Learning Forum

[musaero19]

Hello everyone,

I am studying an oscillating airfoil and I have started using a UDF;

#define pi 3.1415926535897932384626433832795

#define ff 0.87898

DEFINE_CG_MOTION(move_body, dt, vel,omega, time, dtime)

{ 

real alphabar,ome;

/* reset velocities */

NV_S(vel,=,0.0);

NV_S(omega,=,0.0);

/* motion */

    alphabar=20*pi/180; /* amplitude, pi/180 used for have rad instead of deg */

    ome=alphabar * ff*2*pi*cos(ff*2*pi*time);

 vel[0] = 0.0;

 vel[1] = 0.0;

 vel[2] = 0.0;

 omega[0] = 0.0;

 omega[1] = 0.0;

 omega[2] = ome;

}

I am using airfoil as a rigid body and FLUID as a rigid body. so the whole domain is rotating. Is this the correct approach, since there is no dynamic mesh motion?

Thank you.

[Rob]

I'd probably use sliding mesh and recode the above to rotate the sliding zone surrounding the aerofoil. As to whether the model is correct, it's not wrong but I don't know if it's doing what you want it to as opposed to what you're telling it to do.

[musaero19]

Hi Rob, thank you for your answer. Yes, there are 3 different approaches as I read from the literature, moving whole domain, hybrid domain (inner and outer with interface), and dynamic mesh approach. I thought the first approach I am using might have some cons. I am studying the dynamic stall of airfoils at high RE. I am going for a hybrid approach now. Thank you.

[Rob]

That's how I'd do it. :)

[musaero19]

Thank you Rob for your valuable insights.
One question, if I want to oscillate a 3D wing using the same hybrid domain, inside for 3d wing and outside domain..using the same UDF. Does it work for 3D geometry? Or do we have to go for a different approach?

Thank you in advance.

[Rob]

The UDF you gave is for 3d. C (C++ etc) define x=0, y=1 and z=2 for the coordinates: so vel[1] above is the y component.