Expert: Paul Soderman - 7/28/2006

Question
Paul,

I was linked here through google while I was trying to find some info about wind tunnel design.  I am an entering junior at Embry Riddle Aeronautical University studying Aerospace Engineering, and I am very interested in undetaking my own little project...building a wind tunnel.  I've been heavily researching the subject for the past week or so, and I believe I have a reasonable subject knowledge of the wind tunnel, its parts and their functions.

My question to you is, from what most of what I read about the contraction section said, the contraction cone should be somewhat "S" shaped on all four sides to maintain a steady, laminar flow, and reduce flow separation.  Knowing that construction of this section can be fairly expensive and difficult, I was wondering what implications can be had by making the contraction cone reduce its area linearly by using flat plates and posibly adding a screen at the end of the contraction to try and renormalize the flow of air instead?  Would the flow be too un-uniform through the test section?  Are there any suggestions you may have for me?

Note:  I plan on using a 1' x 1' test section (thinking about reducing the size) and using about a 1.5'-2' industrial size fan (already own).

Answer
Hi Josh

The good news is that the flow will accelerate in your contraction.  The bad news is that the test section flow will be pretty non-uniform.  As the flow accelerates along the contraction wall, the boundary layer will grow and retard the flow causing the velocities in the duct center to increase.  The flow in the corners will be especially bad as the two boundary layers mix.  You could try to predict the boundary layer size and resulting velocity distribution, which will depend on the length of contraction and wall angle.  Once the flow gets to the end of the contraction, it will have to pass over the shoulder - flow separation is real a possibility and would be especially bad for the test section. A screen at the test section inlet would be in a high speed flow and, therefore, quite lossy and would generate its own turbulence.  It would be difficult to make the flow uniform.

Whether  the non-uniform flow is acceptable depends on the anticipated use of the wind tunnel.  Most aerodynamicists want uniform velocities so they can test wings and expect nice loading.  If you only need the center of the test section, you might accept less non-uniformity as long as it was repeatable and calibrated.

My NASA friend, Ken Mort, added the following:

*****************************************
It wasn't clear if this was going to be a closed circuit or nonreturn
wind tunnel. If it's a nonreturn wind tunnel he will need a bellmouth
entrance. If it's a closed circuit he will need a contraction that is
sort of S shaped. The last  3/4 or so of the contraction should be a
cubic; otherwise the velocity distribution won't be uniform. A screen
at the small end of the contraction isn't a good idea; it will have a
high loss and introduce a lot of turbulence. If he wants to use a
screen at all it should be upstream of the contraction at least a 1/4
diameter ahead of the contraction.

As a compromise he could use straight sides for the first part of the
contraction and then make the last 1/2 cubics. For the start of the
contraction if it's a nonreturn tunnel he could use a simple lip with
half rounds. It still would be complicated to make.

Just using large radius sections upstream of the test section would
result in nonuniform flow in the test section.
**************************

Hope this helps.