You are on page 1of 7

ISSN (Print) : 0974-6846

Indian Journal of Science and Technology, Vol 9(30), DOI: 10.17485/ijst/2016/v9i30/99195, August 2016 ISSN (Online) : 0974-5645

Design of Low Cost Open Circuit Wind Tunnel
– A Case Study
Mahesh K. Panda* and Amiya K. Samanta
Department of Civil Engineering, National Institute of Technology Durgapur - 713209, West Bengal, India;
mahesh26119@gmail.com, aksnitd@gmail.com

Abstract
Objectives: This paper elucidate the entire procedure to design an open circuit subsonic wind tunnel which will be used
to study the wind effect on the different prototypes of structural elements. Method: The existing guidelines and findings
of the previous research works were followed for design calculation of different parts of the wind tunnel. Three design
trials have been performed so as to achieve a velocity of 25 m/s at the test section. The final trial has been discussed in
the body of the paper. Findings: The design includes a square test section of side 500 mm, to accommodate the model
and required instrumentation in it, for force and pressure measurement. A straight section before the test chamber is
provided to allow the output of contraction section to stabilize before it reaches the test section. Instead of using a curved
wall shape a straight contraction profile with larger contraction length is used. The diffuser extends from the test section
and its cross section changes from a square test section to an octagon of 311 mm side. The pressure drop and power
required for the designed wind tunnel are calculated and tunnel performance curve is plotted. Application: The design
being purely empirical, it requires computational validation before commencement of the construction of the wind tunnel.
The investigators have actually made an attempt to design the same in the existing setup, which may be useful for different
research purpose.

Keywords: Design of Open Circuit Wind Tunnel, Loss in Contraction Section, Power Requirement, Pressure Drop in
Diffuser, Stilting Section, Tunnel Performance Curve

1. Introduction volume of air is allowed to pass through the tunnel circuit
continuously Figure 1.
The wind tunnel is a tool to study fluid flows around a
body and the forces generated by the fluid-structure
interaction. Using such tool, it is possible to measure
global and local flow velocities, as well as pressure and
temperature around the body. A wind tunnel experiment
provides the force and pressure values on the model
and the flow visualization. It is a critical instrument in
the quick and thorough design process of anything that
involves fluid dynamics. Based on the flow through the
tunnel circuit, the wind tunnel is of two types, open
type and closed type. In open type, ambient air enters Figure 1. Schematic diagram showing different parts of
from one side and exits to the atmosphere after flowing wind tunnel.
through the tunnel. In closed type wind tunnel, a constant

* Author for correspondence

R being the hydraulic radius of the contraction section test section. The pressure 0. A contraction wind tunnel to have minimum construction cost and the design satisfying all criteria will be such that the separation running cost. Test chamber dimension depends on the type of test to be performed and the size of the model to be tested. So the length of test section should be as 0. The test 0. Design Methodology starts with the selection of contraction ratio6.4. The length of the test chamber was set 0. Figure 3. satisfying the test requirements. 2. which section keeping an eye on the accessibility and installation is 20% longer than the maximum recommended value.03 loss coefficient goes on increasing with increase in the test section Figure 3. The open circuit wind tunnel is made of several distinct The length of the contraction should be in between 0.e. the to R. From the easier and cheaper construction a straight contraction shape experimental investigations. of the test chamber. For smaller tunnels the contraction ratio should be in between 6 to 9. the pressure drop coefficient values go on decreasing with increase in CR.02 small as possible. The contraction 2. To avoid flow The wind tunnel design starts with deciding the test separation.indjst. The nozzle exit cross section dimensions and shape when it is about 10%2. described as follows: As shown in Figure 4. is just avoided and the exit non-uniformity is equal to or less than the maximum tolerable level for the desired application in shortest possible length5. For of the test model and instrumentation. Design of Low Cost Open Circuit Wind Tunnel – A Case Study The main goal of wind tunnel design is to have uniform flow within the test chamber.org Indian Journal of Science and Technology . Pressure drop coefficent In order to have a blockage ratio less than 10% a square 0.5-3 times its hydraulic diameter3.01 250 500 750 1000 1250 1500 1750 to 1. diffuser. The wind tunnel’s overall dimensions are key factors in its construction and running costs. Principal parts of an open type wind tunnel are the test 2.215 times the hydraulic diameter of the Length of test section (mm) test section. the diffuser and the fan. The test chamber is being designed to test the scaled model 0. in the range of 0. space and budget constraints1. The design process starts by defining the test chamber dimensions and shape. settling chamber and driving The contraction accelerates the flow to the test section. so the maximum 2.15 R sections. Different parts are inlet. unit (Fan).05 test chamber of 500 mm side is chosen Figure 2. Too large contraction section avoids flow separation7.06 of a silo with length 250 mm and minimum L/D ratio 1. The test chamber length has to be are identical to the test chamber.04 will be carried out at a flow speed of 25 m/s. the settling chamber. a contraction of length of 900 mm is used. contraction. This paper deals with the design of an open type further reducing any variations in velocity. 2 Vol 9 (30) | August 2016 | www.25 m i. after that the rest part of the wind tunnel is designed by keeping in view the dimensions Figure 2.1 Test Section allowable contraction ratio 9 is chosen. the contraction cone. it has been found that the is used instead of solving complex equations to get the wall blockage has an almost negligible effect on test results shape.2 Contraction Cone section. Plot to show the pressure loss in a constant area section. The trade-off is required between the opposing needs to get a suitable design of the wind tunnel. Test section.

