betaFeedback

WorksPhysics and Astronomy

Discrete Element Simulations of particles interacting via capillary forces using MercuryDPM

Abstract

We present the implementation of two advanced capillary bridge approximations within the Discrete Element Method (DEM) framework of the open-source code MercuryDPM. While MercuryDPM already includes a simplified version of the Willett approximation, our work involves implementing both the classical Willett approximation and the recently published Bagheri approximation in MercuryDPM. Through detailed descriptions and illustrative simulations using a two-particle collision model, we demonstrate the enhanced accuracy and capabilities of these approximations in capturing the complex dynamics of wet granular matter.

Metadata

Motivation and significance

In the study of particulate systems, understanding the interactions between particles in the presence of liquid phases is crucial for a variety of industrial and scientific applications [1, 2]. One of the key phenomena in such systems is the formation of liquid bridges between particles, which causes forces due to capillarity [3]. These interactions can significantly influence the mechanical properties and behavior of granular materials [5, 6] [4]. Liquid capillary bridges play a vital role in many processes, in the fields of soil mechanics and powder technology [7, 8]. Accurately approximating these forces is therefore essential for reliable simulations of wet granular materials. Incorporating these complex interactions into DEM simulations poses significant challenges. The naïve way to account for capillary forces between particles linked with a liquid bridge (see Figure 1) would require solving the Young-Laplace equation for the capillary force in each time step of the DEM integration, which is certainly not feasible. Therefore, fit formulae were elaborated which approximate the capillary force as a function of the particle and liquid properties. The evaluation of these fit formulae is by orders of magnitude more efficient than the numerical solution of the Young-Laplace equation. The fit formulae by Willett et al. [9] and Bagheri et al. [10] have been shown to be highly accurate, where the latter delivers also the surface area of a liquid bridge which is needed when considering evaporation processes in wet granulate. By now, DEM packages rely on more simple and less accurate formulae.

Capillary bridge spanning between two unequal-sized particles

Figure 1. Capillary bridge spanning between two unequal-sized particles.

C1Current code versionMercuryDPM Master Version
C2Permanent link to code/repository used for this code versionhttps://github.com/meysam-bagheri/MercuryDPM
C3Permanent link to Reproducible Capsule
C4Legal Code LicenseBSD 3-clause
C5Code versioning system usedgit
C6Software code languages, tools, and services usedC++
C7Compilation requirements, operating environments & dependenciesLinux/Windows/macOS; see https://www.mercurydpm.org/
C8If available Link to developer documentation/manualhttps://github.com/meysam-bagheri/MercuryDPM/blob/master/README.md https://www.mercurydpm.org/documentation
C9Support email for questions[email protected]; [email protected]

Table 2.

The current article reports the implementation of highly accurate descriptions of capillary forces in the DEM package MercuryDPM. MercuryDPM is a versatile open-source code designed for simulating the dynamics of particulate materials using DEM [11, 12]. Its flexibility and extensibility make it an ideal tool for simulations of large scale granular systems By enhancing MercuryDPM with accurate liquid bridge approximations, researchers and engineers can gain deeper insights into the behavior of wet granular systems and improve the design and optimization of related processes.

The Bagheri approximation, originally formulated for equal-sized spheres [10], is extended in this work to account for particles of unequal sizes. This application note outlines the implementation of both the Bagheri approximation [10] and the Willett approximation [9] into MercuryDPM for modeling liquid capillary bridges. Additionally, we compare the results with the simplified Willett approximation already integrated into MercuryDPM. This new implementation significantly improves MercuryDPM’s ability to predict the behavior of wet granular materials.

The following sections of this note describe the particle simulation system and detail the contact models. Subsequently, we demonstrate the implementation of the codes. Finally, we show illustrative examples and discuss the performance and applicability of the approximations in various scenarios. A detailed description of the different capillary force approximations is provided in Appendix A.

Software description

Software architecture

Our simulations use a two-particle collision model to analyze interactions between particles of unequal radii, as shown in Figure 2. The contact model combines dissipative Hertz contact forces [13, 14] with a non-linear liquid bridge model for capillary forces. The results from the implementation of different liquid bridge approximations—namely, the simplified Willett approximation, the classical Willett approximation, and the Bagheri approximation—are compared. The parameters used in DEM simulations are given in Table 1 (PDF p. 4). Two different particle size ratios are considered in the two-particle simulations: 0.5 mm and 0.8 mm, and 0.5 mm and 1 mm, respectively. When particles approach each other at a certain distance SS, they are not linked by a liquid bridge and there is no capillary force. Once particles are in contact, a liquid bridge forms. After contact, the capillary force, F(S)F(S), is present at the same distance SS between the particles until the bridge ruptures.

