[cfdblender]

Is there any way to command STAR CCM+ to target a specific range of Y+ values (ie., between 30 to 100) on a surface?

So far, I've been doing 'trial and error' with the number of prisms and thickness of the prism layer, but it's turned into 'trial and horror'!

Surely, there must be a better way ?

[triple_r]

To calculate y+, you need to know the wall shear stress, and usually you don't know that before solving for the problem, so I'm afraid the best you can do is trial-and-error.

But there are better ways to go about it:

Estimating the wall shear:
Based on your simulation and the problem, you might be able to find some correlations to estimate the shear stress at the wall. For example, if it is a pipe flow, you can use the Moody chart to estimate the friction factor, and from that you can calculate the wall shear stress. This is going to be for the fully developed part of the fluid, so your actual shear will be different, so the y+ value will not be exactly what you are shooting for, so be careful and account for non-developed regions (for example, in entrance region, y+ will be a lot higher than what you shoot for because of higher shear).
If you don't have a good estimate, run the simulation with coarse mesh, and see what the shear is, use that to calculate y+ :-) again, because the elements are coarse near the wall in your first mesh, the actual shear will, in general, be higher than what you calculate in your coarse simulation, so take that into account.

Calculating near wall element thickness:
When you have an estimate of wall shear, you can use the definition of y+ to find out what the thickness of first layer has to be to get you in the range of y+ that you need.

first layer thickness = (desired y+) * (kinematic viscosity) / (friction velocity)

where friction velocity is:

square root of (wall shear stress / density)

Setting parameters in StarCCM+:
Instead of using total thickness and stretch factor, use "Wall Thickness" distribution model for prism layers:

If you are using part-based meshing, under "automated mesh" > Meshers > Prism Layer Mesher, change the Distribution Mode to Wall Thickness. Then under "Defaul Controls" you can set the total thickness, number of layers, and Near Wall Thickness (which should be what you calculated in the previous step).

Look-outs:
In this method you won't be able to set the stretch factor, and usually you want the stretch factor to be small (just a tad over 1.0, something like 1.1 for example). To make sure that is the case, you can use the following formula to estimate the number of layers (this works with geometric progression, which is the default stretching function):

(total thickness) = (near wall thickness) x ( (stretch factor)^n - 1) / (stretch factor - 1)

[cfdblender]

Thank you, triple_r, for an excellent step-by-step and comprehensive answer.

I will start by getting a good plot of wall shear stress. Found this article about it, should be useful: https://wiki.anl.gov/tracc/STAR-CCM%2B_Technical_Notes

[triple_r]

Not a problem :-) Yes, that should help, though for estimating wall shear, you can use the field function that comes with StarCCM+ and the error should be negligible for the first estimate when you are just tuning y+. For your final solution though, you probably want to go with the more accurate definition.

Good luck.