Hi,

I need to do a validation study for my airfoil simuation in the initial phase to ensure that my domain size chosen is valid. It is a C-grid. For e.g. my 1st domain size is:

distance from inflow to airfoil : 7 distance from top/bottm to airfoil : 8 distance to far-field : 16

grid is 240x80. Nearest perpendicular distance from airfoil surface to grid is 0.01 units.

Supposed my 2nd domain is

distance from inflow to airfoil : 11 distance from top/bottm to airfoil : 12 distance to far-field : 24

It is about 1.5 times bigger than the 1st domain. So what should be the no. of grids and nearest perpendicular distance? Is it reasonable to simply multiply everything by 1.5

So new grid size is 320x120. 0.015 units?

Thanks in advance

The airfoil case has been studied by so many that if you search, you will find tons of them. The question that you have to ask yourself is what is the meaning of the far field boundary condition (physically and mathematically). When you get the answer to this question clear in your mind, then you will be able to judge whether you have the right dimensions to your new meshes or not. In real life situations, we face objects of all kinds of geometry and not just airfoils, and we face the same question, is the size of the mesh adequate and the results are valid ? Physics then mathematics will give you the answer. Good Luck

Thanks Ahmed, but the farfield boundary also depends on the numerical mtd as well as the boundary condition. Hence, for each numerical scheme, what one does is to increase it to see if there is any chances of change. If it is too small, there may be backflow. Besides the physics, I believe one must also do this form of validation.

Anyway, what I meant is in order to have a valid comparison on the domain, how should I change it so that the comparison is still valid. That is, if I double the domain size, must I double the no. of grids? More imptly, should I also double the the perpendicular distance from airfoil surface to nearest grid?

Far Field BC or FreeStream BC or whatever the name given by the software you are using, is a distance far away from the airfoil (in all directions) where the disturbances caused by the presence of the airfoil in the main stream have died out. So what the magnitudes of the flow variables are supposed to be just one cell inside the domain (or their rates of change). The ideal case would be when the rates of change are all equal to Zero, then here comes your judgment, does the hardware you are using, have enough memory to extend the domain so far. Experimental work published so far suggests 10 chord lengths (or any other number) may be enough. You need to judge this based on your particular flow conditions. As long as you are happy with the rates of change calculated at this outer boundary. Good Luck

Just one question, when designing air crafts, do you really believe the vorticity generated by the wings have died out at the tail?

Thanks once again Ahmed.

I don't think so. At the tail, there should definitely be vorticity (shedding of vortex at the tail). I believe the vorticity will diminish as distance increase. That is why we have to extend the boundary to certain distances.

That is why we have to extend the boundary to certain distances.

Hundreds or thousands of meters?

Have a look at the ``No'' boundary condition which is likely what you want at the outflow boundary.


@article{papanastasiou1992nob,
title={{A new outflow boundary condition}},
author={Papanastasiou, T.C. and Malamataris, N. and Ellwood, K.R.J.},
journal={International Journal for Numerical Methods in Fluids},
volume={14},
number={5},
pages={587--608},
year={1992},
publisher={John Wiley \& Sons, Ltd}
}

Jed Thanks for the link, an interesting paper indeed. Nice fellow you are

You may be very pleasantly surprised to see what 'shape' the vorticity field takes further downstream - at say L/D = 20 - 40 range (L=downstream distance).

mw...

<www.adthermtech.com/wordpress3>