Low-SPEED WIND TUNNEL: DESIGN AND BUILD ABSTRACT ln this chapter the authors deal with a procedure for the design and bui speed wind tunnel for airfoil aerodynamic analyses and micro wind turbine studies, “The designed closed-circuit wind tunnel has atest chamber with a squ section ($00 mm x $00 mm) with a design average flow velocity of about 30 ms along its axis ‘The designed wind tunnel has a square test chamber, two diffusers (one adjacent to the tet section and one adjaent to the fan to slow the flow), four comers (with turning vanes) to guide the flow around the 9° comers, an axial fan to guarantee the mass flow fate and balance any pressure loss throughout the circuit, a settling chamber with a honeycomb (to eliminate any transverse flow), a series of ever-finer mesh screens (to reduce turbulence) and a nozzle to accelerate flow and provide constant velocity over the ‘whole test chamber. The pressure losses of single components were evaluated as well as the global pressure loss (the sum of pressure losses of all the single components). Once the pressure losses were evaluate, the axial fan was chosen to guarantee the design's volumetric flow, balance pressure losses and above all maximise its performance. The definitive dimensions of the wind tunne! are 10.49 m x 3.65 m. ‘Once the design targets were defined, the test chamber dimensions, maximum wind speed and Reynolds numbers were calculated ‘Atte end ofthe design process, the wind tunnel energy consumption was estimated and on-design and off design performance was evaluated t obtain the wind tunnel circuit characteristics fora defined velocity range (0 ~ 50s) The best circuit and axial fan matches were performed in both the open and closed test section configurations. Using the matching procedure between the fan and wind ‘unnel’s mechanical characteristics (global pressure loss as @ function of wind velocity), the fan operating parameters were set up for optimum energy conservation, 1.0. INTRODUCTION ‘This chapter deals with the fluid-dynamics design of a closed-loop wind tunnel for airfoil aerodynamic analyses and micro wind turbine studies.‘Wind tunnels are measurement tools to study gas flows around a body and the forces ‘generated by the gas-body interaction. For the most par, air is used in wind tunnel. Using such tool it is possible to measure global and local flow velocities, as well as pressure and temperature around the body. Moreover, optical tests using special insemination lize flow motion. The components of a wind tunnel are specifically designed and built to ensure the test chamber an as near as possible uniform in space and independent of time airflow. ‘Wind tunnels’ global dimensions depend mainly on the type of testing, so knowing the shape and dimensions of the wind tunnel test chamber as well as the in-chamber flow velocity, the Reynolds numbers can be established Knowing the test chamber dimensions and wind tunnel test type, all the wind tunnel ‘components can be designed. ‘Wind tunnels can be classified into two main types: open and closed circuit. tunnels can also have two basic test section configurations: open and closed. ‘The air flowing through an open wind tunnel circuit follows essentially a straight path from the nozzle inlet to the test section outlet, prosecuting inthe diffuser. ‘The air flowing in a closed-return wind tunnel, recirculates continuously with no air leakage, ‘There are advantages with both the open and closed circuit tunnels and with both open and closed jets. For an open-retum wind tunnel the advantages are related to lower construction costs and. being able to visualize the flow using smoke without needing to purge the tunnel. The disadvantages are needing to mount extensive screens to obtain high quality flow, greater energy to run the wind tunnel and high noise levels which may cause environmental problems. ‘The closed-circuit wind tunnel has high quality flow, independent of weather conditions and other activities in the building. It requires less enery compared to the open-circuit wind tunnel and produces less noise. The disadvantages of closed-circuit wind tunnels are high construction costs, purging after flow visualization using smoke and needing to mount a heat exchanger. In conjunction with an open-circuit tunnel, the open test section will require an enclosure to prevent air leaking to the diffuser while for closed-circuit wind tunnels, the flow in the ‘open test section tends fo have a solid boundary. Hence, the open test sections is best suited to ‘operating with closed-cireuit wind tunnel. ‘This chapter presents the criteria for wind tunnel design applying them to a specific case study. 