CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   Flapping wingsfishlike propulsion (http://www.cfd-online.com/Forums/openfoam-solving/60216-flapping-wingsfishlike-propulsion.html)

joegi May 9, 2006 13:41

Hi everybody, Probably this
 
Hi everybody,

Probably this topic has been dicussed before, but I'm wondering if somebody else is using OpenFOAM to tackle this king of problem, that is : flapping wings/fishlike propulsion.

I just installed OpenFOAM, and now I will start to learn everything, but if somebody can share his experience on this topic with me/everybody I/we will really appreciate it.

For the moment we want to compare the results obtained with an in-house software developed in the uni (this software uses the immersed boundary method), with another high front software, and probably try to implement the same meshing approach used on the IBM in openfoam (if it's not yet implemented).

Thanks in advance for your help,

joegi

hjasak May 9, 2006 13:52

Sounds like a fun project. A
 
Sounds like a fun project. A combination of the automatic mesh motion solver and a laminar/turbulent transient flow solver will do the trick. Do you have any plans on how to specify the boundary motion to get the wings flapping properly?

I have been involved in some fluid-structure interaction projects in the past and several of them are ongoing. For example, have a look at the stuff from prof. Ivankovic's group at university College Dublin (there are some presentation slides from the OpenFOAM Workshop in Zagreb):

http://powerlab.fsb.hr/ped/kturbo/FsbOpenFOAMWorkshop/

Good luck,

Hrv

lr103476 May 10, 2006 03:50

Hi, Here in the Netherlands
 
Hi,

Here in the Netherlands, I am studying the aerodynamics around insect based flapping wings. Up to now my model only consists of a 2-dimensional airfoil which moves according to harmonic functions. The angle of attack and amplitude are simplified to sine functions. Basically the range in angle of attack is about from 90 deg up to -90 deg. The amplitude is about 4 chord lengths.

At this time I am still validating / verifying the unsteady icoFoam solver using static cylinders. When this validation / verification study is finished, I will proceed with the flapping wing model. It is of upmost importance to get your computational domain and grid right. The automatic mesh algorithm in OpenFOAM leads to very accurate meshes, but is not that fast (compared to the spring bases method used in Fluent). Therefore it is advisable to divide your comp. domain into several submeshes. Only near the wing automatic mesh deformation has to be used to minimize the work for this step.

From my experience, no commercial software package is capable of solving this kind of problems sufficiently accurate and efficient. Fluent is only first order accurate in time, and only capable of solving moving tetrahedral meshes. In addition Fluent re-meshes parts of the domain, which may lead to severe mesh quality degradation, especially after multiple flapping periods. The method used by OpenFOAM leads to nice meshes at all times but is slower. On the order hand the icoFOAM solver is twice as fast as the flow solver in Fluent.

I doubt that Immersed Boundary Methods (IBMs) are suitable (in terms of accuracy and efficiency) for this kind of low Reynolds number problems. Since the flow is very viscous (due to Re = O(100-200)) large boundary layers occur on the wing surface which needs to be captured by sufficient (nice) cells in the mesh. Using IBM a lot of interpolation needs to be performed near the surface such that this method is not preferable to solve for moving wings at low Re.

I suggest to try the automatic mesh solver of OpenFOAM. When you setup you problem axi-symmetric you could maybe also use some topological changes to account for rotation. Although I doubt the capabilities op IBMs I am very interested in its behavior since I have no real practical experience with it. So when you write your IBM in OpenFOAM for 3D flapping wings I am very eager to give it a try and see what happens.

When I got some usable results from my moving wing simulations I will post them. After that I want to extend my model to a 3D insect wing wich moves according to real insect kinematics. In my oppinion OpenFOAM has the capabilities to solve for that sufficiently accurate and efficient using its moving mesh algorithm.

Please, keep me informed about your progress.

Regards, Frank

joegi May 14, 2006 14:53

Hi Frank, Do you have expe
 
Hi Frank,

Do you have experimentals results?.
Right now I'm looking for experimentals results on flapping wings, just to compare with the results obtained with the IBM software we are using.

By the way I'm NOT looking for experimentals results on cylinders, I think I have all the possible results on cylinders.

For the moment I'm learning the insight of openfoam, in order to attack my problem. Please keep us update about your progress.

Regarding to the commercial software issue, I'm using cfdesign just to do some validations, it's not as strong as fluent, but I think for this type of case is really good. It's quite easy to setup a 6DOF problems on it, you should give it a look.

Cheers,

joel

aeropeng April 28, 2010 23:49

Hi, joegi
I meet a problem that how can I describe the rigid body with 3DOF rotation by inertial coordinates. I'm pretty eager to know how did you setup a 6DOF of flapping wing.

Good luck
Peng

sanjay December 23, 2011 06:13

do u know how to do flapping simulation on Starccm+ ?


All times are GMT -4. The time now is 17:46.