Figure 2. Numerical setup for two-particle collision model for particles of different sizes with bridges (a) before collision (b) during collision and (c) after collision shown for particle sizes 0.50 mm and 0.80 mm.

variableunitvalue
elastic modulus (EE)MPa5
Poisson’s ratio (ν\nu)-0.35
particle density (ρ\rho)kg/m3^32000
contact angle (θ\theta)degree0
surface tension (γ\gamma)N/m0.079
liquid bridge volume (VV)nl10, 80

Table 1 (PDF p. 4). DEM simulation parameters.

Software functionalities

The simplified Willett approximation is available in MercuryDPM under the name LiquidBridgeWillet. The new adhesive contact models are implemented in MercuryDPM under the names LiquidBridgeBagheri and LiquidBridgeClassicalWillet, respectively. We compare the force versus interparticle distance results from three liquid bridge approximations for different particle size ratios and different liquid bridge volumes. We discuss the accuracy of each approximation, emphasizing the extended applicability of the Bagheri approximation for simulating polydisperse systems, even though it was originally developed for particles of equal radius.

Sample code snippets analysis

In Appendix A, we demonstrate the computation of the capillary forces using the Classical Willett approximation in LiquidBridgeClassicalWillet and the Bagheri approximation in LiquidBridgeBagheri, respectively.

Illustrative examples

As illustrated in Figure 3(a) and (b), the results from the simulation of the two-particle collision model for particle sizes of 0.5 mm and 0.8 mm, with liquid bridge volumes of 10 nl and 80 nl respectively, reveals an overlap between the forces predicted by the classical Willett approximation and the Bagheri approximation as a function of inter-particle distance. In contrast, the simplified Willett approximation shows a deviation in the capillary force between the two particles for both the liquid bridge volumes. This trend is also observed for particles sized 0.5 mm and 1 mm for different liquid bridge volume cases, as shown in Figure 4(a) and (b). However, the force magnitude increases as the effective interacting particle radius increases in Figure 4(a) and (b) compared to Figure 3(a) and (b). Thus, the Bagheri approximation is more accurate and closely aligns with the classical Willett approximation, regardless of the liquid bridge volume between particles and particle sizes.

Capillary force plots for two particle sizes and liquid bridge volumes

Figure 3. Capillary force, FF, exerted by a liquid bridge as a function of particle distance SS for solutions obtained from the fit equations of simplified Willett approximation, Classical Willett approximation and Bagheri approximation for liquid bridge volume (a) V=10V = 10 nl and (b) V=80V = 80 nl for particle sizes 0.5 mm and 0.8 mm.

Capillary force, $F$, exerted by a liquid bridge as a function of particle distance $S$

Figure 4. Capillary force, FF, exerted by a liquid bridge as a function of particle distance SS for solutions obtained from the fit equations of simplified Willett approximation, classical Willett approximation and Bagheri approximation for liquid bridge volume (a) V=10V = 10 nl and (b) V=80V = 80 nl for particle sizes 0.5 mm and 1 mm.

Impact

All contact models in MercuryDPM are defined by a normal, frictional, and adhesive contact models. The normal, frictional and adhesive contact models currently available in MercuryDPM are summarised in [11]. A specific contact model can be chosen by assigning a corresponding species to a particle. The name of the species can be obtained by concatenating the names of the normal, frictional, and adhesive contact model and adding the word Species. For example, users can utilize the interactions provided by particles of type HertzianViscoelasticMindlinLiquidBridgeBagheriSpecies, which include a dissipative Hertz normal force, Mindlin friction, and liquid-bridge adhesion forces based on the Bagheri approximations. Similarly, the adhesive force approximation can be replaced by the classical Willett approximation using interactions of type HertzianViscoelasticMindlinLiquidBridgeClassicalWilletSpecies. The implementations are included in the master branch of MercuryDPM. This integration will expand access to these advanced approximations, providing the entire MercuryDPM user community with tools to perform more detailed and accurate simulations.

A detailed description of the test simulation is provided in the README.md file on GitHub. Following these instructions leads to the force calculation shown in Figure 3(a), using the Bagheri model.

