A geometry-based algorithm for cloning real grains

Jerves, Alex X.; Kawamoto, Reid; Andrade, José E.

doi:10.1007/s10035-017-0716-7

Cited by 19 publications

(20 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Segundo, una mejorada, ahora paralelizada y anada versión del algoritmo estocástico geométrico de clonación [13], el cual está basado en las distribuciones multivariables antes mencionadas, y que, de la misma manera, introduce un proceso de muestro mejorado de radios, así como también una prueba de control de calidad nuevo basado en la relación volumen-radio es discutido. Tercero, el ADN morfológico de los granos (es decir, relación de aspecto, redondez, relación volumen-supercie y diámetro del grano) es también extraído de los nuevos granos clonados y comparado al obtenido de la muestra padre.…”

Section: Introductionunclassified

“…Thus, to ameliorate the high cost, time consuming and size constraints of 3DXRCT scanning, a first computational algorithm was developed to obtain morphological features from a sample of real grains, (i.e., aspect ratio, principal geometric directions, roundness and spherical radius) [13]. This first algorithm successfully reached its goal of generating new grains (clones) similar to ones from a parent sample.…”

Section: Introductionmentioning

confidence: 99%

“…Finally, the algorithm was optimized with the aim of optimizing its run time and parallelize it [29][30][31][32]. Furthermore, the algorithm's parameters were calibrated for better results [13].…”

Section: Introductionmentioning

confidence: 99%

“…This work was carried on has been performed in the stages, as shown by Figure 1. First, the morphological DNA from the sample grains (parents) is extracted [13]. Second, a clone's embryo (equivalent ellipsoid) is generated by sampling values from the multivariable aspect ratio and minimum principal direction's distribution.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A geometry-based algorithm for cloning real grains 2.0

2018

Self Cite

View full text Add to dashboard Cite

We introduce an improved version of a computational algorithm that "clones"/generates an arbitrary number of new digital grains from a real sample of real digitalized granular material. Our improved algorithm produces "cloned" grains that more accurately approach the morphological features displayed by their parents. Now, the "cloned" grains were also included in a Discrete Element Method simulation of a tri-axial test and showed similar mechanical behavior compared to the displayed by the original (parent) sample. Thus, the present work is divided in four parts. First, we compute multivariable probability density functions (PDF) from the parents' morphological parameters (morphological DNA), i.e., aspect ratio, roundness, volume-surface ratio, and particle diameter. Second, an improved, now parallelized and better tuned version of the Geometric Stochastic Cloning (GSC) algorithm [13], which is based on the aforementioned multivariable distributions, and that, in the same way, introduces an enhanced radii sampling process, as well as a new quality control test based on the volume-surface ratio is discussed. Third, morphological DNA of the grains (i.e., aspect ratio, roundness, volume-surface ratio and particle diameter) is also extracted from the new "cloned" grains and compared to the one obtained from the parent sample. Fourth, clones and parents are subjected to a tri-axial compression tests using a Level Set (LS) in Discrete Element scheme (3DLS-DEM), and then, compared in terms of their mechanical response. Finally, the error of the "clones" in the morphology and mechanical behavior is analyzed and discussed for future improvements.

show abstract

Section: Introductionunclassified

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A geometry-based algorithm for cloning real grains 2.0

2018

Self Cite

View full text Add to dashboard Cite

show abstract

“…This paper introduces a micromechanical model to study the liquefaction phenomenon. It is based on the commonly used discrete element method [4] (DEM) to micromechanically model soils and rocks [5,6]. DEM has been used before to model soil shearing during undrained conditions [1,7].…”

Section: Introductionmentioning

confidence: 99%

Micromechanics of Liquefaction in Granular Materials

2018

View full text Add to dashboard Cite

A discrete element method (DEM) implementation is developed to study the micromechanics of liquefaction in granular materials. In a liquefaction event, the pore water acts as a cushion between the grains, reducing the contact and friction forces and the overall soil strength. The proposed model reproduces this phenomenon by introducing the effect of pore water as a constraint over the DEM particles' mechanics. The DEM particles will suffer resistance to any displacement changing the pore volume, which takes into account the very small compressibility of water. It is found that this constraint is enough to simulate soil liquefaction under quasistatic deformation. Finally, it is shown that the initial density of the granular skeleton, defined by the number of contacts between grains, plays a critical role in determining if the soil will liquefy or not. This critical value opens the possibility of treating liquefaction in soils as a bifurcation problem.

show abstract

Synthesizing realistic sand assemblies with denoising diffusion in latent space

Vlassis,

Sun,

Alshibli

et al. 2024

Num Anal Meth Geomechanics

View full text Add to dashboard Cite

The shapes and morphological features of grains in sand assemblies have far‐reaching implications in many engineering applications, such as geotechnical engineering, computer animations, petroleum engineering, and concentrated solar power. Yet, our understanding of the influence of grain geometries on macroscopic response is often only qualitative, due to the limited availability of high‐quality 3D grain geometry data. In this paper, we introduce a denoising diffusion algorithm that uses a set of point clouds collected from the surface of individual sand grains to generate grains in the latent space. By employing a point cloud autoencoder, the three‐dimensional point cloud structures of sand grains are first encoded into a lower‐dimensional latent space. A generative denoising diffusion probabilistic model is trained to produce synthetic sand that maximizes the log‐likelihood of the generated samples belonging to the original data distribution measured by a Kullback‐Leibler divergence. Numerical experiments suggest that the proposed method is capable of generating realistic grains with morphology, shapes and sizes consistent with the training data inferred from an F50 sand database. We then use a rigid contact dynamic simulator to pour the synthetic sand in a confined volume to form granular assemblies in a static equilibrium state with targeted distribution properties. To ensure third‐party validation, 50,000 synthetic sand grains and the 1542 real synchrotron microcomputed tomography (SMT) scans of the F50 sand, as well as the granular assemblies composed of synthetic sand grains are made available in an open‐source repository.

show abstract

A geometry-based algorithm for cloning real grains

Cited by 19 publications

References 25 publications

A geometry-based algorithm for cloning real grains 2.0

A geometry-based algorithm for cloning real grains 2.0

Micromechanics of Liquefaction in Granular Materials

Synthesizing realistic sand assemblies with denoising diffusion in latent space

Contact Info

Product

Resources

About