0 7. A selection of honeycomb and screens for a wind tunnel is very much dependent on the test type to which the tunnel is intended8.org Indian Journal of Science and Technology 3 . diameter (L/D) and porosity (flow area upon total area). A diagram showing the contraction cone with a chamber cross-sectional area matches the dimensions of straight section. which falls in the no stall region in Figure 7 with Contraction Ratio a diffusion angle 5.e.0 as 1.0036 m length is provided.6 Honeycomb drive system. 50 and 0.0040 solving for L. Contraction section pressure loss coefficient with contraction ratio. Figure 6.0038 the diffuser is found to be 2.0 9. 3 mm Vol 9 (30) | August 2016 | www.5 m has been inserted after the contraction.0044 Assuming minimum diffusion angle (θe) i.25 m is used. The area ratio of the withstand applied forces during operation without diffuser should be less than 2.0046 cone.5 and diffuser angle should significant deformation.0042 0. Dh1 is the inlet sections hydraulic diameter 0. The diffuser is mainly used to reduce the velocity of flow in the shortest possible distance to reduce the load on 2. The minimum for honeycomb is the ratio of length to cell hydraulic length of diffuser can be found from Equation. contraction cone inlet i.5 m with a length 2. Figure 4. for non- streamlined shape of contraction wall8.0 6. The diffuser area ratio is calculated 0. hence a diffuser of 2.0 8. .5 7. wall contour and velocity9.38 m. It should have sufficient flexural rigidity to diffuser cross-sectional shapes.5 m x 1. A diagram showing the diffuser. Loss coefficient 0. A square shaped honeycomb. A straight section of length 0.0048 Loss coefficient and is the half of the included angle of the diffuser 0.4 Diffuser equal to 1. The primary design parameter be 5˚ – 7˚ for controlling flow separation. 2.e.5 0. Panda and Amiya K. Samanta Where. Mahesh K. 2.5 6.3 Straight Section Inserting a small settling duct before the test section can reduce the turbulence level to an acceptable level.5 10.0032 side 310 mm Figure 6.5 8. The flow through the diffuser depends on its The honeycomb removes the swirl from the incoming flow geometry defined by area ratio (ratio of outlet area to the and minimizes the variation in both mean and fluctuating inlet area) and diffuser angle (2θ).0034 0. The settling Figure 5.8. 1. the maximum of the minimum length of 0.5 9.0030 5.5 11. (1) Recommended L/D ratio for honeycomb is 6 to 8 with porosity nearly 0.86. Figure 5 shows the straight section attached to the contraction cone.0 10.030.indjst. with an area of cross section equal to the testing chamber size. An octagonal outlet is chosen with 0.5 Settling Chamber The aim of a settling chamber containing honeycombs and screens is to reduce the flow turbulence before it enters the cone.