2.0. WIND TUNNEL DESIGN ‘The first step in wind tunnel design is related to the shape and main dimensions of the test chamber [1] which depends on the type of intended testsGenerally, wind tunnel dimensions are directly related to the test chamber cross-section. ‘The bigger the test chamber cross-section, the greater the overall wind tunnel dimensions, ‘Test chamber main dimensions and air velocity, as well as wind tunnel type bring to the necessary fin power. These and the wind tunnel’s overall dimensions are key factors in its structure and running costs s0 the best trade-off between costs and tests is necessary. ‘The main goal of wind tunnel design is to have uniform flow within the test chamber, Itis best to have a big testing chamber with very high air velocity ‘The design starts by defining the test chamber dimensions and proceeds counter-stream \ise tothe design of other wind tunnel components. ‘The principal components ofa closed-loop wind tunnel are reported in Figure 1. Second Difiser ee isa an Adapt Sale Gi — ae | a contraction razzie Satins chamber =a Smaier Tes section Honeycomb Comer Larger "S Comer ee Figure 1. Closed cireit wind tunnel 2.1. Test Chamber ‘The first step in wind tunnel design is defining a priori the test chamber criteria which are dimensions, shape and desired air velocity. In this test ease, a square testing chamber with a 0.5 m side was used with an air velocity of 30 m/s. From the testing section dimensions the hydraulic diameter can be calculated as in Eq. Ww. D,=WATa a where A isthe test chamber cross sectional area. ‘The test chamber length has to be in the range of 0.5 - 3 times its hydraulic diameter [1]. This choice takes into account that the air flow exiting the nozzle needs 0.5 times the hydraulic diameter to become almost uniform. Moreover, along test chamber (more than 3 times the equivalent hydraulic diameter) could increase boundary layer thickness causing the boundary layer to detach at the tet chamber exit.So, in this case study the length of the testing chamber was set to twice the hydraulic diameter of the testing section. Moreover, to avoid air velocity reduction and an increase in boundary layer thickness at the sharp edges of the test chamber, the sharp edges should be rounded off. 45° chamfers are best. ‘The test chamber also has flanges and windows to allow sample observations and introduce measuring tools, Figure 2 shows a test chamber coupled to a nozzle, Figure 2. Test chamber (above) coupled tothe nozzle (below) 2.2. Nozzle The contraction or “nozzle” accelerates the flow from the settling chamber to the test section, further reducing any variations in velocity Ina wind tunnel, the nozzle is the most difficult component to design. Flow velocity and its uniformity within the test chamber cross-section depend on the nozzle’s design. The nozzle exit cross-section dimensions and shape are identical to the test chamber ones since they are joined together. Consequently, the nozzle also has 45° chamfers. Knowing the nozzle exit cross-section dimensions and shape, those of its inlet cross- section must be determined, ‘The nozzle area ratio should be ‘as large as possible’, to reduce the total-pressure loss, through the screens mounted between the settling chamber and the nozzle (see Fig, 1). Normally, the nozzle inlet/outlet cross-section area ratio should be in the range 6 - 10 [45 Area ratios greater than 10 lead to excessive inlet dimensions while area ratios less than 6 lead to high pressure loss through the sereens, In the test ease, an area ratio of 7 was chosen, With the inlet and outlet nozzle eross- sections, the nozzie’s silhouette is defined by fifth order Bell-Metha polynomials [4] represented mathematically by Eq. (2). Yaad +bE +08 tds +05 +h @ where x gx @ and ih co) Lis the total axial nozzle length and h is half the cross-section side-length. OsX¥sb (5) In order to determine the Bell-Metha polynomial coefficients, the boundary conditions ‘are imposed. Equations. (6) (11) report boundary conditions. E=0>y=% © Mm ® o (19) sri =0 (ot) Figure 3 shows the nozzle shape.yt ye Se Figure 3. Nozde shape. ‘A nozzle should have a total length and double semi side-length of inlet cross-section about equal to I (see Eq. (12) and Fig. 3) [4]. L 21 2) a. (2) Infact, it was experimentally evident that 2 14(2y,) ratio less than 0.667 causes the air flow to detach close to the nozzle exit, whilst a value greater than 1.79 increases boundary layer thickness. In this case study, the value of £/(2y) was set to 0.91 obtaining a nozzle length of 1.3 m. Connecting the nozzle to the testing section requires the sharp edges ofthe nozzle’s outlet to be rounded off with 45° chamfers, Figure 4 shows the nozzle under construction,Figure 4. Nozzle under construction 2.3. Second Diffuser In order to design the second diffuser, the inlet cross-section area must first be calculated while the exit is equal to the nozzle inlet cross-section area. The second diffuser inlet cross- ction area is governed by the fan dimensions (see