Direct implementation of Leonard stress without SGS modeling

eddykendo · October 2, 2018, 00:16

P { margin-bottom: 0.21cm; } Dear. Foamers

At this time, I try to add directly a “Leonard Stress” term (actually, divergence of Leonard stress) in a momentum equation which is calculated using resolved velocity field without the SGS modeling.
I believe that there are two representative ways for this.
The one might be that Leonard stress can be computed in xx.H file such as createFields.H and then added to the momentum equation in UEqn.H. The other might be constructing files, for instance, LeonardStress.H and LeonardStress.C such as linearViscousStress.H, linearViscousStress.C.
Which one is an effective and easy way for that? Otherwise, any recommended method through your experience?
Because I am working on V4.0 and all the files concerning turbulence are based on a template (The second filtering in LES is also needed in order to calculated Leonard stress), I’d like to know the most appropriated way for implementing Leonard stress directly in the momentum equation.
I will appreciate it if you can suggest any opinion for me.

Thank you.

K.C. · October 2, 2018, 05:40

Dear eddykendo,

I would say you have to think about your overall aim of writing this code.

Are you sure, that you won't need it in the future and don't want to implement it in other solvers?
Then writing a single specified xx.H-File and inlcude it to your solver. Compile and have fun with it.

Do you want to implement it to other Foam-Solvers without recompiling every single solver you should use templates, write it as library and link it as turbulence model into existing solvers.

I hope this is usefull for you.

Greets,
K.C.

October 2, 2018, 00:16	Direct implementation of Leonard stress without SGS modeling	#1
eddykendo Member DongHo Join Date: Jul 2013 Posts: 51 Rep Power: 13	P { margin-bottom: 0.21cm; } Dear. Foamers At this time, I try to add directly a “Leonard Stress” term (actually, divergence of Leonard stress) in a momentum equation which is calculated using resolved velocity field without the SGS modeling. I believe that there are two representative ways for this. The one might be that Leonard stress can be computed in xx.H file such as createFields.H and then added to the momentum equation in UEqn.H. The other might be constructing files, for instance, LeonardStress.H and LeonardStress.C such as linearViscousStress.H, linearViscousStress.C. Which one is an effective and easy way for that? Otherwise, any recommended method through your experience? Because I am working on V4.0 and all the files concerning turbulence are based on a template (The second filtering in LES is also needed in order to calculated Leonard stress), I’d like to know the most appropriated way for implementing Leonard stress directly in the momentum equation. I will appreciate it if you can suggest any opinion for me. Thank you.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
LES of turbulent channel flows	cedric_duprat	OpenFOAM Running, Solving & CFD	220	November 18, 2019 05:19
Implicit vs Explicit SGS modelling	syavash	OpenFOAM Running, Solving & CFD	1	December 6, 2016 05:14
CFD Modeling of Direct COntact Membrane Desalination	Ahmad idrees	Main CFD Forum	0	February 19, 2014 17:36
What is really SGS dissipation term in LES?	mrf6166	Main CFD Forum	0	January 5, 2014 07:57
Solvers with "Deterministic Stress" modeling???	Glenn	Main CFD Forum	5	December 22, 2007 11:35

October 2, 2018, 05:40		#2
K.C. Member Join Date: Mar 2015 Posts: 36 Rep Power: 11	Dear eddykendo, I would say you have to think about your overall aim of writing this code. Are you sure, that you won't need it in the future and don't want to implement it in other solvers? Then writing a single specified xx.H-File and inlcude it to your solver. Compile and have fun with it. Do you want to implement it to other Foam-Solvers without recompiling every single solver you should use templates, write it as library and link it as turbulence model into existing solvers. I hope this is usefull for you. Greets, K.C.