[henry]

Hi Sébastien,

Thanks for testing the two variants. For general use we will keep the version with fv3 but if there is interest we could add a switch to allow users to run the standard form or we could derive the version with fv3 from the standard form so that either would be run-time selectable.

Do you have an interest in any of the other SA variants?

H

[paulo]

Thanks a lot Henry for the info,

This means that I can switch wall functions on / off only changing the boundary type ?

Thanks again,

Best regards,

Paulo Rocha

[henry]

With SpalartAllmaras yes wall-functions are selected as boundary conditions on nut. The other low-Re models do not use wall-functions, optionally or otherwise.

H

[paulo]

Thank you very much Henry.

Paulo Rocha.

[paulo]

Thank you very much Henry.

Paulo Rocha.

[egp]

Here is a comparison of 3 cases for a flat plate at Re = 1e6 with sublayer resolution.

Case 1: fv3 != 1, fv2=1.0 - chi/(1.0 + chi*fv1), incorrect, but current distro.
Case 2: fv3 = 1, fv2=1.0 - chi/(1.0 + chi*fv1), baseline SA
Case 3: fv3 != 1, fv2=1.0/pow3(scalar(1) + chi/Cv2_), NASA LaRC fv3 formulation

I would vote for baseline SA as standard implementation in OF. If you read http://turbmodels.larc.nasa.gov/spalart.html, you will see that the fv3 model is not recommended.

[egp]

[seb62]

Hi Henry,

I should say that I am working in industry
and for the moment we are testing openFOAM
on various flows. So we haven't yet studied in details which SA we would like to have, but
talking briefly with some collegues, the SA
with Edwards correction and rotation/curvature correction seem interesting.

Seb

[seb62]

Hi Henry,

I should say that I am working in industry
and for the moment we are testing openFOAM
on various flows. So we haven't yet studied in details which SA we would like to have, but
talking briefly with some collegues, the SA
with Edwards correction and rotation/curvature correction seem interesting.

Seb

[henry]

Hi Eric,

Thanks for the comparison, it confirms what we suspected that with the correct fv2 the fv3 term behaves satisfactorally.

> I would vote for baseline SA as standard implementation in OF. If you read
> http://turbmodels.larc.nasa.gov/spalart.html, you will see that the fv3 model is not recommended.

Not recommended but necessary for stability for complex flows as I have explained in this thread. What I would like to know is what is the "recommended" variant which solved the problem of negative generation?

Hi Sébastien,

I also think that the Edwards variant is interesting. It seems more natural to me that the turbulence production should be based on shear rather than rotation. Perhaps we should give this one a go.

H

[wei_wu]

Hello, Mr Weller
I'm using the Spalart-Allmaras model with roughness wall functions to calculate the wall roughness problem. Do you know what roughness wall functions are they using and what is the source paper?

Thanks

Wei

Originally Posted by henry

Not true, there are several low-Re turbulence models included in OpenFOAM:

LamBremhorstKE
LaunderSharmaKE
LienCubicKELowRe
LienLeschzinerLowRe
SpalartAllmaras

all of which will run without wall-functions. With SpalartAllmaras wall-functions are optional and should be used where the mesh is not fine enough to resolve the boundary-layer.

H

[charlie]

Originally Posted by henry

Hi Sébastien,

Thanks for the interesting link, it seems they are not very happy with the fv3 term: " It was devised to prevent negative values of the source term, and is not recommended because of unusual transition behavior at low Reynolds numbers (see Spalart, P. R., AIAA 2000-2306, 2000)." but the source term going negative does happen in complex flow cases and causes nuTilda to go negative! so we need a fix of some kind. It is not clear if any of the alternative formulations are preferable, perhaps the SA-Edwards or SA-salsa are better; at least they both have a well-posed generation term.

H

Dear Dr. Weller, dear all,

Thanks for this interesting and important discussion. I brushed with this issue recently, when implementing the original SA formulation in our in-house solver: I found the same stability problem in very specific regions of the flow in a "slightly complex" test case. I contacted the model authors and they recommended some additional limiters to prevent Stilde from reaching zero or going negative, which fixes this problem. The limiter we implemented is:

chi >= 1E-03
sTilde >= 1E-05*U_ref/L_ref

(we have a user-defined reference length and velocity, however it is simply important to make sure sTilde doesn't reach zero).

After that discussion, the LaRC web page you have cited was updated with a one-line note at the end of the original SA model description.

I wouldn't recommend the SA-Edwards model, as contrary to the authors' intention, it doesn't give the same results as the original model (rather, significantly reduced skin friction).

My hope is that by implementing the original SA with these limiters, you will be able to keep the model to a single, all-purpose variant that corresponds to the "industry standard".

Best regards,

Charlie.

[henry]

I am not a big fan of clipping; I would prefer an approach where the formulation of the terms naturally obey the physical constraints. Do you have results that show that the original formulation with the clipping is better than the "Ashford" correction we currently use and distribute with OpenFOAM?

H

[charlie]

Dear Henry,

I understand and share your concern - these things are certainly "not nice". However, in this case I was happy with the modification, for the following reasons:

• It does not give rise to discontinuities in the converged solution (i.e. sTilde remains above zero everywhere), rather simply prevents a stable but spurious branch in the system of model equations.
• It is "sanctioned" by the model authors and standard in the implementation of the original S-A model in many solvers, including the NASA solvers.

Now to your question:
Do you have results that show that the original formulation with the clipping is better than the "Ashford" correction we currently use and distribute with OpenFOAM?

