Expert: Paul Soderman - 12/9/2011

Question
QUESTION: Dear Sir,

I am now trying to develop a wind tunnel. Could you please advise me the basis of assigning the specification and arrangement of screens/honeycombs in a settling chamber. Should both screens and honeycombs be used?

What is the best shape and size of wire-mesh screens to be applied? How many screens  are at least needed to improve velocity distribution?

Can a contraction cone be developed with a square-inlet(20 cm x 20 cm) to a rectangular-outlet (13 cm x 5 cm)? I am afraid if such spec. of contraction cone will affect the flow distribution.

Your answers are really much appreciated.

Best wishes
Yati

ANSWER: Yati - You haven't told me what kind of wind tunnel you are designing (open circuit, closed return), nor what type of research or testing is to be accomplished. That information is needed to estimate what turbulence and velocity distributions are to be expected, and what is required. If you don't need especially clean test-section flow you may not need any screens or honeycomb.  If the flow must be very clean, you may need several stages of flow treatment.  

You can find a discussion and guidelines for wind tunnel circuit design including contractions, screens, honeycomb designs, and diffusers in W. Rae and A. Pope:  Low-Speed Wind Tunnel Testing. Although the guidelines are pretty straightforward, the subject is too complex to get into here in detail.

Yes, a square inlet can feed into a rectangular test section. The challenge for contraction design is that there are adverse pressure gradients at the entrance and exit.  Flow separation can occur. In addition, at the corner of rectangular contractions the boundary layer can get large. I offer the advice for a contraction shape from my friend and former NASA wind tunnel designer Ken Mort:

'The shape should be an S shape with cubics at each end. The inflection point should be much closer (1/4-1/3 of the length) to the inlet than the exit.  I like the guides given in the following.

Rouse, Hunter and Hassan, M. M.: "Cavitation-Free Inlets and Contractions". Mechanical Engineering, Vol. 71, March 1949, pp 213-216.

The same guides are also given in NASA TN D-8243.'

If you can get the following references, you can see what others have done with honeycomb and screens:

R. Loehrke and H. Nagib: Experiments on Management of Free-Stream Turbulence, AGARD Report 598, Sept 1972.

R. Loehrke and H. Nagib:  Control of Free-Stream Turbulence by Means of Honeycombs:  A Balance Between Suppression and Generation. J. Fluids Engineering, 342-353, Sept 1976.

It would be good if you could remove the screens for cleaning because they tend to trap dust and change their performance.  If multiple screens are used they should be spaced 30 times the mesh size or 500 times the wire diameter, whichever is larger, according to Rae and Pope.
Paul

---------- FOLLOW-UP ----------

QUESTION: Dear Paul,

Thank you for your kind answer. A lots of useful information are obtained. Actually, I am quite new in this field. I am now studying the flow and pressure distribution in a curve diffuser. FYI, I am using a blow down tunnel instead of a suction tunnel. The curve diffuser is installed at the end of the tunnel.

My concern is to get a fully developed and uniform flow at the entrance of the diffuser. I was introduced a sufficient hydrodynamic entrance length, Lh>20Dh but the flow profile at the diffuser entrance was still not fully developed and uniform (see the attachment). I am using a rectangular duct of 5 cm x 13 cm cross-section. I am expecting the velocity profile is a parabolic (laminar) or flatter (turbulent), similar to the round duct. Is that expectation right? The velocity ranges 2 m/s-30 m/s.

As the flow is not fully developed and uniform yet, instead of flexible hose, I am planning to introduce a settling chamber with multiple screens and a contraction cone just after the blower. Blower supply duct dimension is 20 cm x 20 cm, thus gives contraction ratio to the test section of 6. I have got this idea after looking at an open-circuit wind tunnel system. As the open circuit wind tunnel uses a suction tunnel, I am not sure whether this kind of system can be suitably adopted to the blow down tunnel or not?

It is so much hope that you can advise me in this matter. Many thanks.

Best wishes
Yati

Answer
Yati - Yes, you can use a settling chamber with screens downstream of a blower. But the blower introduces swirl, so you may want to use honeycomb and screens. The diameter of the honeycomb  should be less than or equal to the settling chamber diameter divided by 150. The biggest challenge is to find a cheap source of honeycomb. People have used straws and milk cartons stacked up for low speed tunnels to keep the cost down. Locate the honeycomb upstream of the contraction at least a half diameter of the contraction if possible to minimize the wakes entering the duct, though that advice is usually for wind tunnels with test sections just downstream of the contraction so it may not be so important to you. Rae and Pope discuss combinations of screens and honeycomb.

Your Reynolds number is large enough for turbulent flow. But with a rectangular duct, the flow at the center will be different from the flow at the side walls because of the boundary layer growth in the corners. I am not sure what the profile would look like, though I assume it would be fairly two-dimensional in the center. A round or square duct would provide a more uniform flow, though I assume you would need a diffuser before the corner, which is another complication. It looks like you will just have to try a few things until you get what you want. Good luck.

Paul