Conclusions

We have successfully implemented two accurate capillary bridge approximations—the Bagheri approximation and the classical Willett approximation—within the DEM-based open-source code MercuryDPM. The applicability of Bagheri approximation is extended to particles of unequal sizes. Through illustrative simulations, we have demonstrated the behavior of these models across various particle size ratios and liquid bridge volumes, and have evaluated their performance and applicability relative to a simplified version of the Willett approximation. We believe that this framework will prove valuable to users of MercuryDPM, enabling more precise calculations of capillary forces in particle systems.

Declaration of competing interest

There are no conflicts to declare.

Acknowledgement

We acknowledge Holger Götz for his assistance in integrating the code into the master branch of MercuryDPM.

Data availability

The implementation of the two models presented here and the test scripts are available at https://github.com/meysam-bagheri/MercuryDPM.

Capillary force approximations

The parameters of the capillary force approximations FF, are the contact angle θ\theta, liquid bridge volume VV, and the separation distance SS. The effective particle radius between the interacting particles RR, defined as

2R=1r1+1r2(1)\frac{2}{R} = \frac{1}{r_1} + \frac{1}{r_2} \tag*{(1)}

where r1r_1 and r2r_2 are the radius of the two particles 1 and 2, respectively. Using the effective radius, we obtain the dimensionless liquid bridge volume V∗=V/R3V^* = V/R^3. The liquid bridge rupture distance, ScS_c is given by Sc=(1+θ/2)(V∗1/3+V∗2/3/10)S_c = (1+\theta/2)(V^{*1/3} + V^{*2/3}/10), and the scaled separation distances S+S^+ and S∗S^*, defined as S+=S/(4V/R)1/2S^+ = S/(4V/R)^{1/2} and S∗=(S/R)/ScS^* = (S/R)/S_c, respectively [9, 10]. The dimensionless capillary force is given by F∗=F/(2πRγ)F^* = F/(2\pi R\gamma). The forces corresponding to all the capillary bridge approximations are active until the liquid bridge ruptures at the distance ScS_c.

Simplified Willett approximation

The simplified Willett approximation [9] which already existed in MercuryDPM is defined as follows:

F∗=cos⁡θ1+2.1S++10S+2(2)F^* = \frac{\cos\theta}{1 + 2.1S^+ + 10S^{+2}} \tag*{(2)}

Classical Willett approximation

The classical Willett approximation [10] which is now implemented in MercuryDPM is defined as follows:

ln⁡F∗=f1−f2exp⁡(f3ln⁡S++f4ln⁡2S+)(3)\ln F^* = f_1 - f_2 \exp\left(f_3 \ln S^+ + f_4 \ln^2 S^+\right) \tag*{(3)}

where

f1=(−0.44507+0.050832θ−1.1466θ2)+(−0.1119−0.000411θ−0.1490θ2)ln⁡V∗+(−0.012101−0.0036456θ−0.01255θ2)ln⁡2V∗+(−0.0005−0.0003505θ−0.00029076θ2)ln⁡3V∗\begin{aligned} f_1 ={}& \left(-0.44507 + 0.050832\theta- 1.1466\theta^2\right) + \\ & \left(-0.1119 - 0.000411\theta- 0.1490\theta^2\right)\ln V^* + \\ & \left(-0.012101 - 0.0036456\theta- 0.01255\theta^2\right)\ln^2 V^* + \\ & \left(-0.0005 - 0.0003505\theta- 0.00029076\theta^2\right)\ln^3 V^* \end{aligned}
f2=(1.9222−0.57473θ−1.2918θ2)+(−0.0668−0.1201θ−0.22574θ2)ln⁡V∗+(−0.0013375−0.0068988θ−0.01137θ2)ln⁡2V∗\begin{aligned} f_2 ={}& \left(1.9222 - 0.57473\theta- 1.2918\theta^2\right) + \\ & \left(-0.0668 - 0.1201\theta- 0.22574\theta^2\right)\ln V^* + \\ & \left(-0.0013375 - 0.0068988\theta- 0.01137\theta^2\right)\ln^2 V^* \end{aligned}
f3=(1.268−0.01396θ−0.23566θ2)+(0.198+0.092θ−0.06418θ2)ln⁡V∗+(0.02232+0.02238θ−0.009853θ2)ln⁡2V∗+(0.0008585+0.001318θ−0.00053θ2)ln⁡3V∗\begin{aligned} f_3 ={}& \left(1.268 - 0.01396\theta- 0.23566\theta^2\right) + \\ & \left(0.198 + 0.092\theta- 0.06418\theta^2\right)\ln V^* + \\ & \left(0.02232 + 0.02238\theta- 0.009853\theta^2\right)\ln^2 V^* + \\ & \left(0.0008585 + 0.001318\theta- 0.00053\theta^2\right)\ln^3 V^* \end{aligned}
f4=(−0.010703+0.073776θ−0.34742θ2)+(0.03345+0.04543θ−0.09056θ2)ln⁡V∗+(0.0018574+0.004456θ−0.006257θ2)ln⁡2V∗\begin{aligned} f_4 ={}& \left(-0.010703 + 0.073776\theta- 0.34742\theta^2\right) + \\ & \left(0.03345 + 0.04543\theta- 0.09056\theta^2\right)\ln V^* + \\ & \left(0.0018574 + 0.004456\theta- 0.006257\theta^2\right)\ln^2 V^* \end{aligned}