For an air tunnel. To be effective of proportionality is the friction factor (f). the placed at the same distance from the first screen. The main parameters are at a distance 340 mm from the start of the contraction i.2 times factor in 4 to 6 iterations with an initial value of 1. In our case.8 Drive System given as (7) The drive system compensates the loss in the circuit and determines the movement of the fluid through the test and the expansion loss is calculated by the empirical section.0966 4 Vol 9 (30) | August 2016 | www. A clearance of 0. breaks up the large scale turbulent eddies into a number of hydraulic diameter (Dh).2 times the hydraulic diameter of the settling chamber between inlet and outlet cross section areas (Ar).58 lead to flow instability10. The loss in reducing turbulence a screen must have porosity in the coefficient is given by. (4) range 0. maker’s performance chart can be consulted to choose a so = A3+B3 . . an Axial fan will be used as these are high efficient and produce air flow with The term can be expressed as a geometrical less turbulence. The and the 3 mm wire screen mesh with porosity 0. this loss (p) along the duct is proportional to its length (L).82. In our case > 50. Losses in Wind Tunnel Circuit a porosity value of 0. In the settling chamber diameter is required between the Equation is given as: screens. The energy loss of each section can be written in dimensionless form as pressure drop coefficient (2) Where.7 Screens 3.76 is loss coefficient is the sum of the two loss factors. in passing through the particular section. For loss calculation. the frictional loss coefficient is 2. RPM and required power14. is test section dynamic pressure given as (3) The losses in different parts of wind tunnel can be Figure 8.69 is placed loss and the expansion loss.8 are not The friction loss coefficient can be determined by suitable for good turbulence control while values below using universal law of friction to determine the friction 0. Screen porosity values over 0. for circular section the values of A3 and B3 are -0.1 Pressure Losses in Constant Area Sections Tensioned screens are placed in the settling chamber for Considering a constant-area section (A).e. Pressure loss in the diffuser is due to the skin friction So a screen with 2 mm wire and porosity 0. The pressure loss in a tunnel section is defined as the mean loss of total pressure sustained by the stream. first relates to friction and the second to expansion (6) For one dimensional flow. (9) fan with optimum efficiency. the tunnel can be divided into sections. Figure 8 shows an arrangement of Honeycomb and screens in settling chamber. equivalent conical expansion angle ( ) and the ratio 0. Design of Low Cost Open Circuit Wind Tunnel – A Case Study thick and 27 mm size and 150 mm length is provided with 3.indjst. Diagram showing the arrangement of honeycomb calculated as below: and screen. 2. After the maximum required fan static function depending upon the expansion angle and the pressure and volume flow rate have been estimated the cross section shape of the diffuser.58–0. the pressure the reduction of turbulence level of incoming flow.89. two primary drive systems are relationship (8) a compressor and fan. Experimentally it is found that better turbulence (5) control can be achieved when the finest screen is placed 3. fluid density (ρ) and constant small scale eddies that subsequently decay.org Indian Journal of Science and Technology .2 Pressure Loss in Diffuser farthest downstream11–13.

3553 + 0.04672 respectively. and f is loss in kinetic energy. for a speed of 25m/s at the test section. These losses are due to kinetic Where.36 Pa (13) Vol 9 (30) | August 2016 | www.e. Di the inlet cone diameter. Figure is the mesh factor equal to 1. Mahesh K.5*0. (15) The loss in the honeycomb can be calculated from the The required power for a given section size and flow formula conditions depends on the sum of the pressure drop (11) coefficient (ki) in the individual tunnel sections.3 for average circular metal wire. D0 the test energy being dissipated by vortices and turbulence.05866 respectively. L0 is the length of contraction.2 Stilting Section Loss the equation balancing the energy input to the stream to The losses in the settling chamber are primarily due to the the energy losses in the tunnel is honeycomb and screens used for turbulence reduction.144 =641. The section diameter.5*1. Thus if the power input to the fan is P (i.1 fir silk thread ratio for different diffusion angle. pressure must be compensated by a pressure rise. and kRn is given as (14) From the Equations it is clear that the pressure drop through the honeycomb and the screen depends only upon the characteristics of the same and becomes constant when the Reynolds number values exceed some critical value. The pressure loss equation for subsonic flow in wind tunnel is given as (16) Where In the designed tunnel the various loss coefficient are (12) given in Table 1. which appears as a decrease in total friction coefficient. 4. The loss coefficient in the contraction is given by15: Power required for maintaining a steady flow through the wind tunnel is equal to the total losses occurring in the (10) flow through the tunnel. of A3 and B3 are -0.indjst.5)/ (0. For square section the values Is screen solidity and is the screen porosity.142*.75) = 14. Honeycomb loss depends on L/D ratio of honeycomb and porosity while the pressure drop through the screen is dependent on the type of wire and the porosity.0132 and 0. and 2.225*252 *1.org Indian Journal of Science and Technology 5 .75*.5*1. Panda and Amiya K.225*14.  ower required in Wind P 3. the motor shaft output) and the fan has an efficiency 3.144 m/s The pressure loss in the tunnel circuit is equal to: =( ) ∑ ki + loss at fan section (17) The screen loss coefficient can be calculated = 0. Samanta and 0. Plot showing the pressure drop for various diffusion angle.144*14. usually provided by a fan.2 Contraction Section Loss Tunnel The pressure loss in contraction is due to skin friction. Figure 9. Speed at fan section: = (25*0.0 for new metal wire 9 shows the variation of pressure loss coefficient with area .25*3. 1.

