[gfoam]

Hi everybody:
I want to simulate an aircraft in a roll maneuver in FLUENT. The maneuver is stationary, namely, the forward and angular velocity are constant and the flow is incompressible. I think this could be done via the moving refrence frame formulation or via the sliding mesh approach. The second approach has a higher computational cost than the former but, which one is the correct? Or which one has the resilt more accurate? Thanks in advance
Gonzalo

[husker]

Hi,

Is the maneuver steady? I mean is the aircraft rolling steadily like a bullet disrupting the air? This sounds a little interesting to me.

If the problem is time dependent, you may need the use of dynamic mesh. At that time, I think only one steady boundary condition (cruise speed) will enable to simulate maneuver.

Regards

[Centurion2011]

Sliding mesh in usualy modelled like unsteady problem.

How could you model rolling maneuver as steady problem?

check this link, and start playing with unsteady problems...

Do you have manuals for sliding mesh problems? Be careful with interfaces!

[duri]

Originally Posted by gfoam

Hi everybody:
I want to simulate an aircraft in a roll maneuver in FLUENT. The maneuver is stationary, namely, the forward and angular velocity are constant and the flow is incompressible. I think this could be done via the moving refrence frame formulation or via the sliding mesh approach. The second approach has a higher computational cost than the former but, which one is the correct? Or which one has the resilt more accurate? Thanks in advance
Gonzalo

Both approaches are correct. With MRF you can avoid interface errors and faster. There is no need use costly sliding mesh as your problem can be well represented with frame motion.

[gfoam]

Originally Posted by husker

Hi,

Is the maneuver steady? I mean is the aircraft rolling steadily like a bullet disrupting the air? This sounds a little interesting to me.

If the problem is time dependent, you may need the use of dynamic mesh. At that time, I think only one steady boundary condition (cruise speed) will enable to simulate maneuver.

Regards

Yes, the maneuver is a steady one as the example of the bullet. And because the forward velocity an angular velocity are constant, I think the problem is a steady one.

Originally Posted by Centurion2011

Sliding mesh in usualy modelled like unsteady problem.

How could you model rolling maneuver as steady problem?

check this link, and start playing with unsteady problems...

Do you have manuals for sliding mesh problems? Be careful with interfaces!

As I said before the forward and angular velocity are constant, so I think the problem is a steady one, and I actualy want to avoid using the sliding mesh approach as far as possible.

Originally Posted by duri

Both approaches are correct. With MRF you can avoid interface errors and faster. There is no need use costly sliding mesh as your problem can be well represented with frame motion.

Do you have any paper or publication where it is explained with some detail?
Thank you all for your answers.

[husker]

Hi,

Please pardon me, but I'm still in doubt that you would like simulate steady roll action. Roll action is a time-dependent maneuver, lets say the aircraft completes its rolling in a 20 seconds. In this manner, transient solver will let you see what would happen in that 20 second.

On the other hand, you can also simulate the rolling steadily. In that approach, you can use MRF and as duri offered, less computational resources are required by MRF rather than dynamic mesh. Practically you don't need an academic paper on how to apply MRF, instead, you can have a look at MRF tutorials such as pumps and turbines. Fluent has 2D fan tutorial.

Regards

[gfoam]

OK, I will try MRF formulation and I will see what are the results. Thank you all for the responses