[alexksei]

Hello guys,

I am calculating drag force on an aircraft.
I am new to CFD and can't understand how I should model properly the boundary layer. So we have main fuselage and wings and rudders, which are all of different reference lengths.
For example, the fuselage is 10m and rudders are 0.5m. It means that for the same velocity of the flow, the boundary layer thickness of these two bodies are different (flat plane theory). Thus we need to model near-wall mesh differently for these two regions.
At the same time we need to ensure smooth transition from layer mesh to outer volume mesh for both regions. But as the mesh on the fuselage is more coarse than on the rudder, such smooth transition is possible on either one of them.
Does it mean, that we need to choose which region should be modelled well and which one needs to be sacrified? Or any other recommendations. My final goal is to estimate drag coeficient properly.
Thanks!

[crizpi21]

Hi Aleksei,

I am also trying to model accurately the boundary layer to calculate the drag coefficient, although in my case is for a flow around a 3D cube with rounded edges (using OpenFOAM).

For the mesh, I want to have three regions, so I also have this issue with transition between them. My regions are:

1. BL around the obstacle: I need cells with a very small height (in the normal direction to the surface of the cube) to fulfill y+<1 criteria.

2. After the boundary layer, I need a region with a cell size around 1/20 the characteristic length of the cube.

3. In the rest of the domain, the cells should get bigger as I move waya from the obstacle.

May I ask you which software you are using to generate the geometry and mesh? So far, I have not found a good solution.

Thanks in advance

Cristina

[alexksei]

Hi Cristina,

From what I see in your description I think our problems are different.

In your case you need to model boundary layer with smooth transition to the outer mesh.
Your steps should be:
1. Calculate the first cell hight based on your desired y+ value.
2. Create layers, the number of layers should be such, that the last layer has the more or less the same thickness as the first cell of outer mesh.

I think for your basic cube you can use openFoam mesher, though I am not familiar with it.

[crizpi21]

Hi Aleksei,

You are right, in my case I am not trying to model two different boundary layers. I think the steps you suggested seem reasonable.

Could you tell me which software you are using for the mesh and explain a little bit how you are defining the cell size in the different regions, if you do it manually or maybe by writing an script?

So far I have not found any suitable utility for doing the mesh I want so maybe yours could be useful (provided the format can be exported to OpenFOAM).

Thanks and good luck!

Cristina

[alexksei]

In my case I use ANSA as preprocessing (mesh generation, boundary conditions). It supports OpenFoam files export also. Everything is done with GUI. May be you have it at your disposal too.

But in openFoam there is free SnappyHexMesh meshing software (see corresponding tutorials).