Bagheri approximation

The approximation recently proposed by Bagheri et al. [10] which is now implemented in MercuryDPM is defined as follows:

F∗=F0∗1+asS∗1+cθasbsS∗+cθbsS∗2(4)F^* = F_0^* \frac{1+a_sS^*}{1+c_\theta a_sb_sS^*+c_\theta b_sS^{*2}} \tag*{(4)}

where

F0∗=(1−0.3823V∗0.2586)(1−aθsin⁡bθθ)F_0^* = \left(1-0.3823V^{*0.2586}\right)\left(1-a_\theta\sin^{b_\theta}\theta\right)

with the following parameters

as=−0.3319V∗0.4974+0.6717V∗0.1995bs=13.84V∗−0.3909−12.11V∗−0.3945cθ=acθ3+bcθ+1ac=−0.007815(ln⁡V∗)2−0.2105ln⁡V∗−1.426bc=−1.78V∗0.8351+0.6669V∗−0.0139aθ=0.4158V∗0.2835+0.6474bθ=−0.2087V∗0.3113+2.267.(5)\begin{aligned} a_s &= -0.3319V^{*0.4974}+0.6717V^{*0.1995}\\ b_s &= 13.84V^{*-0.3909}-12.11V^{*-0.3945}\\ c_\theta&= a_c\theta^3+b_c\theta+1\\ a_c &= -0.007815(\ln V^*)^2-0.2105\ln V^*-1.426\\ b_c &= -1.78V^{*0.8351}+0.6669V^{*-0.0139}\\ a_\theta&= 0.4158V^{*0.2835}+0.6474\\ b_\theta&= -0.2087V^{*0.3113}+2.267. \tag*{(5)} \end{aligned}

Code snippets

Code snippets for implementation of LiquidBridgeClassicalWillet
<snippet>
    <content><![CDATA[
void LiquidBridgeClassicalWilletInteraction::computeAdhesionForce()
{
    const LiquidBridgeClassicalWilletSpecies* species = getSpecies();
    if (getOverlap() >= 0)
    {
        wasInContact_ = true;
        Mdoubl effectiveRadius = 2.0 * getEffectiveRadius();
        Mdoubl V_star = species->getLiquidBridgeVolume() / std::pow(effectiveRadius,3);

        Mdoubl f1 = (-0.44507 + 0.050832*species->getContactAngle() - 1.1466*std::pow(species-
>getContactAngle(),2)) +
            (-0.1119 - 0.000411*species->getContactAngle() - 0.1490*std::pow(species-
>getContactAngle(),2))*std::log(V_star) +
            (-0.012101 - 0.0036456*species->getContactAngle() - 0.01255*std::pow(species-
>getContactAngle(),2))*std::pow(std::log(V_star),2) +
            (-0.0005 - 0.0003505*species->getContactAngle() - 0.00029076*std::pow(species-
>getContactAngle(),2))*std::pow(std::log(V_star),3);

        Mdoubl F_star = std::exp(f1);

        Mdoubl fdotn = -2.0 * constants::pi * effectiveRadius * species->getSurfaceTension() * F_star;
        addForce(getNormal() * fdotn);
    }
    else if (wasInContact_)
    {
        Mdoubl effectiveRadius = 2.0 * getEffectiveRadius();
        Mdoubl s_c = getOverlap() * std::sqrt(effectiveRadius / species->getLiquidBridgeVolume())/2;
        Mdoubl V_star = species->getLiquidBridgeVolume() / std::pow(effectiveRadius,3);

