force models: definition of drag coefficient

Submitted by kalei on Tue, 09/13/2016 - 12:55

Hallo,

I am trying to understand the implementation of the different force models in CFDEMcoupling and to extract the applied definition of the "drag coefficient" Cd0.

Looking at SchillerNaumannDrag, for me everything is clear. The drag force is given by the following expression (compare SchillerNaumannDrag.C):

drag=0.125*Cd*rho*M_PI*ds*ds*magUr*Ur

and I get the "drag coefficient" Cd0=Cd

Looking at DiFeliceDrag, I can also follow the implementation (compare DiFeliceDrag.C)

drag=dragCoefficient*Ur with dragCoefficient = 0.125*Cd*rho*M_PI*ds*ds*scaleDia_*pow(voidfraction,(2-Xi))*magUr*scaleDrag_;

In this case following my previous procedure, I would extract the "drag coefficient" Cd0=Cd*pow(voidfraction,(2-Xi))

However, I do not understand the implementation of GidaspowDrag (compare GidaspowDrag.C)

drag=dragCoefficient*Ur with dragCoefficient=Vs*betaP*scaleDrag_;

In this case I get the following "drag coefficient" Cd0=dragCoefficient/(0.125*rho*M_PI*ds*ds*scaleDia_*magUr*scaleDrag). However, if I do so Cd0 is off about 7 orders of Magnitude to the values extracted from the previous models. The variable dragCoefficient however seems to be in the right order of magnitude. Therefore, for me it seems that the definition of dragCoefficient ind GidaspowDag seems to be different to that applied in DiFeliceDrag.

Can anybody help me how to Interpret the source code of CFDEM correctly in order to learn which formulations of the "drag coefficient" Cd0 are implemented for the different drag models?

Best
kalei

kalei | Wed, 09/14/2016 - 10:03

Hallo,
I think I had a mistake in the formula I used for analysis. Now I understand the implementation of GidaspowDrag and get the rigth order of Magnitude for the drag coefficient.
Best
kalei