I have been researching the air flow effects on frost formation at fin and tube heat exchanger and I need uniform flow media in front of heat exchanger. However, I haven't ensured at my tunnel illustrated attached image due the fact that it doesn't a contaction section including honeycomb or screen. I would like to thankful if you share documentation regarding contraction design at low-speed wind tunnel (my velocities are max. 5 m/s) and your recommendations. Also, Is there another method to form more uniform flow at entrance? Thanks and regards.
Hi Ergin,
The contraction follows the flow conditioner unit that houses honeycomb section, the screens etc. Honeycomb section first breaks the large whirl into smaller ones at its entrance and then makes the flow align along its length. The screens attenuate the turbulent fluctuations. Large-scale turbulent eddies are broken into smaller eddies and finally made to decay. Then follows the important contraction which accelerates and aligns the flow and feeds the test section with flow with reduced turbulence.
Useful References:
1. A very digested publication on wind tunnel testing can be found on Researchgate profile of Louis Cattafesta:
2. Barlow, Rae and Pope. Low-Speed Wind Tunnel Testing, John Wiley and Sons, NY 1999.
3. Bradshaw & Mehta. Wind Tunnel Design, http://www-htgl.stanford.edu/bradshaw/tunnel
Hope this helps.
Rajeev
I wrote a paper on the design of a wind-tunnel. It includes a method on design of contractions and settling chambers.
Best regards, Borja Rojo
Article Facility for investigating the flow in a low pressure turbin...
Hi, It is not so easy to answer without a schematic of you equipement, but I try to give you some indications...The length of establishment of the fluid flow (unfiorm velocities at entrance) depends onto the velocities and therefore of the flow regime given by the Reynolds number.calculated with an hydraulic diameter based on the entrance diameter of your tunnel. (are you in laminar or turbulent conditions at entrance?). Then, if you are in subsonic conditions (under the sonic velocity), the flow in only dependant of the generated pressure denoted Pi. Then, your flow conditions into your wind tunnel will be found through sonic tables (charts) according to your throat section. These table are established from classical compressible air equations as a function of the Mach number of your fluid flow. If not, in a first time,you can approximate the length of establisment of your fluid flow taking a establishment length equals 20 timesthe diamter of the duct..
Follow the British Standard. This what I did when I designed my wind tunnel in my MSc research.
Thank you very much for your helping, supporting and sharing. I have identified my contraction curve, but contraction ratio have to be about 1,5-2 due to fact that I don't have enough space. Normally, it has to be 5-8. Is it any problem? and I would like to remind that the air velocity is 4-5 m/s. Thanks
Many methods to calculate contraction wall contours have been available in technical literature. For low speed aerodynamic (and hydrodynamic) tunnels I use the method proposed by Cohen & Ritchie for calculation of small and large contractions with relative high area contraction ratio and short length .
(COHEN, M.J. & RITCHIE, N.J.B. Low-Speed Three-Dimensional Contraction Design. Journal of the Royal Aeronautical Society, Vol.66, pp.231-236, 1962.)
In my ResearchGate profile, the technical communication "Design and Construction of Small Axisymmetric Contractions Part A: Design" shows in detail the application of Cohen & Ritchie method.
A octagonal contraction with 16:1 area contraction ratio of a for a vertical hydrodynamic tunnel produces very stable flow with very small relative turbulence levels - see my papers REBUILDING AND TESTING A VERTICAL HYDRODYNAMIC TUNNEL and EXPERIMENTAL STUDY OF FLOW PAST AN ELLIPTIC CYLINDER NEAR A FLAT PLATE.
Dear Sir, I performed a study but my contraction ratio had to about 1,5-2 due to limitation of laboratory area. Normally, our wind tunnel air maldistribution ratio (standard deviation) before the revision was about %15 with respect to the measurement result taking from 300 points in front of heat exchanger. After the revision, it was seemed an improvement of air uniformity by about %3-4 (it would be %13 according to literature) and the temperature differences at each circuit outlets is less than 1 ˚C but it couldn’t be obtained this condition before the revision as you can see attached doc file. I would like to learn your opinions about my design and Does the usage of honeycomb ensure a improvement of uniformity? Thanks for helping.