At which Re number its considered turbulent? So, the Re number is in the range where the vortex shedding happening, but this is probably for a smooth surface right? Numerically speaking, the surface of the contour of the model is smooth rigth now, which I can compare with this image. But, I assume that if the surface of the contour has roughness, this result could be different?

Nope

I am having such bounding errors, should I get worried and reduce relaxation? I already use 1st order upwinding with gradient limiters. (Resolved k-omega-sst. I set k as 1e-15 at walls. Could this be the culprit, maybe I should set it to something even smaller?)

bounding k, min: 4.28442e-16 max: 2.03794 average: 0.00614192
bounding k, min: -4.05643e-11 max: 2.03743 average: 0.00617002

What is the normal run time of a simulation for pipe flow in pimplefoam?

Is it close to an hour, or between an hour to a day, or between a day to a week, or between a week to a year?

I know it depends on mesh refinement, but I want a crude estimate of what you consider as acceptable run time

I have an i7 Intel CPU with 16GB RAM

how many cells?

steady state or transient?

About 2 years should be enough I'd say, to get a proper fully developed profile. /s

Using a fully transient sim, obviously

100M cells

pimplefoam whenever steady state can be reached

it's hard to answer question like that, it depends on a million other factors

mesh size and quality, boundary conditions, physics modeles, solvers and schemes used, ...

Let me change my question. In practice if you were to set up a case for pipe flow with an 8mm diameter pipe that is 15cm long, are you going to give up the pimplefoam run after an hour, after a day, after a week, or after a year

of all million things that this answer depends on, physical size ISN'T the one

monitor your quantity of interest or residuals

I meant you'd choose your cell size

when QOI stops changing, you're done

residuals must be below , like 1e-6...1e-10, again depends on the problem

I heard running with ddr5 is 2 times faster, so that will save a year!

I got some questions regarding the constant values in the Smagorinsky turbulence model. I've noticed that the orginal formulation for nuSgs is the eq. 12, but openfoam use nuSgs at eq. 11. So, the constant Cs is fonction of Ck and Ce shown in eq. 13. It seems that the constant Ce is from the balance between the subgrid scale energy production and dissipation (eq. 5). So, why this doesn't seems to be present in the original formulation? And finally, what is your method to find Ck and Ce in order to calibrate Cs? Thank for your help.

Is it possible to prescribe turbulent fluctuations (like with turbulentInlet) when using flowRateInletVelocity BC? My inlets and outlets are not oriented along any axes, and I'd preferably prescribe the inlet with a volume flow rate.... Couldn't figure how to do that with the turbulentInlet BC as I only found options to provide velocity components along xyz. Any suggestions are appreciated 🙏 Thanks!

I don't know if its what you looking for or encounter this type of Inlet condition before, but if not, I think its worth to take a look. https://www.openfoam.com/documentation/guides/latest/doc/guide-bcs-inlet-velocity-dfm-fsm.html https://www.openfoam.com/documentation/guides/latest/api/classFoam_1_1turbulentDFSEMInletFvPatchVectorField.html

Thanks for this!! This also seems to have only have options to prescribe velocity in terms of velocity and xyz components, but not as flow rate normal to the selected patch if I understand it correctly.... The problem with using this for my case is that my inlet can't be along coordinate planes, so it is difficult to prescribe components

Did you saw this one? https://www.openfoam.com/documentation/guides/latest/doc/guide-bcs-inlet-flow-rate-inlet.html

My suggestion is to create a bodyforce which is active only in the region near the outlet. The bodyforce can be a randomly geenrated force

Yes, that's what I'm finally using.... Although this BC allows to prescribe flow normal to the patch, it wasn't possible to add turbulent intensity or fluctuations, like it was possible with the other BC

I suppose the only option is to create a small volume ahead of current inlet such that the velocity inlet patch's normal is along one of the coordinate axes and then use the turbulent inlet BC you mentioned earlier