Introduction to transient turbulence modeling (RAS,LES) - Part 1

[@DaylightDigital]

PLEASE do not stop making these, they are great!!!

[@yiannis.3623]

Just Thank you! Your work is a great help for many people!

[@leimiao9185]

In openFOAM 3.*.*, the "LESProperties" and the "RASProperties" are combined in "turbulenceProperties".

[@OpenFOAMJozsefNagy]

+LEI MIAO Yes, things were changed in the newest versions.

[@nourelislamlounis3328]

u don't know how much u helped me man u don't know for real :'(

[@007aha1]

perfekt !!!

[@kennydoritos]

How did you arrive at the definition of the turbulent viscosity at 6:00? nu_tau=0.09 X k^2/epsilon

[@OpenFOAMJozsefNagy]

That is the implementation in the source code.

[@sumitzanje5207]

If we want to use the DNS model for the same configuration, how to modify (or build new) case file for it? Please suggest how to use the DNS turbulence model so that we can compare it with the RANS and LES model.

[@joseomardavalosramirez6380]

Hi József. All your tutorials are great, best Openfoam tutorials ever. Let me ask you a question, could I to use k-omega turbulence model instead k-epsillon?

[@somethingsomething4103]

The videos are quite nice. I hope you come around to make more :).

[@OpenFOAMJozsefNagy]

+Oliver Haas I'm thinking about a tutorial on snappy, but that would take a lot of time and I have a lot to do at work... Let's see, what the future brings.

[@fedezabaleta]

I haven't found a didactic tutorial on snappy. One would be excelent!

[@oscar99999]

Hi József, I want to simulate the flow of a seminfinite 2D domain which corresponds to a horizontal wall moving in the horizontal x-direction. In practice the fluid extends infinitely far away in positive/negative x-direction and positive y-direction. The initial conditions are such that the fluid is at rest and then the plate is started from rest. The pressure should be atmospheric everywhere from the start as well. What are the appropriate boundary conditions for the velocity, pressure on my 'top' and 'inlet' 'outlet' faces? If I where to use a Smagorinsky model what would be the appropriate boundary conditions for NuSgs? many thanks, Oscar

[@OpenFOAMJozsefNagy]

Oscar Wilsby I'm not sure, if I am the right expert for that, but you can shoot me a mail and I can try to answer at leas some of your questions.

[@hardiksharda9673]

Thank You so much You are legend

[@cartoonsindia4072]

kindly make video for internal compressible gas flow turbulence modelling

[@filippopagnoni4517]

Hi József, great video ! Just a little question, why for nuSGS u compute k using the deviatoric part of symm(gradU) tensor? Normally, would not be sufficient to evaluate the magnitude of symm(gradU) in your formula (slide at 8':51'') ? Thank you in advance for your really helpful videos! Filippo

[@OpenFOAMJozsefNagy]

A lot has happened since then in turbulence modeling. Check out the source code for improved models.

[@sacramentofwilderness6656]

I've implemented, using this tutorial as a template, a similar simulation, but the situation on the outlet in the initial time doesn't look very realistic - drive.google.com/drive/folders/1Q03FdTvThdflvtU7lnmqIn5IOdIESEnA?usp=sharing. Is it due to the wrongly chosen boundary conditions? I've set pressure p = 0 on the outlet, inletOutlet on U, and zeroGradient on nut, and for the inlet velocity of (1 0 0). I apologise, If I am asking this question in not a proper place

[@smrutimayapattanaik4077]

Just Great ....

[@simonedebellis4423]

Hi Jòzsef, I have a problem: I'm doing a CFD simulation around a NACA profile of 0.055 m that is in an air flow at the velocity of 25 m/s on x axis using the model K- omega, but I don't how can I estimate the values of K and omega... What values would you insert? Thanks. Simone

[@OpenFOAMJozsefNagy]

I am not an expert in profile modeling. I would suggest you take a look at publications.

[@lizarettflavour]

thank you

[@McHumaty]

József.. I would like to ask if possible, make videos of 'postChannel' solver, for other non-channel geometries.

[@OpenFOAMJozsefNagy]

Dear Marcelo, I am aorry, but I am not an expert in that solver.

[@panittkamma4573]

Hi József I saw you've slightly changed the discrete schemes in this tutorial. Now, I'm working with reacting flow problem but when I've tried to change the scheme (discrete) then the result largely changed. I've been confusing about which discrete is exactly suitable with my problem. Could you recommend about where can I find or get more knowledge to bring me out from this problem. Thank you a lot for your help.

