CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (https://www.cfd-online.com/Forums/openfoam-solving/)
-   -   Not sure which solver to use for an FSI case (https://www.cfd-online.com/Forums/openfoam-solving/202416-not-sure-solver-use-fsi-case.html)

i.sabahi May 30, 2018 09:30

Not sure which solver to use for an FSI case
 
1 Attachment(s)
Hello dear foamers

The problem i want to solve using OpenFOAM 3.0 or Foam-Extend 4.0 is analyzing a 2D, 2-dof airfoil section(pitching and plunging movements) in the wind tunnel. The airfoil is considered a rigid body. It doesn't yet differ to me if the fluid is laminar or turbulent. I intend to solve this problem using the FSI solvers included in either the Foam-Extend 3.0, or the Extend-bazaar toolkit found here.

A simple schematic of my problem, taken from Hodges's "Aeroelasticity class notes" book is attached to the post.

My questions:
1. Using which assumptions and geometry do i get the simplest version of this problem? I'm guessing a laminar fluid, a simple shape like a circle(cylinder) attached at the center to a normal string, moving only at y direction(vertically)?
2. Using which solvers, conditions and assumptions would I have the most validity and generate the best results without having to face extreme difficulties?
3. What solver, case and tutorial do i use first? I understand that to the extent of my knowledge, the following solvers and tutorial cases are present:
Foam-extend:
icoFsiElasticNonLinULSolidFoam(HronTurekFsi)
icoFsiFoam(flappingConsoleSmall)
Extend-bazaar:
fsiFoam(3dTube,beamInCrossFlow,HronTurekFsi3)
stressFoam(plateHole)
thermalStressFoam(flange)

More explanations:

My (simply stated) understanding of the problem's physics:
As the flow approaches the airfoil, the forces exerted on it from the fluid(air) makes the airfoil move vertically or rotate about the fixed point. These displacements generate forces in the torsional and normal springs, which in turn will displace the airfoil and hence, impact the flow again. This will likely cause oscillations to occur to the airfoil.

The final goal of my simulation:
The final goal is to analyze the flutter speed for different Reynolds numbers, spring properties, airfoil shapes, point placements and etc.

A little more advanced understandings and stuff:
This problem is a 2 way Fluid-Solid Interaction problem, and because the final goal is to analyze the flutter speed, large displacements will occur, which means big and frequent vortexes and disturbed flow, hence more complex fluid solvers need to be used, and implicit(tight) coupling will be required to make sure of the convergence of the results in each iteration. Also using moving and conforming mesh will be needed. OpenFOAM utilizes the partitioned FSI approach, which is beneficial to me, but since the airfoil is rigid, a very simple solver for the rigid solid is sufficient.( for which i developed and verified a very simple FORTRAN code, for which the input is: point forces of the polygon-assumed shape and it's properties, and the output is: final velocity and coordinates of each point after delta_t.)

Somethings i should add:
I understand only a little C++ code, and my deadline is in 2 months, so even thought I'm spending like 6 hours a day on this and I'm willing to study any material deemed necessary by the knowledgeable users, I really need advice on how to be able to run my simulation as fast as possible.

What I've done till now:
Ran some cases with OpenFOAM, read and done most parts of the OF tutorial, searched a LOT on this forum regarding similar cases. I've also ran the flappingConsoleSmall case with icoFsiFoam solver till endTime=1s, and viewed the results with Paraview.


Please keep in mind that my major is similar to mechanical engineering and even though I'm not a total stranger to fluids and CFD(Passed fluid mechs subjects 1&2, and a CFD subject in bachelor's), there's a lot i don't know.


I understand the forum members' value of time and would gladly and naturally provide feedback and info on my progression as to help future foamers and dear readers of the forum. Also, if this thread is posted on the wrong forum, my apologies, i read this guide and found this forum the most relevant.

Thanks to everyone in advance for taking the time to read and/or provide answers to this post.

i.sabahi June 2, 2018 06:43

Dear forum members

It would be greatly appreciated if you could take the time to answer my questions or guide me towards a source by reading which my questions would be answered.

Regards
Iman

i.sabahi June 17, 2018 08:13

Dear all


I've come up with the conclusion that the best solver for my case(more info in the post #1) is pimpleDyMFoam, and even though that'll not be a true FSI simulation, it'll be the closest to what I need.

In short, basically:
Concept of FSI, + if the body is considered rigid = concept of dynamic meshing.

Am I on the right track here? Any thoughts regarding this post or post #1 would be greatly appreciated.


Wishing everyone a good day
Iman


All times are GMT -4. The time now is 01:35.