        Mdoubl f1 = (-0.44507 + 0.050832*species->getContactAngle() - 1.1466*std::pow(species-
>getContactAngle(),2)) +
            (-0.1119 - 0.000411*species->getContactAngle() - 0.1490*std::pow(species-
>getContactAngle(),2))*std::log(V_star) +
            (-0.012101 - 0.0036456*species->getContactAngle() - 0.01255*std::pow(species-
>getContactAngle(),2))*std::pow(std::log(V_star),2) +
            (-0.0005 - 0.0003505*species->getContactAngle() - 0.00029076*std::pow(species-
>getContactAngle(),2))*std::pow(std::log(V_star),3);

        Mdoubl f2 = (1.9222 - 0.57473*species->getContactAngle() - 1.2918*std::pow(species-
>getContactAngle(),2)) +
            (-0.0668 - 0.1201*species->getContactAngle() - 0.22574*std::pow(species-
>getContactAngle(),2))*std::log(V_star) +
            (-0.003375 - 0.006988*species->getContactAngle() - 0.01137*std::pow(species-
>getContactAngle(),2))*std::pow(std::log(V_star),2);

        Mdoubl f3 = (1.268 - 0.01396*species->getContactAngle() - 0.23566*std::pow(species-
>getContactAngle(),2)) +
            (0.198 +0.092*species->getContactAngle() - 0.06418*std::pow(species-
>getContactAngle(),2))*std::log(V_star) +
            (0.02232 + 0.02238*species->getContactAngle() - 0.009853*std::pow(species-
>getContactAngle(),2))*std::pow(std::log(V_star),2) +
            (0.0008585 + 0.001318*species->getContactAngle() - 0.00053*std::pow(species-
>getContactAngle(),2))*std::pow(std::log(V_star),3);

        Mdoubl f4 = (-0.010703 + 0.073776*species->getContactAngle() - 0.34742*std::pow(species-
>getContactAngle(),2)) +
            (0.03345 + 0.04543*species->getContactAngle() - 0.09056*std::pow(species-
>getContactAngle(),2))*std::log(V_star) +
            (0.0018574 + 0.004456*species->getContactAngle() - 0.006257*std::pow(species-
>getContactAngle(),2))*std::pow(std::log(V_star),2);

        Mdoubl F_star = std::exp(f1 - f2*std::exp(f3*std::log(s_c)+f4*std::pow(std::log(s_c),2)));

        Mdoubl fdotn = -2.0 * constants::pi * effectiveRadius * species->getSurfaceTension() * F_star;
        addForce(getNormal() * fdotn);
    }
}
    ]]></content>
    <tabTrigger></tabTrigger>
    <description></description>
    <!-- Optional: Set a scope to limit where the snippet will trigger -->
    <!-- <scope >source.python</scope> -->
</snippet>

Figure B.5. Code snippets for implementation of LiquidBridgeClassicalWillet.

Code snippets for implementation of LiquidBridgeBagheri
<snippet>
  <content><![CDATA[
void LiquidBridgeBagheriInteraction::computeAdhesionForce()
{
    const LiquidBridgeBagheriSpecies* species = getSpecies();
    if (getOverlap() >= 0)
    {
        wasInContact_ = true;
        Mdouble effectiveRadius = 2 * getEffectiveRadius();
        Mdouble V = species->getLiquidBridgeVolume()/std::pow(effectiveRadius,3);

        // full model according to Eq. 24-26 from Bagheri et al. 2024, CPM
        Mdouble f0 = 2*constants::pi*(1 - 0.3823*std::pow(V,0.2586));
        // std::cout << species->getMeanRadius() << std::endl;
        Mdouble a = 0.4158*std::pow(V,0.2835)-0.6474;
        Mdouble b = -0.2087*std::pow(V,0.3113)+2.267;
        Mdouble ft = (1 - a*std::pow(std::sin(species->getContactAngle()),b))*f0;

        Mdouble force_full_fit = ft;

        Mdouble fdotn = -force_full_fit * effectiveRadius * species->getSurfaceTension();

        addForce(getNormal() * fdotn);
    }
    else if (wasInContact_)
    {
        Mdouble effectiveRadius = 2 * getEffectiveRadius();
        Mdouble V = species->getLiquidBridgeVolume()/std::pow(effectiveRadius,3);
        Mdouble S_rupture = (1.0 + species->getContactAngle()*0.5)*
(std::pow(V,1.0/3.0)+0.1*std::pow(V,2.0/3.0));
        Mdouble S = -getOverlap()/(effectiveRadius+S_rupture);
        Mdouble Log_V = std::log(V);