[@OpenFOAMJozsefNagy]

+Panitt Kamma Which scheme did you change? Of course. This is a tutorial, you have to find the order of discretization, that you need for your case.

[@panittkamma4573]

+József Nagy OK I will move following your recommendations.Thanks a lot for your answers plz don't stop to make these beneficial video.

[@pingchang5584]

Hi Jozsef Nagy,, Firstly, thanks for your videos, really helpful. And I have two questions about fvsolution.In fvsolution, tolerance is already exist to control the convergence of the equation,why OF set relTol to control convergence . And I found the value of reTol are varies in many cases. Could you tell me the reason of this? Thanks a lot.

[@OpenFOAMJozsefNagy]

+Ping Chang I guess youtube is not the best place for this discussion, but I would suggest, that you play around with the values in simpleFoam and then in pisoFoam and see the influence.

[@yumengwang151]

Hi I want to know is there similar paper, which the parameter of the BC and domain size almost identical with the video

[@OpenFOAMJozsefNagy]

Un fortunately I am not aware of such a paper. Please do a lit research.

[@yumengwang151]

@@OpenFOAMJozsefNagy Thanks anyway for the response. But could u please also tell me more about the file 0/U. Did you set the inlet velocity as fixedvalue, or create a channel flow simulation before and used the obtained velocity vektorlist as the initial conditions in 0/U. Cause I am doing a progamm, which need to validate my own turbulenz modell for unsteady flow in backward facing step simulation. It would be very thankful if you could help

[@OpenFOAMJozsefNagy]

@@yumengwang151 I did the first simple approach.

[@sumitzanje5207]

What about y+ value? If we are comparing different turbulence models such as k-Omega, k-epsilon, realizable k--epsilon, SST k -omega, etc. Do we need to check y+ value in each case? Do we need to check even after using wall functions? please clarify on this.

[@OpenFOAMJozsefNagy]

Yes, you should check it.

[@TheHarishkg]

Hi, Am new to openfoam. While i run my simulation using overset method in openfoam(OverpimpleDyMFoam) i get following error: FOAM Warning : From function bool Foam:versetPolyPatch::master() const in file oversetPolyPatch/oversetPolyPatch.C at line 149 The master overset patch is not the first patch. Generally the first patch should be an overset patch to guarantee consistent operation. What do you think is the issue?

[@OpenFOAMJozsefNagy]

Sorry, I do not have a lot of experience with overset yet.

[@ashishtiwari1912]

I am looking for implementation of k-omega SST model in openfoam. Can you suggest where could i find those? And I have question related implementation of k-omega SST model. I know that theoretically k=0 at the wall but why is omega not zero at the wall?

[@OpenFOAMJozsefNagy]

It is called kOmegaSST. You usually use wall functions at walls.

[@DaylightDigital]

József, what exactly do you mean when you say that the NS equations are being solved with the pressure gradient term?

[@OpenFOAMJozsefNagy]

Which part of the video do you mean?

[@DaylightDigital]

József Nagy At 12:21 you said "...we are solving this vector-matrix with the pressure gradient term." How, exactly is the pressure gradient being used to solve the equation? Providing me with a reference would be fine, no full explanation is required here.

[@OpenFOAMJozsefNagy]

DaylightDigital It is a question of what parts of the equation you put into the matrix on the left hand side and which part is explicit on the right hand side. Here, the pressur egradient is not inculded into the vector matrix. Check out the phd thesis of Prof. Jasak and/or Henrik Rusche. Both are available online as pdf.

[@killua9369]

Can we have a 2D mesh for LES cases? I always find people use 3D and I don't know why!

[@OpenFOAMJozsefNagy]

yes usually LES is only for 3D. this is only a tutorial here

[@michaelvillame]

Hello Prof. , when I run your tutorial, I got this error in OpenFOAM 7, and I don't know why. any ideas?? --> FOAM FATAL ERROR: Attempt to cast type zeroGradient to type nutWallFunction

[@OpenFOAMJozsefNagy]

This is new... You have to use the BC nutWallFunction in 0/nut

[@symeonsavvopoulos3117]

Very interesting videos. Let me ask you a question as I am new to OpenFOAM. How can you add algebraic equations to your system?

[@OpenFOAMJozsefNagy]

Directly into the source code. Or over function object, if I know it correctly.

[@YangJuntao]

Is there a higher resolution version？

[@OpenFOAMJozsefNagy]

1080p is the max.

[@YangJuntao]

Hi, Jozsef, thanks for your reply. Sorry, I just found out it's my linux firefox browser's issue.

[@hassnhadia8510]

Hi can we contact by e-mail