity distributions, and how results generalize to domains outside the training distribution for metals undergoing damage evolution. Further, we investigate how our techniques can be used to predict damage evolution in viscoplastic materials undergoing cyclic failure. Our work demonstrates how deep learning can be used effectively to predict mechanics in material microstructures without requiring expensive retraining when the simulation geometry is modified, a critical step towards the use of surrogate models in real-world 3D applications.
- Open Access
- 1 March 2025
- Published by Office of Scientific and Technical Information (OSTI)
Reliability and Lifetime Predictions of Fine Pitch Ball Grid Arrays in Thermal Cycling
- 1 March 2025
- Published by Office of Scientific and Technical Information (OSTI)
Open AccessReliability and Lifetime Predictions of Fine Pitch Ball Grid Arrays in Thermal Cycling
- 5 July 2024
- Published by Elsevier in International Journal of Solids and Structures
- Vol. 301
Multiscale simulation of spatially correlated microstructure via a latent space representation
- 5 July 2024
- Published by Elsevier in International Journal of Solids and Structures
- Vol. 301
Multiscale simulation of spatially correlated microstructure via a latent space representation
- Open Access
- 1 May 2024
- Published by Office of Scientific and Technical Information (OSTI)
Modeling shock attenuation in granular materials: the role of particle shapes
Proof-of-concept computational study of shock in granular materials of various shapes using LAMMPS
- 1 May 2024
- Published by Office of Scientific and Technical Information (OSTI)
Open AccessModeling shock attenuation in granular materials: the role of particle shapes
Proof-of-concept computational study of shock in granular materials of various shapes using LAMMPS
- 23 August 2021
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 118 (36)
Unearthing real-time 3D ant tunneling mechanics
Granular excavation is the removal of solid, discrete particles from a structure composed of these objects. Efficiently predicting the stability of an excavation during particle removal is an unsolved and highly nonlinear problem, as the movement of each grain is coupled to its neighbors. Despite this, insects such as ants have evolved to be astonishingly proficient excavators, successfully removing grains such that their tunnels are stable. Currently, it is unclear how ants use their limited information about the environment to construct lasting tunnels. We attempt to unearth the ants’ tunneling algorithm by taking three-dimensional (3D) X-ray computed tomographic imaging (XRCT), in real time, of Pogonomyrmex ant tunnel construction. By capturing the location and shape of each grain in the domain, we characterize the relationship between particle properties and ant decision-making within an accurate, virtual recreation of the experiment. We discover that intergranular forces decrease significantly around ant tunnels due to arches forming within the soil. Due to this force relaxation, any grain the ants pick from the tunnel surface will likely be under low stress. Thus, ants avoid removing grains compressed under high forces without needing to be aware of the force network in the surrounding material. Even more, such arches shield tunnels from high forces, providing tunnel robustness. Finally, we observe that ants tend to dig piecewise linearly downward. These results are a step toward understanding granular tunnel stability in heterogeneous 3D systems. We expect that such findings may be leveraged for robotic excavation. Significance Predicting the stability of granular materials under particle removal has wide-reaching applications, including automating tunnel excavations. Searching for general laws that govern granular stability is challenging given the complexity of granular material dynamics. However, knowledge may be gained from the natural world, where organisms have evolved adaptive tunneling strategies. Among these, subterranean-nesting ants execute an innate tunneling behavioral program that can lead to remarkably stable tunnel excavation. We use X-rays to image the process of ant tunnel construction through a particulate substrate. We use these data to create a grain-scale accurate simulation for estimating particle mechanics in the sample during real-time excavation. We present evidence that ants benefit from force redistributions during incremental digging, suggesting techniques for robotic mining.
- 23 August 2021
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 118 (36)
Unearthing real-time 3D ant tunneling mechanics
Granular excavation is the removal of solid, discrete particles from a structure composed of these objects. Efficiently predicting the stability of an excavation during particle removal is an unsolved and highly nonlinear problem, as the movement of each grain is coupled to its neighbors. Despite this, insects such as ants have evolved to be astonishingly proficient excavators, successfully removing grains such that their tunnels are stable. Currently, it is unclear how ants use their limited information about the environment to construct lasting tunnels. We attempt to unearth the ants’ tunneling algorithm by taking three-dimensional (3D) X-ray computed tomographic imaging (XRCT), in real time, of Pogonomyrmex ant tunnel construction. By capturing the location and shape of each grain in the domain, we characterize the relationship between particle properties and ant decision-making within an accurate, virtual recreation of the experiment. We discover that intergranular forces decrease significantly around ant tunnels due to arches forming within the soil. Due to this force relaxation, any grain the ants pick from the tunnel surface will likely be under low stress. Thus, ants avoid removing grains compressed under high forces without needing to be aware of the force network in the surrounding material. Even more, such arches shield tunnels from high forces, providing tunnel robustness. Finally, we observe that ants tend to dig piecewise linearly downward. These results are a step toward understanding granular tunnel stability in heterogeneous 3D systems. We expect that such findings may be leveraged for robotic excavation. Significance Predicting the stability of granular materials under particle removal has wide-reaching applications, including automating tunnel excavations. Searching for general laws that govern granular stability is challenging given the complexity of granular material dynamics. However, knowledge may be gained from the natural world, where organisms have evolved adaptive tunneling strategies. Among these, subterranean-nesting ants execute an innate tunneling behavioral program that can lead to remarkably stable tunnel excavation. We use X-rays to image the process of ant tunnel construction through a particulate substrate. We use these data to create a grain-scale accurate simulation for estimating particle mechanics in the sample during real-time excavation. We present evidence that ants benefit from force redistributions during incremental digging, suggesting techniques for robotic mining.