[mechem]

Hi all,
There are two output files in each time folder when running MPPICFoam tutorial:

kinematicCloud:UCoeff

kinematicCloud:UTrans

Dimension of scalar "UCoeff" and vector "UTrans" are kg and kg.m/s, respectively.
I think they should be relevant to momentum transfer from solid phase to gas phase for each cell. As I need to plot over line the magnitude of particle velocity or the averaged mass flux of particles in a horizontal row of cells in paraview, UTrans seems to be a good parameter to show that.
I have two questions:
1- What are and how are calculated "UCoeff" and "UTrans"?
2- Why are kinematicCloud:UCoeff and kinematicCloud:UTrans files in time folders not visible in paraview besude U.air, alph.air, p, ...?
Thanks,

[jarpve]

Hello, have you already figure out what are Utrans and Ucoeff? And how are they calculated?. Thanks

[kenyi89]

Hi,

To answer your questions:

Utrans is the transferred momentum (the interaction term) between the fluid phase and the particle phase. The calculation of this parameter is shown in kinematicParcel.c and it basically is a field where each cell contains the sum of the drag forces and the pressure gradient force acting on all the particles in the cell. The pressure gradient force is only included when activated as a force in the case you are looking at.

Ucoeff is a field where each cell contains the sum of the drag coefficients, it's calculation can again be found in kinematicParcel.c

I have written it all out on paper, I have attempted to take a photo, it should at least clarify something .

[jarpve]

Originally Posted by kenyi89

Hi,

To answer your questions:

Utrans is the transferred momentum (the interaction term) between the fluid phase and the particle phase. The calculation of this parameter is shown in kinematicParcel.c and it basically is a field where each cell contains the sum of the drag forces and the pressure gradient force acting on all the particles in the cell. The pressure gradient force is only included when activated as a force in the case you are looking at.

Ucoeff is a field where each cell contains the sum of the drag coefficients, it's calculation can again be found in kinematicParcel.c

I have written it all out on paper, I have attempted to take a photo, it should at least clarify something .

Thank you, I am going to watch it.

[openfoammaofnepo]

Dear Jeroen,

Your explanation is very useful. UTrans is used in KinematicCloud.H [or .C], however, based on your explanation, it has been defined in KinematicParcel.H [or .C]. What is the relation between these class? I did not find the explicit inheritance between them. Thank you so much!

best regards,
OFFO

[kenyi89]

Hi OFFO,

I am no longer working with these codes, but the kinematicParcel refers to the parcel (so essentially a particle representing many particles with same velocity and position) while the kinematicCloud represents a collection of kinematic parcels.

Explicit inheritance in the code is, due to the extreme amount of template programming going on, very difficult to see. For the moment I would just go with the logical relation I described above; how the actual inheritance is done in the code (it might not even by an inheritance relation) is not so interesting I think .

[openfoammaofnepo]

Dear Jeroen,

Thank you so much for your help. So based on your explanations, a cloud is a collection of parcels, and a parcel is a collection of many particles. Does it mean that the ODEs for mass , momentum and energy are solved for the parcel? or the cloud?

These terms are common in multi-phase literature, but for different codes, they may have different meanings. Thank you for your help.

best regards,
OFFO

[kenyi89]

Dear OFFO,

The ODE for momentum for the fluid phase (so not the particles) includes momentum transfer terms which take into account (per fluid cell) what the transferred momentum between the phases is. Due to numerical tricks it appears in two places, once as UCoeff and once as UTrans, but they represent the same momentum transfer.

As far as the parcels go, there you just solve Newtons equation of motion for the parcel (= particle with adjusted radius and mass). Here momentum transfer is included by the drag formulas.

[openfoammaofnepo]

Dear Jeroen,

Thank you for your reply. So the properties for the parcels will be updated for each time step. The interaction (like momentum and heat exchange) between gas and solid or liquid phases happens between Eulerian gas framework and parcels ? Right? Do we need to do some summations of some parcels to obtain the properties of clouds? I am not very clear about the usage of cloud in terms of the phase interactions?

Thank you so much!
OFFO

[kenyi89]

Dear OFFO,

You are correct. For the phase interaction you use the summation fluid-cell wise of the particles, so if for instance 5 parcels are in one cell you add the drag forces up to get the momentum exchange for that cell. These summations give you the UTrans and UCoeff fields needed as input for the fluid momentum equations. Instead of simple adding you can also use different weighting that distribute the momentum exchange over the cells.

As far as I know the cloud is just a collection of parcels and the corresponding class describes properties shared between the parcels.

[openfoammaofnepo]

Thank you for youe explanations!

[ms6918]

Hello,
Thank You so much for the thoughts on this topic. I have a small doubt.
In the fils U.Eqn.H, the source term is written as coalParcels.SU(U)
And in KinematicParcel file, 2 source terms are defined as
Su (Explicit momentum source for particle) and dUtrans (Momentum transfer from the particle to the carrier phase)
Can you kindly clarify the difference between the two.
Since it is written that dUtrans is the Momentum transfer from the particle to the carrier phase, shouldn’t this be used in as a source term in U.Eqn.H instead of Su.

Thank you again for the valuable feedback