Problems of Heat Transfer Modelling in OpenFOAM

foamF · February 18, 2019, 06:41

I just started the heat transfer analysis with OpenFOAM, and I have some questions:

1. In the tutorial "hotRoom" of bouyantBossinesqPimpleFoam, Prt are defined in (i) "transportProperties" file (in "constant" folder), and (ii) "alphat" file with alphatJayatillekeWallFunction (in "0" folder).

How do I understand why Prt are inputted in both files? Are different Prt used in different locations of the domain, e.g. Prt in alphatJayatillekeWallFunction used only for wall boundary and Prt in transportProperties used for elsewhere?

2. In the tutorial "hotRoom" of buoyantPimpleFoam, I didn't aware the input value of Prt, which is unlike bouyantBossinesqPimpleFoam. Anyone knows how to define Prt in buoyantPimpleFoam?

3. why both "p" and "p_rgh" files are required in the "0" folder for buoyantPimpleFoam and bouyantBossinesqPimpleFoam?

It appears that it only needs to set "calculated" for all patches in "p" file. What is the exact purpose of such setting in "p"?

Hope some experts can give me hints on the above questions!!

clapointe · February 18, 2019, 15:15

Off the top of my head, (mostly) in order :

To address points 1 and 2 : the boussinesq approximation is for "compressible" fluids for which density is only a function of temperature. This affects its implementation such that an incompressible thermo/turbulence model is created an then a correction is applied to account for variation in temperature.

So for 1 -- you are correct that the prt in the alphat boundary condition is used for the boundary, and the prt defined in transportProperties is used inside the domain. Many boundary conditions will take parameters from their own dictionary inside their respective 0 folder (independent from a "global" value).

And for 2 -- buoyantPimpleFoam likely creates/uses compressible thermo/turbulence. So prt will not be defined like in buoyantBoussinesqPimpleFoam. OF assumes a prt of unity, but you can input a different value inside the turbulenceCoeffs dictionary in turbulenceProperties (see eg https://github.com/OpenFOAM/OpenFOAM...enceProperties).

Finally, for 3 -- solvers that take gravity into account (like the heat transfer solvers) solve the pEqn in terms of p_rgh, which is the pressure minus the hydrostatic variation. They enforce consistency between p and prgh by making one a function of the other, hence the "calculated".

Caelan

foamF · February 22, 2019, 07:16

thx, Caelean.

I just setup the case and run the simulation with buoyantPimpleFoam.

I used compressible:alphatJayatillekeWallFunction at the solid walls. Anyone can give me advice if there is any requirement on the first cell size with this kind of temperature wall function?

Since I used high Re kOmegaSST, I check only the yPlus between 30 and 300.

February 18, 2019, 06:41	Problems of Heat Transfer Modelling in OpenFOAM	#1
foamF Member Join Date: Aug 2018 Posts: 47 Rep Power: 7	I just started the heat transfer analysis with OpenFOAM, and I have some questions: 1. In the tutorial "hotRoom" of bouyantBossinesqPimpleFoam, Prt are defined in (i) "transportProperties" file (in "constant" folder), and (ii) "alphat" file with alphatJayatillekeWallFunction (in "0" folder). How do I understand why Prt are inputted in both files? Are different Prt used in different locations of the domain, e.g. Prt in alphatJayatillekeWallFunction used only for wall boundary and Prt in transportProperties used for elsewhere? 2. In the tutorial "hotRoom" of buoyantPimpleFoam, I didn't aware the input value of Prt, which is unlike bouyantBossinesqPimpleFoam. Anyone knows how to define Prt in buoyantPimpleFoam? 3. why both "p" and "p_rgh" files are required in the "0" folder for buoyantPimpleFoam and bouyantBossinesqPimpleFoam? It appears that it only needs to set "calculated" for all patches in "p" file. What is the exact purpose of such setting in "p"? Hope some experts can give me hints on the above questions!!

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Current state of conjugate heat transfer in OpenFOAM	Dreoasteh	OpenFOAM	2	April 4, 2023 13:47
Domain Reference Pressure and mass flow inlet boundary	AdidaKK	CFX	75	August 20, 2018 05:37
mass flow in is not equal to mass flow out	saii	CFX	12	March 19, 2018 05:21
Modelling the heat transfer during compression and cooling of natural gas	pano	Main CFD Forum	0	December 10, 2010 15:53
OpenFoam for heat transport problems	Martin Lorenz (Lorenz)	OpenFOAM Running, Solving & CFD	16	March 18, 2009 03:05

February 18, 2019, 15:15		#2
clapointe Senior Member Join Date: Aug 2015 Posts: 494 Rep Power: 14	Off the top of my head, (mostly) in order : To address points 1 and 2 : the boussinesq approximation is for "compressible" fluids for which density is only a function of temperature. This affects its implementation such that an incompressible thermo/turbulence model is created an then a correction is applied to account for variation in temperature. So for 1 -- you are correct that the prt in the alphat boundary condition is used for the boundary, and the prt defined in transportProperties is used inside the domain. Many boundary conditions will take parameters from their own dictionary inside their respective 0 folder (independent from a "global" value). And for 2 -- buoyantPimpleFoam likely creates/uses compressible thermo/turbulence. So prt will not be defined like in buoyantBoussinesqPimpleFoam. OF assumes a prt of unity, but you can input a different value inside the turbulenceCoeffs dictionary in turbulenceProperties (see eg https://github.com/OpenFOAM/OpenFOAM...enceProperties). Finally, for 3 -- solvers that take gravity into account (like the heat transfer solvers) solve the pEqn in terms of p_rgh, which is the pressure minus the hydrostatic variation. They enforce consistency between p and prgh by making one a function of the other, hence the "calculated". Caelan

February 22, 2019, 07:16		#3
foamF Member Join Date: Aug 2018 Posts: 47 Rep Power: 7	thx, Caelean. I just setup the case and run the simulation with buoyantPimpleFoam. I used compressible:alphatJayatillekeWallFunction at the solid walls. Anyone can give me advice if there is any requirement on the first cell size with this kind of temperature wall function? Since I used high Re kOmegaSST, I check only the yPlus between 30 and 300.