| Line 1: | Line 1: | ||
| − | Published in ''Computer Methods in Applied Mechanics and Engineering'', Vol. 420, Art. Num.116705, 2024<br> DOI: 10.1016/j.cma.2023.116705 | + | Published in ''Computer Methods in Applied Mechanics and Engineering'', Vol. 420, Art. Num.116705, 2024<br> |
| + | DOI: [https://www.sciencedirect.com/science/article/abs/pii/S0045782523008289 10.1016/j.cma.2023.116705] | ||
==Abstract== | ==Abstract== | ||
Reconstruction methods for granular materials are an essential step in achieving an accurate geometrical basis in the use of particle methods such as the well-known discrete element method. In this study, a novel continuous analytical method will be proposed to achieve individual 3D reconstructions of granular materials. For this purpose, the 2D Fourier Descriptor theory is used and a generalization of it to the 3D case is obtained, which is quite different from the spherical harmonics theory. Some limitations existing in the original model will be addressed by developing and calibrating a generalized approach that results in a total of four main models. Particles from a study sample are reconstructed, and the models are evaluated based on four generalized error metrics, as well as the presence of some specific limitations detected in the models. Finally, the models are compared using a decision-making method in different decisional scenarios. | Reconstruction methods for granular materials are an essential step in achieving an accurate geometrical basis in the use of particle methods such as the well-known discrete element method. In this study, a novel continuous analytical method will be proposed to achieve individual 3D reconstructions of granular materials. For this purpose, the 2D Fourier Descriptor theory is used and a generalization of it to the 3D case is obtained, which is quite different from the spherical harmonics theory. Some limitations existing in the original model will be addressed by developing and calibrating a generalized approach that results in a total of four main models. Particles from a study sample are reconstructed, and the models are evaluated based on four generalized error metrics, as well as the presence of some specific limitations detected in the models. Finally, the models are compared using a decision-making method in different decisional scenarios. | ||
Revision as of 11:55, 27 September 2024
Published in Computer Methods in Applied Mechanics and Engineering, Vol. 420, Art. Num.116705, 2024
DOI: 10.1016/j.cma.2023.116705
Abstract
Reconstruction methods for granular materials are an essential step in achieving an accurate geometrical basis in the use of particle methods such as the well-known discrete element method. In this study, a novel continuous analytical method will be proposed to achieve individual 3D reconstructions of granular materials. For this purpose, the 2D Fourier Descriptor theory is used and a generalization of it to the 3D case is obtained, which is quite different from the spherical harmonics theory. Some limitations existing in the original model will be addressed by developing and calibrating a generalized approach that results in a total of four main models. Particles from a study sample are reconstructed, and the models are evaluated based on four generalized error metrics, as well as the presence of some specific limitations detected in the models. Finally, the models are compared using a decision-making method in different decisional scenarios.
Document information
Published on 01/01/2024
DOI: 10.1016/j.cma.2023.116705
Licence: CC BY-NC-SA license
Share this document
claim authorship
Are you one of the authors of this document?