[bees1995]

Hi guys,

I was wondering whether it is possible to make small changes to the model once grid indpendence has been achieved? Specifically in my case I'm looking to model flow around an aircraft empennage and I wanted to determine the effects of rudder and elevator input. Rather than doing a grid indpendent study for each case I was hoping to save time and do it for all in one go.

Thanks for your help!

[lukas.lebo]

I don't really have much experience with this, but I'm sure other people can only help you when you get a bit more specific. Like are you considering compressible or incompressible flows? What turbulence model are you using?

When you found a "best practice" for mesh-refinement for one case, you can expect that it will at least behave very close for similar cases. You can always confirm this by doing one more simulation of the same case with a more refined or coarser mesh (±20% computational cost is already enough to tell). If your case runs under extreme conditions and at the edge of grid independency, I'd recommend to validate it for at least one case-variation.

People usually rely on said "best-practice" because you can't really do a mesh study for every commercial project (when they are all similar).

Cheers,
Lukas

[bees1995]

Hi Lukas,

I'm doing an incompressible simulation and using an SST turbulence model.

[bees1995]

But the cases I am running are fairly "safe" in their setup, running at low-speed, incompressible flow over structures that are aerodynamically smooth i.e. no overly turbulent flow, so I would expect some resilience in the mesh. I was just wondering if it's common practice to do so. I'm not looking for super accurate results, so slight errors that occur when the model is altered and the resulting mesh not being independent is fine for me.

[lukas.lebo]

Incompressible RANS (especially SST) simulations have rather stable grid-behaviour when it comes to variations in geometry. You should be good.