Variable Prandtl number on patches

doerschlott · October 7, 2020, 04:38

Hey guys,

I'm currently working on heat transfer phenomena in long pipes (e.g. heat exchanger), which are highly turbulent. Therefore, I'm using the CHT solver "chtMultiRegionFoam" and, for now, k-epsilon turbulence model.

The turbulent heat transfer is partly stated in the fluid field "alphat" with the boundary condition "alphatWallFunction" on patches to neighbouring regions, where one can state an optional constant Prandtl number (0.85 being default) for the heat transfer process. The problem with heat exchangers is strongly increasing fluid temperatures and therefore deviating material properties, including the Prandtl number.

So, is anyone aware of a way to have the Prandtl number being calculated locally, to better resolve the changing heat transfer along the pipe?
Appreciate all input.

Best regards
doerschlott

Wenyuan · October 8, 2020, 04:15

Hi doerschlott,

The Prandtl number used in turbulence models is Prt, which is different from the molecular Prandtl number Pr. Prt is just a closure constant.

doerschlott · October 11, 2020, 06:35

Hi Wenyuan,

thank very much you for your reply. I didn't think of that for some reason. After some research, I found that alphat is calculated as follows:

$alpha_t = \frac{\nu}{Pr} + \frac{\nu_t}{Pr_t}$

Am I right, that Pr is actually calculated based on the given material properties (may it be constant or polynomials e.g.)?

Best regards
doerschlott

October 7, 2020, 04:38	Variable Prandtl number on patches	#1
doerschlott New Member Join Date: Feb 2020 Location: Germany Posts: 17 Rep Power: 6	Hey guys, I'm currently working on heat transfer phenomena in long pipes (e.g. heat exchanger), which are highly turbulent. Therefore, I'm using the CHT solver "chtMultiRegionFoam" and, for now, k-epsilon turbulence model. The turbulent heat transfer is partly stated in the fluid field "alphat" with the boundary condition "alphatWallFunction" on patches to neighbouring regions, where one can state an optional constant Prandtl number (0.85 being default) for the heat transfer process. The problem with heat exchangers is strongly increasing fluid temperatures and therefore deviating material properties, including the Prandtl number. So, is anyone aware of a way to have the Prandtl number being calculated locally, to better resolve the changing heat transfer along the pipe? Appreciate all input. Best regards doerschlott

October 8, 2020, 04:15		#2
Wenyuan New Member Wenyuan Fan Join Date: Mar 2017 Posts: 27 Rep Power: 9	Hi doerschlott, The Prandtl number used in turbulence models is Prt, which is different from the molecular Prandtl number Pr. Prt is just a closure constant. doerschlott likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
decomposePar no field transfert	Jeanp	OpenFOAM Pre-Processing	3	June 18, 2022 12:01
AMI speed performance	danny123	OpenFOAM	21	October 24, 2020 04:13
parallel run OpenFoam	Srinath Reddy	OpenFOAM Running, Solving & CFD	13	February 27, 2019 09:15
[snappyHexMesh] Problem with decomposePar, snappyHexMesh	luca1992	OpenFOAM Meshing & Mesh Conversion	12	August 23, 2017 19:06
Compressor Simulation using rhoPimpleDyMFoam	Jetfire	OpenFOAM Running, Solving & CFD	107	December 9, 2014 13:38

October 11, 2020, 06:35		#3
doerschlott New Member Join Date: Feb 2020 Location: Germany Posts: 17 Rep Power: 6	Hi Wenyuan, thank very much you for your reply. I didn't think of that for some reason. After some research, I found that alphat is calculated as follows: $alpha_t = \frac{\nu}{Pr} + \frac{\nu_t}{Pr_t}$ Am I right, that Pr is actually calculated based on the given material properties (may it be constant or polynomials e.g.)? Best regards doerschlott