An iterative Bayesian filtering framework for fast and automated calibration of DEM models

Cheng, Hongyang; Shuku, Takayuki; Thoeni, Klaus; Tempone, Pamela; Luding, Stefan; Magnanimo, Vanessa

doi:10.1016/j.cma.2019.01.027

Cited by 86 publications

(46 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this paper, a DEM representative volume previous calibrated against oedometric experiments as obtained from 3DXRCT images Cheng et al (2018a)is probed with small perturbations. The numerical probes are performed with a wide range of magnitudes, along both forward and backward loading directions.…”

Section: Discussionmentioning

confidence: 99%

“…Because the calibration procedure and a posterior probability distribution of micromechanical parameters is already available from the Bayesian calibration Cheng et al (2018a), it would be interesting to see how the uncertainties at the microscale affect the dispersion relations at various wavelengths. Future work also involves using existing continuum models to better approximate the dispersion branches obtained from the DEM simulations, so as to accurately capture the change of dispersion relations with respect to stress history and anisotropy in the materials.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Elastic wave propagation in dry granular media: Effects of probing characteristics and stress history

Cheng

Luding

Saitoh

et al. 2020

International Journal of Solids and Structures

Self Cite

View full text Add to dashboard Cite

Elastic wave propagation provides a noninvasive way to probe granular materials. The discrete element method using particle configuration as input, allows a micromechanical interpretation on the acoustic response of a given granular system. This paper compares static and dynamic numerical probing methods, from which wave velocities are either deduced from elastic moduli or extracted from the time/frequency-domain signals. The dependence of wave velocities on key characteristics, i.e., perturbation magnitude and direction for static probing, and maximum travel distance and inserted signals for dynamic probing, is investigated. It is found that processing the frequency-domain signals obtained from dynamic probing leads to reproducible wave velocities at all wavenumbers, irrespective of the perturbation characteristics, whereas the maximum travel distance and input signals for the time domain analysis have to be carefully chosen, so as to obtain the same long-wavelength limits as from the frequency domain. Static and dynamic probes are applied to calibrated representative volumes of glass beads, subjected to cyclic oedometric compression.Although the perturbation magnitudes are selected to reveal only the elastic moduli, the deduced wave velocities are consistently lower than the long-wavelength limits at various stress states, and thus sensitive to sample size. While the static probes investigate the influence of stress history on modulus degradation, dynamic probing offers insights about how dispersion relations evolve during cyclic compression. Interestingly, immediately after each load reversal the incremental behavior is reversibly elastoplastic, until it becomes truly elastic with further unload/reload. With repeating unload/reload, the P-or S-wave dispersion relations become increasingly scalable with respect to their long-wavelength limits.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Elastic wave propagation in dry granular media: Effects of probing characteristics and stress history

Cheng

Luding

Saitoh

et al. 2020

International Journal of Solids and Structures

Self Cite

View full text Add to dashboard Cite

show abstract

“…Exciting pulse Iterative prior resampling (e.g. Cheng et al, 2019) is relatively simple and may contribute to better estimate the uncertainty. The algorithm (Fig.…”

Section: Snmrmentioning

confidence: 99%

Improving Bayesian Evidential Learning 1D imaging (BEL1D) accuracy through iterative prior resampling

Michel

Nguyen

Hermans

2020

Preprint

View full text Add to dashboard Cite

“…Statistical or optimisation‐based algorithms become popular to determine particle‐scale parameters in DEM as an inverse problem. Existing algorithms include the response surface methodology, artificial neural networks, Latin hypercube sampling and Kriging, random forest, the genetic algorithm, the sequential quasi‐Monte Carlo, and the Bayesian approach . Although these algorithms are useful to quantify a wide range of complex problems, their applications alone still suffer from numerical issues such as local optimum and time‐consuming iterative process.…”

Section: Introductionmentioning

confidence: 99%

A hybrid calibration approach to Hertz‐type contact parameters for discrete element models

Feng

Zhao

et al. 2020

Num Anal Meth Geomechanics

View full text Add to dashboard Cite

This study aims at providing a hybrid calibration framework to estimate Hertztype contact parameters (particle-scale shear modulus and Poisson ratio) for both two-dimensional and three-dimensional discrete element modelling (DEM). On the basis of statistically isotropic granular packings, a set of analytical formulae between macroscopic material parameters (Young modulus and Poisson ratio) and particle-scale Hertz-type contact parameters for granular systems are derived under small-strain isotropic stress conditions. However, the derived analytical solutions are only estimated values for general models.By viewing each DEM modelling as an implicit mathematical function taking the particle-level parameters as independent variables and employing the derived analytical solutions as the initial input parameters, an automatic iterative scheme is proposed to obtain the calibrated parameters with higher accuracies. Considering highly nonlinear features and discontinuities of the macro-micro relationship in Hertz-based discrete element models, the adaptive moment estimation algorithm is adopted in this study because of its capacity of dealing with noise gradients of cost functions. The proposed method is validated with several numerical cases including randomly distributed monodisperse and polydisperse packings. Noticeable improvements in terms of calibration efficiency and accuracy have been made. K E Y W O R D Sadaptive moment estimation, calibration, constitutive law, discrete element method, Hertz-Mindilin contact model, nonlinear elastic model

show abstract

An iterative Bayesian filtering framework for fast and automated calibration of DEM models

Cited by 86 publications

References 60 publications

Elastic wave propagation in dry granular media: Effects of probing characteristics and stress history

Elastic wave propagation in dry granular media: Effects of probing characteristics and stress history

Improving Bayesian Evidential Learning 1D imaging (BEL1D) accuracy through iterative prior resampling

A hybrid calibration approach to Hertz‐type contact parameters for discrete element models

Contact Info

Product

Resources

About