Turbulence and Surface Roughness Issues in Wind Modeling

Hello FEATool community,

I am a college undergraduate engineering student using FEATool to model how wind flows around a simple building (cube or rectangle) for a senior year project. I have come across 2 big issues when trying to model this.

1.)Because air's Reynolds Number is so high, my velocity spikes and goes extremely turbulent. I've attached two images of a small scale and large scale study. I've tried increasing the maximum non-linear iterations with a finer grid size with some improvement, but I was wondering if there was any way to better model what I am trying to. I am currently using the stationary solver and using mumps linear solver in the pdf attachedFEAToolWindHelp.pdf

2.)Is there any way to implement some sort of surface roughness at the walls? I am currently using no slip at the walls. I had seen the Fluid-Structure equations seem to be able to do it but it the documentation says it is more for laminar flow.

Help would be greatly appreciated and thank you in advance.

Best,

Dear Stefano,

1) It looks like either your simulation has not converged and/or is quite underresolved (requiring a much finer grid). Simultion with the built-in Navier-Stokes physics mode would require some additional turbulence model, such as implemented in the post on Custom Implementation of an Algebraic Turbulence Model in FEATool Multiphysics (which will be available as a built-in option with version 1.12).

2) There are currently no pre-defined boundary conditions for accounting for surface roughness. One option would export OpenFOAM problem case files (by using the "Export" button in the OpenFOAM Solver settings dialog box) and manually edit and run the simulations for example replacing all instances of "nutkWallFunction" with "nutkRoughWallFunction" which should be in the file (0/nut) when using a turbulence model (wallfunction definitions for k/epsilon/omega might also need to be modified).

Stefano Fav wrote

I am a college undergraduate engineering student using FEATool to model how wind flows around a simple building (cube or rectangle) for a senior year project. I have come across 2 big issues when trying to model this.

1.)Because air's Reynolds Number is so high, my velocity spikes and goes extremely turbulent. I've attached two images of a small scale and large scale study. I've tried increasing the maximum non-linear iterations with a finer grid size with some improvement, but I was wondering if there was any way to better model what I am trying to. I am currently using the stationary solver and using mumps linear solver in the pdf attachedFEAToolWindHelp.pdf

2.)Is there any way to implement some sort of surface roughness at the walls? I am currently using no slip at the walls. I had seen the Fluid-Structure equations seem to be able to do it but it the documentation says it is more for laminar flow.

Hi admin

when will the version 1.12 be released?
Can you simply update the version 1.11 to 1.12 after the release?

The target releasedate for FEATool version 1.12 is some time early spring (2020), and yes is should be possible to upgrade as ususal from v1.11 (the release naming indicates that all 1.x releases should be backwards compatible, that is you should be able to load an old model into a newer version (although models made in newer versions might not load in older versions due to missing features). The 2/3/4....x releases will indicate that some incompatibilities might occur).

Leon Robers wrote

Hi admin

when will the version 1.12 be released?
Can you simply update the version 1.11 to 1.12 after the release?