Implementation of Gidaspow drag model

Hello all,

I was looking at the implementation of Gidaspow drag model which is done in the following way in CFDEMcoupling.

Wen and Yu :

betaP = 0.75*( //this is betaP = beta / localPhiP!
rho*voidfraction*CdMagUrLag
/ (ds/scaleDia_*Foam::pow(voidfraction,2.65)) );


Ergun :

betaP = (150 * localPhiP*nuf*rho) //this is betaP = beta / localPhiP!
/ (voidfraction*ds/scaleDia_*phi_*ds/scaleDia_*phi_)
+ (1.75 * magUr * rho)
/((ds/scaleDia_*phi_));


These expressions of betaP, as clearly mentioned in the commented part are divided by localPhiP which is (1- voidfraction). My doubt is, I am unable to track where is this localPhiP compensated? All the expressions should have an additional localPhiP to be consistent. Is this done outside the GidaspowDrag.C file?
I don't see any such division in other drag model implementation. Seemed inconsistent to me.
Can anyone please explain me what I am missing here?

Thanks.

kind regards,
Vinay