I'm not familiar with the "Ashford" correction and didn't manage to find the PhD thesis easily, but assume this refers to the f_v3 modification. The f_v3 was by the way proposed by Spalart as an attempt to fix the problem of negative sTilde without clipping. I don't have results to show that the original+clipping is better than f_v3, since I haven't tried f_v3. All I have are citations that discourage the use of f_v3, for example Spalart writes in "Trends in Turbulence Treatments" (AIAA 2000-2306):

"...the f_v3 function, which we have recommended privately to some users. The f_v3 formula has an odd effect on transition at low Reynolds numbers, which we never fully understood. It was devised to prevent negative values of sTilde. We now recommend, instead, taking for sTilde the larger of the original f_v1 definition and 0.3 x S."

So, in summary, I just have "hearsay" and no concrete results of my own to say that one is better than the other. However, reputable and authoritative sources discourage the use of the unpublished f_v3. I would therefore personally be happier to see the original model with clipping as the standard OpenFOAM implementation. However, due to the source code acces it would of course be no problem for me to implement my preferred version should you decide to go for f_v3!

I hope this information is useful.

Best regards,

Charlie.

[impecca]

Hi, I am following up LES version of Spalart-allmaras and have some questions.

Why fv3 is (which is from V1.7) ?

00067 tmp<volScalarField> SpalartAllmaras::fv3() const
00068 {
00069 volScalarField chi = nuTilda_/nu();
00070 volScalarField chiByCv2 = (1/Cv2_)*chi;
00071
00072 return
00073 (scalar(1) + chi*fv1())
00074 *(1/Cv2_)
00075 *(3*(scalar(1) + chiByCv2) + sqr(chiByCv2))
00076 /pow3(scalar(1) + chiByCv2);
00077 }


rather than

fv3 = (1 + chi*fv1)*(1-fv2)/chi;

which is from SA-fv3 (http://turbmodels.larc.nasa.gov/spalart.html )

If this is because 'negative source' term which Henry mentioned on this thread, is there any reference for the current version?


Thanks

Yusik

[boger]

This appears to be an issue again in 1.6-ext. In particular, 1.6-ext has fv2 = 1.0 - chi/(1.0 + chi*fv1), but also has fv3 != 0, which is the "Case 1" combination in Eric's earlier post that produces incorrect results.

It looks like the fv2 term was changed recently to move away from the "SA-fv3" model, but in that case, fv3 needs to be removed from the calculation of Stilda as well.

[vkrastev]

In OF-1.6 the SA implementation is the so called SA-fv3 one, with fv2 consistent with the presence of fv3...This means that using this kind of implementation my calculations should be stable, but after reading all the replies in this tread still I don't understand if they should be also reliable...ok for the stability of the SA-fv3 model in complex cases, but what about the accuracy? However, I have also another question about the SA model usage: I've had a look into the source code for the nutSpalartAllmarasWallFunction and the nutSpalartAllmarasStandardWallFunction....Well, the latter is consistent with the nutWallFunction used in other RANS turbulence models, but the former seems to me quite obscure: can someone give me a brief explanation of what exactly it does? And, finally, does the SA model require to set nutSpalartAllmarasWallFunction for low-Re calculations? If it so, is this wall function suitable also for coarser grids or should it be replaced by nutSpalartAllmarasStandardWallFunction in such cases?

Thank you in advance for any opinion and/or suggestion

V.

[boger]

I can't answer all of your questions, but I can tell you that nutSpalartAllmarasWallFunction is finding the friction velocity by solving Spalding's law of the wall, which is a non-linear equation for the law-of-the-wall that is valid from the wall out through the logarithmic region. The only place I could quickly find it written is in Eq 59 of this link.

[vkrastev]

Originally Posted by boger

I can't answer all of your questions, but I can tell you that nutSpalartAllmarasWallFunction is finding the friction velocity by solving Spalding's law of the wall, which is a non-linear equation for the law-of-the-wall that is valid from the wall out through the logarithmic region. The only place I could quickly find it written is in Eq 59 of this link.

Ok, so if I understand properly from the reference this is simply a more general form of wall function (compared with the standard one), wich gives a continuos solution for the friction velocity also in the buffer layer: this solution is then used to enforce a proper value for nut in the first computational node, right? In practice, this wall function could be (in principle) applied also to other RANS models, but if so I'm wondering why the OF programmers have decided to call it "SpalartAllmarasWallFunction"...However, thanks a lot, your link and reply were really helpful!

Best Regards

V.

[0.1 watts]

Originally Posted by impecca

Hi, I am following up LES version of Spalart-allmaras and have some questions.

Why fv3 is (which is from V1.7) ?

00067 tmp<volScalarField> SpalartAllmaras::fv3() const
00068 {
00069     volScalarField chi = nuTilda_/nu();
00070     volScalarField chiByCv2 = (1/Cv2_)*chi;
00071 
00072     return
00073         (scalar(1) + chi*fv1())
00074        *(1/Cv2_)
00075        *(3*(scalar(1) + chiByCv2) + sqr(chiByCv2))
00076        /pow3(scalar(1) + chiByCv2);
00077 }

rather than

fv3 = (1 + chi*fv1)*(1-fv2)/chi;

which is from SA-fv3 (http://turbmodels.larc.nasa.gov/spalart.html )

If this is because 'negative source' term which Henry mentioned on this thread, is there any reference for the current version?


Thanks

Yusik

This calculation of fv3 is still present in 2.1 and 2.0.1 in both the DES and RANS formulations (and by extension the DDES model too). I believe it must produce different results to the NASA description.

Is this term perhaps referenced to Ashford's thesis? I can't access it so I would be grateful if anyone can confirm whether it came from his thesis, or if not where it came from (or whether it is a bug).

EDIT: From looking at open access papers that do cite his thesis, I see that the fv3 term is the same as the NASA description but different to what's in the OF source code, so I suspect it doesn't come from Ashford.