Learning Forum

[Iwashita]

I conducted the simulation of unsteady aerodynamic force of flat plate.

The condition of analysis is following pitcture. U=2m/sec, B=1.0m

Turbulance model : LES Smagorinsky-Lilly (Cs=0.1)

[Iwashita]

At the same time, Moment is correct. Only phase of Lift force is not correct. Why?

[Iwashita]

I also calculated about rotational oscillation, and it was correct result both Lift and Moment.
I think something wrong with monitoring Lift force of parallel motion object. But I could not identify setting problem yet.
Or Dynamic mesh zone setting is wrong (Ex. center of gravity)

[Rob]

Where are the cells created to account for the motion?

[Iwashita]

ItÔÇÖs center (x,y)=(0,0)
Here is fluent settings.
Following figure describes the concept of unsteady aerodynamic force.
The theodorsen function says when U/Bf(Non-dimensional wind speed)=0, both LyR and LyI =0 , which means no lift force when U=0(m/sec).But actually, lift force work on oscillating flat plate in no wind by my analysis model.

[Rob]

Hex mesh so it's smoothing rather than remeshing. Can you alter the model to fix the mesh around the body and move that within an outer mesh?

[Iwashita]

Thank you for your comment. I tried it, but the result did not change. May be it takes long time to resolve this problem.

[Rob]

How big is the displacement, and how many time steps per cycle?

[Iwashita]

The amplitude of vibration is B/100 =1cm. And 500 steps/cycle(dt=0.005sec, U×dt=B/100). I tried more cases to know what causes this problem.

[Rob]

Looking at the mesh the jump in cell size won't help stability. Which zones are moving, and which are stationary?

[Iwashita]

Thank you very much for your help.
I repaired the mesh round plate.Dynamic mesh zones have been defined for the plate and the around the plate (both are specified as edge zones).
In addition, the inlets and outlets do not have dynamic meshes defined, or the "Exclude mesh motion in boundary conditions" checkbox is checked.
I compared next analysis. Right side result(0.00065N) is near to manual calculation value(0.000776N). But left side analysis is almost 100 times force of right side model.
As you say, there may be something wrong with the definition of dynamic mesh, otherwise transient analysis setting is wrong.
And here is dynamic mesh settings

And this is UDF (I think this is not problem)

[Rob]

I think the problem is with the way you've set up the moving parts. For detailed flow around an object we tend to move the object and a region around it. That whole fluid zone moves as a rigid body and we than remesh the bits around that. Unfortunately the tutorial and video in Help both show it as being an overset mesh problem, that works but MDM tends to be more efficient.

[Iwashita]

Is MDM the method I'm working with now? I have modified the around plate area to Surface, but the result did not change. I'm sorry I haven't been able to solve this problem.
(Excuse me. In the post of November 9, last case in Table 1 is U=2m/sec, not 0m/sec.)
I found similar discussion, but he may have given up.
/forum/discussion/26263/dynamic-mesh-drag-coefficient

[Iwashita]

I'm not sure you can open this file. This is analysis files.

[Rob]

That's better. The region around the plate should move with no remeshing. The outer region will then remesh. I think may know more - I know what the model does but not which buttons to use to make it work.

[Iwashita]

Thank you very much for introducing him to me. I have sent him a message.