Calibration of parallel bond parameters in bonded particle models via physics-informed adaptive moment optimisation

Qu, Tongming; Feng, Y. T.; Wang, Min; Jiang, Shengqiang

doi:10.1016/j.powtec.2020.02.077

Cited by 49 publications

(13 citation statements)

References 66 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The procedure includes the several steps as following: 1) establish an optimization model, 2) optimize the optimized objective function through an optimization algorithm (such as adaptive moment estimation (Adam), Genetic Algorithm (GA), Ant-colony algorithm (ACA), Surface fitting method (SFM), Deeping learning network(DLN), etc., and 3) train and get the best model. The optimization-based method can be viewed as a "model of model" to calibrate the DEM parameters [67].…”

Section: Optimization-based Methodsmentioning

confidence: 99%

“…The research results show that after about 80 iterations, the calculated results tend to be stable. Feng et al [67] used the optimization method to carry out an automated calibration procedure for bonded SiC ceramic materials. The UCT and TPBT model were performed with Adam algorithm in a few iterative steps to match the bond parameters.…”

Section: Optimization-based Methodsmentioning

confidence: 99%

Section: Optimization-based Methodsmentioning

confidence: 99%

“…Compression test is a common mechanical property test method used to evaluate the shear failure of brittle materials [64,66,68,67,73]. Its schematic diagram and DEM simulation failure diagram are shown in Fig.…”

Section: Compression Testmentioning

confidence: 99%

See 3 more Smart Citations

Discrete element modeling of the machining processes of brittle materials: recent development and future prospective

Jiang

Tang

et al. 2020

Int J Adv Manuf Technol

Self Cite

View full text Add to dashboard Cite

In recent years, the discrete element method (DEM) has been used to model the bulk material, especially for the brittle materials (such as rocks, ceramics, concrete, ice, etc.) with various mechanical properties or responses by setting serials of contact properties (such as bonds) in the particle assembly. These bonds can withstand a certain amount of force and/or moment, so that the stresses executed in the bond can be used for determining the initiation and propagation of micro-crack. There are increasing evidences over the last twenty years that the DEM is becoming an effective numerical method to simulate the cracking, crushing and deformation of continuous media under external loads. The DEM has now been widely used in the field of processing and machining of rock, ceramic, concrete and other brittle materials. In this paper, the theoretical principles, formulations, contact models as well as the numerical solving processes of DEM are introduced. The applications of DEM for the machining processes of brittle and rigid materials such as ceramics are described and reviewed in detail, with future development trend also discussed.

show abstract

Section: Optimization-based Methodsmentioning

confidence: 99%

Section: Optimization-based Methodsmentioning

confidence: 99%

Section: Optimization-based Methodsmentioning

confidence: 99%

Section: Compression Testmentioning

confidence: 99%

See 2 more Smart Citations

Discrete element modeling of the machining processes of brittle materials: recent development and future prospective

Jiang

Tang

et al. 2020

Int J Adv Manuf Technol

Self Cite

View full text Add to dashboard Cite

show abstract

“…Chehreghani et al [2] performed calibration with a response surface methodology and a central composite design approach. The adaptive moment optimization technique has been applied by Qu et al [26] to calibrate a BPM. Also, strategies based on neural networks or genetic algorithms can be found in the literature [3,19,28].…”

Section: Introductionmentioning

confidence: 99%

Linearization-based methods for the calibration of bonded-particle models

Jarolin

Dosta

2020

Comp. Part. Mech.

View full text Add to dashboard Cite

In the work at hand, two methods for the calibration of the elastic material parameters of bonded-particle models (BPMs) are proposed. These methods are based on concepts of classical mechanics and enable a faster calibration compared to the conventional trial and error strategy. Moreover, they can be used to counter-check the consistency of the BPM. In the first method, the mathematical model is linearized and solved in a classical matrix-vector formulation similar to a finite element scheme. Further linearization in the second approach enables the direct calculation of the elastic stiffness tensor, reducing computational costs massively. To analyze the capabilities and limitations of both methods, they have been applied in three different case studies. Obtained results have shown that the new strategy allows us to significantly reduce the calculation time.

show abstract

A multi‐fidelity residual neural network based surrogate model for mechanical behaviour of structured sand

Zhou,

Yin,

et al. 2024

Num Anal Meth Geomechanics

View full text Add to dashboard Cite

The structured sand presents significant interparticle bonding and anisotropy, resulting in significant differences in the physical and mechanical properties from the pure sand. This study proposes a new surrogate model based on the concept of multi‐fidelity residual neural network (MR‐NN) as an alternative to DEM simulation for predicting the mechanical behaviours of structured sand with different initial anisotropy and saving largely computational costs. The model is initially trained using low‐fidelity (LF) data to focus on capturing the main underpinning correlations between macroscopic mechanical parameters with inter‐particle properties and anisotropic state variables (i.e., microscopic fabric and tilting angle of sand), where the LF data are generated from previously proposed anisotropic macro‐micro quantitative correlation. Subsequent training on sparser high‐fidelity (HF) data is used to calibrate and refine the model with HF data generated from DEM simulations for anisotropic structured sand. Feedforward neural network (FNN) is adopted as the baseline algorithm for training models. The macroscopic anisotropic parameters of structured sand predicted by the surrogate model are compared with DEM simulations to examine its feasibility and generalization ability. Furthermore, the robustness of the surrogate model is examined by discussing the effect of LF data on the performance of MR‐NN. The superiority of the MR‐NN is further verified by comparing the performance of the trained MR‐NN with the one‐shot trained FNN based on the same HF data. All results demonstrate that the proposed surrogate model can provide a fast and accurate simulation of the anisotropic parameters of structured sand.

show abstract

Calibration of parallel bond parameters in bonded particle models via physics-informed adaptive moment optimisation

Cited by 49 publications

References 66 publications

Discrete element modeling of the machining processes of brittle materials: recent development and future prospective

Discrete element modeling of the machining processes of brittle materials: recent development and future prospective

Linearization-based methods for the calibration of bonded-particle models

A multi‐fidelity residual neural network based surrogate model for mechanical behaviour of structured sand

Contact Info

Product

Resources

About