        // full model according to Eq. 27-30 from Bagheri et al. 2024, CPM
        Mdouble f0 = 2*constants::pi*(1 - 0.3823*std::pow(V,0.2586));

        Mdouble a = 0.4158*std::pow(V,0.2835)+0.6474;
        Mdouble b = -0.2087*std::pow(V,0.3113)+2.267;
        Mdouble ft = (1 - a*std::pow(std::sin(species->getContactAngle()),b))*f0;

        Mdouble a_s = -0.3319*std::pow(V,0.4974) + 0.6717*std::pow(V,0.1995);
        Mdouble b_s = 13.84*std::pow(V,-0.3909) - 12.11*std::pow(V,-0.3945);

        Mdouble a_t_a = -0.007815*std::pow(Log_V,2)-0.2105*Log_V -1.426;
        Mdouble b_t_a = 1.78*std::pow(V,0.8351)+0.6669*std::pow(V,0.01391);
        Mdouble a_t = a_t_a*std::pow(species->getContactAngle(),3)+b_t_a*species->getContactAngle()+1.0;

        Mdouble force_full_fit = ft*(1+a_s*S)/(1+a_t*b_s*a_s*S+a_t*b_s*std::pow(S,2));

        Mdouble fdotn = -force_full_fit * effectiveRadius * species->getSurfaceTension();

        addForce(getNormal() * fdotn);
    }
}
  ]]></content>
  <tabTrigger></tabTrigger>
  <description></description>
  <!-- Optional: Set a scope to limit where the snippet will trigger -->
  <!-- <scope>source.python</scope> -->
</snippet>

Figure B.6. Code snippets for implementation of LiquidBridgeBagheri.

References

  1. [5]S. Roy, A. Singh, S. Luding, T. Weinhart, Micro-macro transition and simplified contact models for wet granular materials, Computational Particle Mechanics 3 (2016) 449–462. doi:10.1007/s40571-015-0061-8.DOI
  2. [6]K. Johanson, Y. Rabinovich, B. Moudgil, K. Breece, H. Taylor, Relationship between particle scale capillary forces and bulk unconfined yield strength, Powder Technology 138 (2003) 13–17. doi:10.1016/j.powtec.2003.08.037.DOI
  3. [7]C. McKenna-Neuman, W. G. Nickling, A theoretical and wind tunnel investigation of the effect of capillary water on the entrainment of sediment by wind, Canadian Journal of Soil Science 69 (1989) 79–96. doi:10.4141/cjss89-008.DOI
  4. [8]R. Price, P. M. Young, S. Edge, J. N. Staniforth, The influence of relative humidity on particulate interactions in carrier-based dry powder inhaler formulations, International Journal of Pharmaceutics 246 (2002) 47–59. doi:10.1016/s0378-5173(02)00359-9.DOI
  5. [9]C. D. Willett, M. J. Adams, S. A. Johnson, J. P. K. Seville, Capillary bridges between two spherical bodies, Langmuir 16 (2000) 9396–9405. doi:10.1021/la000657y.DOI
  6. [10]M. Bagheri, S. Roy, T. Pöschel, Approximate expressions for the capillary force and the surface area of a liquid bridge between identical spheres, Computational Particle Mechanics (2024). doi:10.1007/s40571-024-00772-5.DOI
  7. [11]T. Weinhart, L. Orefice, M. Post, M. P. van Schrojënstein Lantman, I. F. Denissen, D. R. Tunuguntla, J. Tsang, H. Cheng, M. Y. Shaheen, H. Shi, et al., Fast, flexible particle simulations – An introduction to MercuryDPM, Computer Physics Communications 249 (2020) 107129. doi:10.1016/j.cpc.2019.107129.DOI
  8. [12]A. R. Thornton, T. Plath, I. Ostanin, H. Götz, J.-W. Bisschop, M. Hassan, R. Roeplal, X. Wang, S. Pourandi, T. Weinhart, Recent advances in MercuryDPM, Mathematics in Computer Science 17 (2023) 13. doi:10.1007/s11786-023-00562-x.DOI
  9. [13]N. V. Brilliantov, F. Spahn, J.-M. Hertzsch, T. Pöschel, Model for collisions in granular gases, Physical Review E 53 (5) (1996) 5382. doi:10.1103/PhysRevE.53.5382.DOI
  10. [14]C. Thornton, Granular Dynamics, Contact Mechanics and Particle System Simulations, Springer, 2015. doi:10.1007/978-3-319-18711-2.DOI

Paper details

vlnc:W_zbkq4ty4
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents