mesh resolution for DNS with high prandtl number

hnemati · January 10, 2014, 10:14

Dears
I am doing DNS with high prandtle number fluid (Pr=15).
1- Could someone tell me the best resolution for DNS.
2- For high Pr number I should go finer mesh, Could someone tell me what is the relation between Pr and factor of mesh resolution.
Bests

FMDenaro · January 10, 2014, 10:39

Quote:

Originally Posted by hnemati

Dears
I am doing DNS with high prandtle number fluid (Pr=15).
1- Could someone tell me the best resolution for DNS.
2- For high Pr number I should go finer mesh, Could someone tell me what is the relation between Pr and factor of mesh resolution.
Bests

Just consider the Peclet number = Re*Pr in the energy equation as the same way as you consider only the Re number in the momentum

beer · January 10, 2014, 23:40

I think I remember from University that the resolution should be the kolmogorov length.

http://en.wikipedia.org/wiki/Kolmogorov_microscales

It is not alway ease to find the right epsilon. You can try to estimate it due to Reynolds number and boundary condition or run a coarse grid simulation with k-e turbulence.
How do you want to run a DNS anyway. I remember running some LES-code at University for several weeks on clusters (and DNS can be factors of magnitude more expensive).

Anyway, good luck and fun

January 10, 2014, 10:14	mesh resolution for DNS with high prandtl number	#1
hnemati Member SAM Join Date: Apr 2012 Posts: 74 Rep Power: 13	Dears I am doing DNS with high prandtle number fluid (Pr=15). 1- Could someone tell me the best resolution for DNS. 2- For high Pr number I should go finer mesh, Could someone tell me what is the relation between Pr and factor of mesh resolution. Bests

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
how to set periodic boundary conditions	Ganesh	FLUENT	15	November 18, 2020 06:09
Turbulent Prandtl number	Ravenn	FLUENT	14	June 20, 2017 11:13
ParaView: Number of cells/points in mesh and field don't match	Jaggy_Snake	OpenFOAM Bugs	2	March 1, 2017 09:06
[snappyHexMesh] non-smooth mesh	Svensson	OpenFOAM Meshing & Mesh Conversion	11	January 18, 2012 09:13
DecomposePar unequal number of shared faces	maka	OpenFOAM Pre-Processing	6	August 12, 2010 09:01

January 10, 2014, 23:40		#3
beer Member Join Date: Dec 2012 Posts: 92 Rep Power: 13	I think I remember from University that the resolution should be the kolmogorov length. http://en.wikipedia.org/wiki/Kolmogorov_microscales It is not alway ease to find the right epsilon. You can try to estimate it due to Reynolds number and boundary condition or run a coarse grid simulation with k-e turbulence. How do you want to run a DNS anyway. I remember running some LES-code at University for several weeks on clusters (and DNS can be factors of magnitude more expensive). Anyway, good luck and fun