Hi, I am looking forward to implement an open boundary (far field boundary condition) in an unstructured code which does not account for these situations. I am looking for some information )papers or simmilar) which could help me on this task.

Does anybody know any references concerning the subject???

Boundary conditions depend upon lots of things; please be specific so that any suggestion would be meaningful.

However, some of the papers ( I had posted a reference list sometime ago ) will give you far field boundary conditions for the solution of compressible inviscid/viscous flow problems on unstructured grids. Jawahar

Sorry I didn't make myself clear. I am trying to implement an open boundary condition in an unstructured code. My aim is to study the air flow around a standard NACA airfoil and I want to do it, not in a wind tunnel, like a closed boundary will give me, but on an open air boundary. I am aiming to reduce considerably the size of my domain. At the moment, I need to simulate it under a domain of 50!!! times the chord of the arfoil so the closed boundary effect can be minimized. Any suggestion will be welcomed. Thankjs very much, Jawahar

I would use absorbing boundary conditions (Riemann based). The standard reference is a Jameson paper... I haven't got it available here but do a search on "Jameson" in this forum (or the forum archive) and I think you'll find it - I have a vague memory that this has been discussed here previously.

The extent of the computational domain should be at least 10 times the chord, preferably 20, even with this kind of boundary conditions. With an unstructured code this shouldn't be a problem though since you can expand the cells further away from the object.

If your aim is to reduce the size of the computational domain the paper "Global far-field computational boundary conditions for C and O grid topologies" by A.Verhoff in AIAA Journal, Vol.36, No.2, pp. 148-156, may be useful for you. Whether structured or unstructured the concepts are same.

Hope this helps.

Uma

Hi,

I did a serch with Jameson....and nothing other than a post from a discussion on Density time marching based methods vs pressure based methods......so it must have expired!

Can you find any more details on this Jameson paper??

Thanks............................................ .Duane

Try the following far-field boundary conditions. I find it very effective for both structured and unstructured grid computations. You may locate the outer boundary between 8-12 chords. Again, you can try a couple of numerical experiments to fix the outer boundary ! I've assumed you are using circle as the outer boundary !

Create an extra ring of cells beyond the last ring of cells.

For free-stream Mach number less than 1: ---------------------------------------

Prescribe asymptotic far-field solutions in those cells. i.e uniform flow on which is superimposed the circulatory flow caused by a point vortex at the mid-point of the airfoil's chord. The instantaneous circulation GAMMA at the airfoil decides the strength of the vortex. The flow induced by the vortex is then computed from small perturbation theory after including the compressibility effects (Prandtl-Glauert transformation). (Also refer to " Foundations of Aerodynamics " by Kuethe & Chow)

For supersonic free-stream Mach numbers, just prescribe the free-stream values in the outer ring cells.

Everything I've stated applies to cell-vertex methods also. You can prescribe these conditions at points lying on the outer circle.

Good Luck ! Jawahar

Thanks very much indeed for your kindness and interest. Your advices have been extremly useful. XAVI

To complement Jawahar's posting, I think you would find the following paper very useful :

J.L. Thomas and M.D. Salas "Far-Field Boundary Conditions for Transonic Lifting Solutions to the Euler Equations", AIAA Journal, July 1996, Vol. 24, No. 7, pp 1074-1080.