Figure 1). ‘Therefore, itis essential to first design the fan. From specialist literature [4], itis known that the ratio between the fan cross-section area Ayand the test chamber cross-section area 4, has to be in the range 2 - 3 (see Eq, (13). <3) To use an 4j/Ay ratio value greater than 3 is not recommended because irregular flow velocities at the fan entrance may be generated. To use & Ay/d,, ratio value less not recommended because it may increase the overall wind tunnel dimensions (higher wind ‘tunnel construction costs), A Ay/dq ratio equal to 2 is a good choice for maintaining low wind tunnel dimensions and ‘costs. In the present case study an 4y/4 ratio equal to 2 was used. ‘Taking into account the area ratios and using the mass conservation law, the air vel at the fan exit can be calculated using Eq, (14).«aay Eq, (20) yields 15 m/s for the case study (cj velocity in test section), ‘Also, from the area ratios, the diameter of the fan cross-section (circular) ean be calculated. In this testcase it equals to 0.800 m, ‘The usual design rule for subsonic diffusers is that the total (29, ) included angle of a Portion of a circular cone with the same length (2) and area ratio (4,) as the diffuser should not exceed a maximum value (6 °[1)). ‘The second diffuser inlet cross-section area is equal to the fan’s, while the second diffuser outlet cross-section area is equal to the nozzle’s inlet. So, they are known values at this stage of the design process. With the hydraulic diameters of the inlet and outlet diffuser cross-sections, the equivalent ‘cone expansion angle can be calculated given by Eq. (15). Te “TD, } a5) ‘Dj, is the inlet section's hydraulic diameter. To avoid having a very long diffuser, 3 ° was chosen (used at the National Full-Seale ‘Acrodynamics Complex at NASA Ames Research Center [9]). Solving Eq. (15) for Z, the minimum diffuser length can be determined (6.58 m for this test ease). Figure 5 shows the diffuser in 3D. flow velocity at fan outlet; c. - air low 9 = Figure 5. Second diffuserIn Figure 5, it is evident that the second diffuser inlet cross-section is circular like the fan's, while the second diffuser outlet cross-section is square. Consequently, the second diffuser also acts as a shape adapter. ‘The second diffuser also has flanges which connect with other wind tunnel parts. ‘Another shape adapter is connected to the fan inlet section so the fan and other wind ‘tunnel components (smaller corner) match. The shape adapter also has flanges. In this ease study, the shape adapter between the fan and smaller comer has been designed. Its total length equals 0.3 m, Figure 6 shows shape adapter in 3D. Figure 6, Shape adapter. 2.4, First Diffuser The inlet cross-section area and shape of the first diffuser are known because they equal the cross-section area and shape of the test chamber. The outlet cross-section area is also known because it equals the inlet fan cross-section area. ‘Since the fan inlet cross-section is round and the frst diffuser’s is square, the side / of the ccan be calculated equaling the areas as in Eq. (16). 16) 11 equals 0.710 m for this case study. As in the case of the second diffuser, the minimum diffuser length can be calculated using Eq. (15). With a maximum total cone angle of 4 °, the first diffuser length equals 3.32 m, Figure 7 shows the first diffuser in 30,10 Figure 7. First diffuser in 3D. Figure 8. Tuming vanes inside a comer,2.5. Corners ‘This wind tunnel is a closed-loop type so flow must be deflected by 90° four times with minimum turbulence at the four comers. For this reason the comers are equipped with blades (more efficient but more costly) or bent flat plates (more economical - our case study). The comers’ dimensions match their related wind tunnel components. Thus, comers are ‘equal in pair and than they will design in couple. 2.5.1. Smaller Corners (1 and 2) Comers | and 2 have a square cross-section equal to the first diffuser outlet’s. The bent flat plates’ leading edges are set to $°, while the trailing edges are 0°. Thus, the bent flat plates” chord is calculated for 85 ° (see Fig. 8). ‘To determine the minimum number of bent flat plates, the chord value must first be calculated and is given by the ratio between the corner section width and the corner divisions chosen. A vane number of 25 is a good choice [3]. a7 Using the first diffuser outlet cross-section (J-0.710 m in the case study), a vane gap (hy2) of 2.84 10° m is obtained. ‘The vane gap-chord ratio has must be less than 0.25, the minimum chord value being calculated by Eq. (18). (18) A value of 0.1136 m was obtained, From chord value, the minimum bent flat plate radius (see Eq. (19) is calculated, (1% where cy. 18 the chord, 1. is the bent flat plate curvature radius, 7 being the central angle subtended by the chord. A minimum radius of 8.4*10° m was obtained and 24 blades were needed. The blades could be installed on a removable frame for cleaning and ‘maintenance. Figure 9 shows the corner in 3D, Figure 10 showing the bladesFigure 9. The comer in 3D, Figure 10. The tuming vanes in 3D. 2.5.2, Larger Corners (3 and 4) Using comers 1 and 2 design criteria, corners 3 and 4 can be designed. They have a square cross-section equal to the second diffuser outle’s (inlet nozzle cross-section). The1B vane gap can be calculated using Eq. (20) with eomers 3/4 side length (1.320 m in the case study). In the 3/4 comer design, a vane number of 25 is also a good choice [3]. 