Roy S. Design rules for small low speed wind tunnels. Hong YT. 6 Vol 9 (30) | August 2016 | www. manufactures. To ensure good performance of the structures subjected The power required by the tunnel is equal to: to wind loads.372 found then the wind tunnel can be built. 2015 Mar. So it is a useful tool in wind engineering. 83(827):443–9. Barlow JB.2200 the design is required to be verified computationally as Screen 1 0.6049 tested for the desired flow quality and pressure loss in the Contraction 0. 1999. Honeycomb 0. IOSR Journal of Engineering. The Aeronautical Journal. Mehta RD. as per the guidelines of the previous researchers. 1979 Nov. Model studies through = (0. Encyclopedia of Aerospace Engineering.indjst. Fundamentals of wind‐ performance curves.4117 well as experimentally. Experimental The fan/motor to be fitted in the tunnel should be chosen investigation on turbulence reduction in subsonic wind conservatively to have desired flow velocity in the wind tunnels. Figure 10 shows the tunnel performance curve for the designed wind tunnel. can be plotted on the tunnel performance 2010 Dec. Cattafesta L. Avail- able from: http://www. p. Ain Shams Engineering Journal.org Indian Journal of Science and Technology . 2. 6. 89(3-4):247–62. velocity 25 m/s Further due to the purely empirical design of wind Wind Tunnel Section Loss coefficient tunnel.752*14. 2012 Oct. Pope A. Figure 10. First the tunnel design is to be Screen 2 0. Gharbanian K. 2008.149 software (ABAQUS or ANSYS) if satisfactory results are Test Section 0. Bradshaw P. the wind tunnel performance is to be Total 1. 2011 Mar.3553 tested and calibrated. Chen JC. Rae WH. Low-speed wind tunnel test- ing. Journal of Wind Engineering and Industrial Aero- dynamics. Energy Conversion and Management.pdf 9. The power required for given flow in the wind tunnel test section is calculated and the same is plotted as the 6. Before carrying Diffuser 0. Design op- timization of three dimensional geometry of wind tunnel contraction. 2014.144)*641. their behavior must be anticipated in = Q * ∆Pa = (A*V)* ∆Pa (18) advance by the design engineer16. 15(2):137–47. Design of Low Cost Open Circuit Wind Tunnel – A Case Study Table 1.0037 tunnel computationally through any high-end simulation Straight section 0. Yassen YEl-S.tomcarlone. Arifuzzaman. Tunnel performance.36 = 4008. Tunnel pressure loss coefficient at test section tunnel test section to cater the unseen losses in the tunnel. Wind Tunnel Design. 3rd Ed. 6(1):281–8. Soltani MR. Saha UK. Conclusion ect/WindTunnelReport. 8. the tunnel.142*0. Elssaka MM. MM. Fang FM. which are supplied by the fan tunnel design. Design construction and performance test of a low cost subsonic wind tunnel. 4. Bahr CJ. 2(10):83–92.5 Watt wind tunnel facilitate the same for the intended structure. Experimental and analytical evaluation of flow in a square to square wind tunnel con- traction. Mathew J. United States of America: John Willey and Sons Publ.com/Wind TunnelProj- 5. 7. Fan 1. 5. 3. 86(6):418–27.0629 out any experiment. Building a wind tunnel: It will blow your mind. Aerospace Science and Technology. 2011 Mar. An adopted blockage factor correlation curve and the fan with maximum efficiency that intersects approach in wind tunnel experiments of a Savonius-style the tunnel performance curve is to be chosen for use in wind turbine. Abdalhamed AS.25*3. References tunnel performance curve. Manshadi D. 1–10.

Indian Journal of investigation on turbulence reduction in subsonic wind Science and Technology. Floren. 2011 Sep. 2011 Mar. Swami BLP. 197:139–55. Samanta 10. Turbulence reduction by screen. Turbulent boundary layer perturbed by a screen. 15(2):137–47. New York: McGraw Hills Publications. 2010 May. Wind tunnel at LABINTHAP (Updated). Velazquez MT. tunnels. 16. Wind effects on tall building 13. 3(4):565–73. Soltani MR. 15. 1998. Experimental frames-influence of dynamic parameters. Groth J. 1988 May. Mehta RD. 1985 Sep. Vol 9 (30) | August 2016 | www. Quiterio PR. Eslava GT. 23(9):1335–42. D Space. 14. 3(5):583–7. Francis JA. AIAA Journal. Panda and Amiya K. Fan handbook: Selection application and design. 2016. wiki/DSpace Journal of Fluid Mechanics. 11. Kumar BD. Available from: https://en.org Indian Journal of Science and Technology 7 . Johansson AV. Mahesh K. cio SS. Bleier FP. Aerospace Science and Technology.indjst. Energy and Power Engineering. Eslava RT.org/ 12.wikipedia. Manshadi MD. Gharbanian K.