search

RCHMT / Pore-Network-Modeling

Pore network model for reactive coupled heat and mass transfer
3 stars 3 forks source link

diffusive conductance #6

Open masoodmoghaddam opened 5 years ago

masoodmoghaddam commented 5 years ago

Until now we've spent a lot of time in the diffusive conductance calculation. I find this issue today: https://github.com/PMEAL/OpenPNM/issues/1069 Its not clear!

masoodmoghaddam commented 5 years ago

here's a sketch of how the new models work: from: https://github.com/PMEAL/OpenPNM/pull/1062 image

masoodmoghaddam commented 5 years ago

@samanjalilian have we considered all of this steps?

masoodmoghaddam commented 5 years ago

I find also: Thermal Conductance image where U is the conductance, in W/(m2 K). But Amin has said: (https://github.com/PMEAL/OpenPNM/issues/900) Units of (thermal_conductance) * Kelvins = Watts. ==> W/(K). So it seems that unit of conductance in openpnm is multiplied by m2 of area:

Units of (conductance) * Units of (Dirichlet BC) = Units of (Neumann BC)

Diffusion conductance: m^3/s

masoodmoghaddam commented 5 years ago

Today we are locking for the source of error in our 2d simulation. We start from the description provided by Amin: https://github.com/PMEAL/OpenPNM/issues/911 which refers to: https://github.com/PMEAL/OpenPNM/pull/935

samanjalilian commented 5 years ago

I implemented what mentioned in : https://github.com/PMEAL/OpenPNM/pull/935/commits/04cdd5bcfcbb43776c6f472d7856eca2820278ae and also I considered the true surface of pores (cross section area of throats is subtracted from surface area of pores (pi*d^2/4)) dasdad

untitled8.txt

masoodmoghaddam commented 5 years ago

To day we received an email form Amin: I spent an afternoon trying to reproduce the data, but so far I can't match the reaction rates (they're a couple orders of magnitudes off).

Since we have moved to OpenPNM v2.0, I had to dig into v1.6 which we used for that paper. I think I might have found the problem that causes the mismatch.

  1. Previously in v1.6, we used to include molar density in diffusive conductance calculations, but we no longer do that in v2.0. For air, the molar density is about 40 mol/m3 which explains almost 2 orders of magnitude of the discrepancy.
  2. However, the discrepancy I observed is more than that. I dug up the old code (you can also view it here) and found a bug that has been fixed in v2.0. The bug is that instead of looking for throat.area on the network, we used to manually calculate it, assuming 3D cylindrical throats (line 50), whereas 2D ones should have been assumed in this case *.
masoodmoghaddam commented 5 years ago

We wait Amin to makes an example. we implement an analytical solution for this problem for verification.