The effect of turbulence on traditional wind mill sails is important. They provide extra forces on the sail rods. The question is how do we calculate (using Imcompressable) these turbulance induced forces: gustloading.
At this moment my thinking is:
- I can simulate the airspeed U and k at a certain point in an air flow (the point is where the sail tips would be: so I don’t simulate the sail).
- This provides an idea of the Turbulence Intensity: TI=sqrt(2*k/3)/U
- I take as a formula for the turbulance force gustloading: Fgust:: U^2(1+g*TI)^2
g is the number of sigmas (TIs) one incorporates to determine the gust force. I understand in general g=3.5
- if we remove the steady-state gust force due to the steady-state speed: Fgustnorm:: (1+g*TI)^2
I am not sure if my above thinking is correct. Any ideas/hints/links? THANKS
An additional question is: is the turbulence model (for instance Realizable k-epsilon) accurate enough to provide an idea of these forces?
All the best,
Hi @molen_victor,
Your thinking is on the right track for a preliminary analysis. Using a steady-state simulation to find the turbulence intensity (TI) and then feeding that into an empirical formula is a sound engineering approach to get a good first estimate of the gust loading.
However, it’s important to be aware of the limitations of this method. Since you’re using a steady-state (RANS) simulation, the turbulence intensity you’re calculating is a time-averaged statistical value. It doesn’t capture the dynamic, transient nature of real-world wind gusts. While your formula will give you a reasonable approximation, it won’t be as accurate as a more direct simulation of the gust event itself.
Turbulence Model Accuracy (Realizable k-epsilon)
You are correct that the Realizable k-epsilon model is a robust and widely-used turbulence model that is well-suited for many industrial applications, although k–\omega SST tends to be a more natural choice . For your semi-analytical approach, Realizable k-epsilon should provide a good estimation of the time-averaged turbulence quantities like k and epsilon.
The key thing to remember is that RANS models like Realizable k-epsilon are designed to model the effects of turbulence on the mean flow, not to resolve the turbulent eddies and gusts themselves. So, while it is accurate for what it’s designed to do, it has inherent limitations when it comes to predicting the peak forces from transient gusts.
Cheers
Igor
I had the issue that k-omega SST sometimes more unstable is in its soluation then (Realizable) k-epsilon. But what I now decided is to start using Realizable) k-epsilon; and when that is succesful see how k-omega SST works. If stable, than I choose the k-omega SST results.
I am interested in indicatins; not [yet] in as accurate as possible…
Thanks for your feedback, really helped.
All the best