Learning Forum

[Saeed]

Hello,

1-I wanted to use coarse graining for a CFD-DEM simulation. since in cfd-dem the size of particle should be smaller than CFD cell size and the CGM scale factor is multiplying the size of the particle (particle dimension) , I should be careful that the size of the particle do not get larger than the CFD cell size after multiplied by the CGM scale factor or the size of the original particle should be smaller than CFD cell size ?

 

2- I did simulations with different porosity using CFD-DEM, the porosity was 0, 0.2, 0.6 and I my results for 0.6 are not very good. what could be the reason and how can I improve it? all parameters are same for all simulations. 

[Jackson Gomes]

 

Hi Saeed.

Ideally, particles should be smaller than the CFD cell, but if Vp > Vc, the diffusion algorithm redistributes the solid fraction to neighboring cells, ensuring accuracy. You can refer to the on CFD-DEM coupling in the CFD Coupling Technical Manual, Chapter 6, which discuss the relationship between particle size and CFD cell size, as well as the use of diffusion algorithms to redistribute solid fractions.

We need more information about what is meant by “porosity” (solid fraction or geometric arrangement) and “poor results” (convergence, discrepancies, or another criterion) to identify and address the issue effectively.

Best Regards

 

[Saeed]

Thank you for you answer. 

1- My question is in coarse graining , ideally, the original size of the particle should be smaller or the size of the particle mutiplied by coarse graining factor should be smaller than CFD cell ?

2- in Rocky particle tab, after adding particles, its pososity can be defined too( the proportion of empty space in a material) .

in density simulation, particles are agglomorating at some parts or moving in a way which does not match the experiments and does not make sense and in porosity simulation, particles are disappearing from the system. But there is no converngence issue. 

 

[Jackson Gomes]

Dear Saeed,

The final parcel size is what Rocky will use to perform the averaging calculations to send the Solid Volume Fraction to the CFD solver.

CGM is always a trade-off between computational speed and accuracy. Please check whether your CGM factor is too large - general recommendation is for the smallest scaled up particle to be still at least an order of magnitude smaller than the smallest characteristic length of the domain.

The porosity defined in Rocky’s particle tab refers to the internal porosity of the particles (proportion of empty space within a material).

[Saeed]

1- Thank you for your answer and help.

 

2-My question is , As I am experiencing the issues exalpined, by lowering the density or increaing the porosity what else should be changed that particle behaviors be realistic and system be stabled?

Best

[Jackson Gomes]

Dear Saeed,

Ensure that the Numerical Softening Factor (NSF) is set to an appropriate value, extremely low values can lead to excessive overlap and numerical issues. You can also try to decrease the simulation time step to enhance numerical stability.