20) Using corners 3/4 data, a vane gap of 5.28 10 m is obtained for the ease study. ‘The minimum chord value and minimum blade curvature radius are obtained using Eq, (21) and Eq, (22) as per comers 1 and 2. en 2.6. Settling Chamber Joined to the 4" comer there is a setting chamber with a constant cross-sectional area. The aim of a settling chamber which contains hhoneycombs and screens is to reduce the flow turbulence before it enters the nozzle Figure 11. The settling chamber in 3D.chamber cross-sectional area matches the dimensions of the other components it is joined to, while the setting chamber length is designed to fit the gap between the components close to the wind tunnel loop. In this study the total settling chamber length is 2.070 m and is able to contain one honeycomb and three screens. Figure 11 shows the setting chamber in 3D. 2.7. Honeycomb ‘A honeycomb with its cells aligned in the flow direction is able to reduce fluctuating variations in transverse velocity [8]. The honeycomb has litle effect on stream-wise velocity 5° AABS, if 0<9,<1.5° K, donne) =} 4+ By9,4C, 9, + D,9;'+ E,9,' + F,9'+G,9,° if 15°S9,<5° (55) AAB SI if 9,>5° ‘Table IV. Eckert’s Ke parameters for circular and square cross-sections Gireular Square 0.1033, 0.09623 0.02389 0.008152 0.1709 0.1222 SOLITON 0.08590 0.03260 0.02203 ‘o.001078 0.003269 0.005076 =0,0006145 0.000013) 0.000280 0.0001388 0.00002337, 0109661 0.01522 0.08672 0.05866 3.3. Pressure Losses In Corners inlet section [1 ‘As far as eneray loss in corners is concerned, the critical comer sections are the two after the test chamber because of greater dynamic pressure and the need for uniform flow at the fan i2 The losses in the comer vanes can be minimised with an efficient blade cross-section, as well as an appropriate chord-to-gap ratio, Vanes with cambered airfoils and straight leading ‘edges are less sensitive to approaching flow angularties than sharp leading edze vanes, To estimate the comer loss coefficient, Eq, (56) can be used [1], assuming that skin friction asa function of the Reynolds number is similar to that on a flat plate. 4.55 KenO0 oes 56 (log, Re.) & where Re, isthe local Reynolds number based on the vane chord. 3.4, Pressure Losses in Screens In wind tunnels, sereen is installed just before the fan section for security reasons. This kind of screen has a significant impact on pressure loss since this isa relatively high velocity section. ther screens contro turbulence just before the nozzle inlet. In terms of energy loss, all screens ate treated in the same manner. W. T. Bekert [3] proposes an empirical relation for the screen loss coefficient based on three main parameters: porosity or its complement solidity, the Reynolds number calculated with wire diameter Rey, and mesh factor Kaa. The latter was studied by 1. E. Idel'chik [6] who assigned it a value of 1.0 for new metalic wires, 1.3 for circular metallic wires, and 2.1 for silk fibres. An average value of 1.3 for Kua is @ good choice in most of cases. Eckert’s empirical equation for calculating the sereen loss coefficient is reported in Eq, (, KK gs Ky Ao “ Bi wr o.7es(1-Bes ) 0400 3.5. Pressure Losses in Honeycombs ‘To determine the pressure loss in honeycombs, the three main parameters of stream-wise length to cell hydraulic diameter ratio, porosity and Reynolds number based on cell hydrauli diameter must be accounted for. W. T. Eckert, K. W. Mort, and J. Jope [3] proposed the relation reported in Eq. (59).na(2a](4}-( 69) where: oxns(2] Rex”! Re, $275 4y ei (0) A 0.214] — Re, > 275 (a) In Eq, (60) Re, is the Reynolds number based on material roughness A and Dy, is the cell hydraulic diameter. 