Learning Forum

[info]

I am trying to evaluate the 'flutter derivatives' of a very simple 'bridge-like' section geometry, based on the works of ScanlaN & Tomko (1971).

For that purpose, I need to move the wall of the immersed body with periodic (either translational or rotating) motion and check the forces on the body at specific time instants. It is a Time-dependent simulation obviously.

I set up a simple case in AIM where I try to model the periodic vertical body motion, but it seems nothing happens, i.e the body doesn't seem to move. Please check the attached screens and animations, also the link for the archive.

Please help me to understand the possibilities and limitations of the moving wall function, especially in the sense of 'Time-dependent' Fluid simulation.

Thank you in advance.

Csaba

🛈 This post originally contained file attachments which have been removed in compliance with the updated Ansys Learning Forum Terms & Conditions

[Gaurav Sharma]

Hi steelxpert

For a stationary wall with no slip condition, the fluid particles adjacent to the wall have zero velocity. But if the wall is non stationary in practical application, it would impart some velocity to the neighboring particles. Moving wall function is to capture such scenarios.

Please note, the use of this feature doesnt' end up in a visibly moving wall, but if we zoom in on the faces where wall is defined,  we must see the particles to be moving with the specified velocity at any time point. This is the velocity that the wall imparted on these particles due to no slip condition.

Please let me know if you are having an observation that doesn't abide to above behavior. If required, I can share a securetransfer link to get your files and have a look at your setup.

Thanks & Regards,

Gaurav

[info]

hi Gaurav_ANSYS

Thank you for your response and explanation.

Since there is no instructional video for such important case, neitherthe Help provide examples of proper usage, the most important would be to clarify the problem set-up.

Please arrange the securetransfer link, so that I can be sure that we are on the same page for the model set-up for this type of problem.

In the meantime, can you be more specific and let me know which result type is capable to help visually identify the moving particles?

Other important aspect that need further clarification is the distinction between Steady and Time-dependent solution for this type of simulation.

Thank you in advance.

steelxpert

[Gaurav Sharma]

steelxpert I performed a dummy test and applied a velocity of 0.1 m/s on wall. After solution , I checked velocity contour on same wall surface and it was in agreement with provided inputs.   You can perform something similar to identify the velocity of fluid particles at the wall surface. A time dependent or steady solution depends on what is your application. For a continuously rotating wall surface, you might be interested in steady state but if the wall velocity changes in the duration, you may need to go for time dependent solution. With this explanation, i think there is no further value addition in transferring the model. Thanks & Regards, Gaurav

[info]

hi Gaurav_ANSYS

Thank you for the further explanation.

I was confused with the meaning of the 'moving wall' feature. I guess, I expected similar results to the 'moving domain' feature implemented in Fluent, where one can expect to see the how the flow changes due to bluff-body motion on e.g. a velocity magnitude contour plot.

Now it is clear, that this 'moving' wall feature of AIM is a very-very basic version of that Fluent's feature. I regret to aknowledge it.

Can you please briefly explain how this phenomena, i.e. the moving wall, is handked within the solver? I guess that the additional pressure due to the quasi-motion of the wall (1/2*rho*U^2) is added/subtracted to the Steady/Transient results beside the wall?...

As I started in my original question, my goal with this 'moving wall' feature is to somehow get realistic forces (i.e. pressure distribution) to my bluff-body, so that I can further assess the aeroelastic behaviour of my structure. When I approach this with a Transient run (with the 'moving wall' feature is 'ON') I need the instantaneous forces/moments exerted on the body. Since unfortunately the 'moment reaction' is not yet available in AIM for fluid, neither I can send each CFD step result to the structural solver in the implemented FSI, that's why I need a lot of post-processing (Excel, Matlab) to extract the Moment Reaction. For that a 'Line chart' representing the pressure forces would be a conveniant tool, but unfortunately the implementation for this within AIM doesn't make any sense at all for Transient runs, because still only the Line Chart of the last step can be exported!

And this will bring us again to my other question that is still unanswered from ANSYS's side: https://discoveryforum.ansys.com/t/y7a7sx/line-chart-result-discontinues-dashed

[Gaurav Sharma]

Hi steelxpert 

Sorry for the delay in my response. I rant into confusion with the use of term "Moving wall". While you used the term for physical representation f your requirement, I took it for the standard feature only.

Anyways, it seems your requirement needs the dynamic mesh/sliding mesh feature of Fluent which is not available in AIM. Unfortunately, the moving wall feature in AIM doesn't seem to cater to such complex requirements. 

Thanks & Regards,

Gaurav

[info]

Hi Gaurav_ANSYS

Thank you for your answer.

Can you please then briefly enlighten me (and possibly other users) about the theory behind (and its limitations) of the 'Moving wall' feature encapsulated in AIM?

Regards,

Csaba

[Gaurav Sharma]

steelxpert 

Not sure what you are looking for here. Moving wall is just a special case on ensuring no slip at wall, when the wall is moving and not stationary. This comes handy when the wall or adjacent solid body is not modeled in simulation.

The no slip condition states that there is no relative motion between the fluid and the solid at the interface. That indicates if the solid is moving with a certain velocity, adjacent fluid particles will also have the same velocity. When we are not modeling a complete FSI problem, this effect can still be incorporated using the moving wall feature. 

Thanks & Regards,

Gaurav

[info]

Gaurav_ANSYS

Thanks for the explanation again - I think I still don't understand for what practical problems I can use this AIM feature. It's probably because of my lack of sufficient theroretical knowledge, I don't know.

What I am looking for, is very simple, is that how i.e. the surface pressure distribution changes on an immersed body when the wall is moving compared when it is still. Is this feature capable of doing that?

In bridge aerodynamics there is an instability called 'flutter'. This instability can be handled by close-coupled FSI (it is clear that I need FLUENT for that), or either indirectly.

The indirect (i.e. not closely coupled) solution (let's think about a 2d section of a bridge deck) is when we distinctively apply 2-DOF movement-type harmonic excitations (displacement and rotation) to the immersed body and then we postprocess the resulting lift forces and moment reaction as the function of the excitations. From these force/moment functions we can calculate the 'flutter derivatives' which we can use to further implicitly solve the partial differential equation system of the complex motion, thus get a picture of the instabilities (critical wind speed).

In this sense can I use this AIM feature in a time-dependent type CFD?

Regards,

Csaba

[Gaurav Sharma]

steelxpert 

When we talk about flutter analysis, it generally takes into account the flexibility of solid body. It needs a 2 way FSI (or atleast dynamic mesh for some special cases) to model flutter thoroughly. 

However, only for cases, when the solid body is acting as Rigid, the moving wall option of AIM can be potentially used. As I mentioned earlier, the complete movement of rigid body is again not feasible but its effect at any particular instant on the flow can be captured using "moving wall feature of AIM.

Thanks & Regards,

Gaurav