what I mean, both do time-averaging, right?

simpleFoam doesn't "know" it's transient so it's not like it aims for the average of the transient hmmm.. To summarize, I'm not sure, and it probably depends 👍 no need to thank me for all the help

for example AhmedBody works well with simpleFOAM I think. But the sphere we were struggling with not as much

exactly the opposite, I did both on the sphere and got the same results after averaging

I don't know the exact math (yet) but if I understand the principle, URANS is like moving time average where window is each timestep

Then the norm is probably that RANS mean value = URANS mean value. Probably some exception but overall I think you're good with RANS for that purpose

I can thank you now 🙂 thanks

Yeah that's pretty close, based on how statistical descriptions go.

therefore I'm better off with steady-state, unless I want to study transient phenomena

Why are residual tolerances for simpleFoam usually a few orders of magnitude smaller than pimpleFoam's? I mean the

fvSolution.residualControl

With SIMPLE I've seen 1e-6 ... 1e-9 but with PIMPLE more like 1e-3 ... 1e-4 ?

I would add a question to kandelabr's. In case of running PISO with Co<1, do residuals "matter"? Since we go to the next time step anyway

It has bin some time that I understood the difference between PIMPLE and PISO. A good explanation is here https://www.simscale.com/forum/t/cfd-pimple-algorithm/81418

I guess you are simulating RAS?

yes

PIMPLE and SIMPLE are very different algorithms, for very different purposes. You are supposed to use higher residuals on PIMPLE because you are running into stiff problems, and want to advance the solution as fast as possible. Residuals in SIMPLE converge, if the do, to very low residuals when you approach a steady solution.

in fact I never messed around too much with PIMPLE, but with CFX, s steady solution might not get RMS residuals below 1e-4 while the transient one went to 1e-5. Would it be true to say for unsteady solution that PIMPLE residuals drop below a SIMPLE solution? (provided timestep isnt excessive I suppose)

In general, no. The only thing you cmay be able to say about PIMPLE & SIMPLE is that you can arbitrarily increase the Co in PIMPLE in order to obtain a steady solution that resembles that of SIMPLE, but not the opposite.

Thanks a lot, that'll calm me down for a while

until I finish Malalasekera, that is :)

I couldn't get pass the preface of that book 😛

*didn't have to

No no, I honestly couldn't. Somehow I liked more Peric's book.

Hello LES world. Do you have good reference for going through LES modeling ?

Hello, you may have a glance on "Fluid mechanics, turbulent flow and turbulence modeling" book/course notes of Lars Davidson

Here is the link: http://www.tfd.chalmers.se/~lada/comp_turb_model/

See Chapter 18

I see, that one is quite similar... Well, I'll finish what I started and then see what next