Hello everybody,
I could not simulate wind speed fluctuation by assigning turbulence intensity (TI).
I simulate the fluid-structure interaction by using STAR-CCM+ and Abaqus, and I simulate the flow field of through IDDES. I set field functions to assign inlet velocity and turbulence intensity, respectively. Some points are set to record the time history of wind speed in freestream direction. However, results show that the wind speed do not fluctuate obviously (i.e. the wind speed just slightly vary with time). I calculated the turbulence intensity (TI) according to the definition of TI. The TI was 0.0001%, indicating that the wind speed was almost constant.
I set “Synthetic Turbulence Specification” from “none” to “intensity + length scale” in “continuum” and “region-boundary -inlet-physical conditions” when I just use STAR-CCM+ and do not use Abaqus. In this case, the wind speed vary obviously with time. However, there was warning when I set “Synthetic Turbulence Specification” in fluid-structure interaction:
“The Synthetic Turbulence Specification applied to boundary ‘S.Con.inlet’ in region “flow”, is not compatible with the Motion Specification in that region.”
The Synthetic Turbulence Specification should be suppressed when the mesh morphing of flow field was considered.
The setting of regions and simulation are shown in attachments.
The computational region was 1000 m (x) *500 m (y) * 500 m (z)
The section of structure was 6 m (x) *6 m (y) * 50 m (z)
Is there anything I forget to check? How can I get a certain level of turbulence at a specific spot in the field through inlet boundary specification?
Any feedback on the approach/idea itself is welcome too.
first, to resolve fluctuations in turbulence you have to approach a DNS or, at least, a LES formulation.
then, to see physical fluctuations you have to let the solution developing, that is forget the arbitrary initial condition. That requires to wait some turnover times before to see the correlated fields.
you just need to set up a user-defined function for the inlet flow velocity.
Firstly, I might assume you need LES to get u' under this case, then talk about turbulent parameters like intensity of integral length scale. Then, the mesh quality is poor, looks like each is a rectangular and big one. You might set the base size smaller, or add controlling volume for refinement.
Simce you are using Les, be careful with your inlet meshing and turbulence length scale. Your mesh should be fine enough to resolve the incoming turbulence.
- Badges
- Science topic