3.6. Pressure Losses in Nozzle ‘The pressure loss in a nozzle is considered only due to skin fretion. Since pressure loss in the nozzle is about 3% of total loss, errors evaluating, Ky are less significant than those made in the high velocity wind tunnel sections so the approximated ‘expression proposed by F. L. Wattendorf [5] (Eq. 61) can be used, K,=032f, ( (61) where Z, is the nozzle length, D,. is the settling chamber hydraulic diameter, and fy is the average friction factor between nozzle inlet and outlet sections. ‘The average friction factor f=, can be evaluated by Eq. (49), with Re equal to the mean value between Re evaluated atthe inlet and outlet of the nozzle, 4.0. CASE STuDY Losses With the above criteria, the loss coefficients for each wind tunnel component can be caleulated. Tab. V shows pressure drops for each wind tunnel component Summing all the wind tunnel section pressure drop values produces the total pressure drop (in this case total pressure drop equalled 141,50 Pa). This pressure drop has to be ‘compensated by the wind tunnel fan.4 ‘Table V. Component Pressure loss at ¢,=30 mvs (c,, = test section air speed) [Fomponens pra jsed Test Secon 12531 Firs Difuser 33.409 fimaller Comer B33 Constant-Area Section 2.068 Smaller Comer 2235 adapter a an Seréen T [ran Constat-Area Seaton Tae Second Diflser 46 [anger Comer 7.968 Sonstant- Arca Section D015 ICarger Comer 7.965 ting Chamber 02138 [Honeycomb 216 First Sereen 1025 Second Seren TORS [Third Sereen 13 ale a7 [Total pressure Toss as Based on the loss coefficients and wind tunnel section pressure drops, and assuming a null relative pressure value in the testing section, the relative pressure values in the wind tunnel sections can be calculated (ideal, without energy loss see Eq, (62);) rea, with energy loss see Eg. (63). (62) (63) where Arun sue i the pressure loss between inlet and outlet cross-sections of the component correlated to the K; factors Static pressure variation within the wind tunnel in ideal and real cases is reported in Figure 14, while Figure 15 shows incremental pressure loss. Figure 14 clearly shows lower pressure values in the real case compared to the ideal case ‘up tothe fan section, From the fan on, the real pressure curve is always greater than the ideal ‘one which is due to an energy gap in the fan section which balances out the pressure losses throughout the wind tunnel ‘The wind tunnel sections* contribution to pressure loss is shown in Figure 16. Clearly ‘most occurs in the first diffuser section,Test Section air speed = 30 m/s. Figure 14. Relative static pressure inside the Wind Tunnel "TestSection air speed = 30 m/s ] T ] ‘cumulative Pressure loss [Pa} eel ee ee < ‘Figure 15. Cumulative static pressure inside the Wind Tunnel. LOLS, BOLLS AES A 2526 Pressure loss[Pa} Figure 16, Wind Tunnel component pressure losses. 5.0. ENERGY RATIO OF THE WIND TUNNEL The wind tunnel energy ratio was calculated as the ratio between flow power in the testing section and the power lost along the cireuit due to pressure losses in all the wind tunnel components (Eq. (64)) 1 E, (64) where i is the /-th component of the wind tunnel. Figure (17) shows energy ratio against testing section air speed. Energy ratio measures the energy efliciency of a wind tunnel [I] and is typically in the range 3-7 for closed circuit wind tunnels. The greater the Energy Ratio the better the wind tunnel energy efficiency.2 Energy Ratio o +} ° 10 20 30 40 50 60 Test section air speed [m/s] Figure 17, Fans. Static Pressreloss Pa] Figure 18, Wind Tunnel Energy Ratio 6.0, AXIAL FAN CHOICE ‘generate air motion within wind tunnel circuits and generally these axial fans are stalled atthe exit of the second corner (see Fig. 1). Closed Test ection ‘Tpen Test Section ‘Wind Tunnel mechanical characteristics.‘The design process consists essentially of matching fan choice from the manufacturer's catalogue 0 the wind tunnel circuit. The first step is determining the wind tunnel's mechanical characteristics (overall pressure drop as a function of air velocity in the test section). The air velocity in the test section can also be expressed as a function of the fan’s air-flow rate using the Mass Conservation Equation. Two mechanical characteristics can be ‘ealculated: the eases of closed and open test sections. These curves are shown in Figure 18. Firstly, for an open testing section, a reasonable friction factor is f= 0.08 [1]. Secondly, the two curves above can be superimposed onto the mechanical characteristics (supplied by the fan manufacturer) of possible fans with matching points graphically determined. Thirdly, the best match indicates the most suitable fan, Figure 19 shows atypical axial fan’s mechanical characteristic. Static Preeue loses] Fan ow Rates) 9. Axial fan mechanical characteristic In that figure pressure gain as funetion of flow rate is reported. Moreover, different ccurves, each for a specific fan rotational speed, are shown in Figure 19. By combining the two mechanical characteristics on the same graph (see Fig. 20), the right fan will match the wind tunne! circuit by intersecting the curve, Choosing the most suitable fan takes into account fan efficiency and design requirements like low rate and air velocity in the testing section and cost. Different graph intersects for different fan rotational speeds provide fan pressure gain and volumetric flow rate. The Mass Conservation Equation can help determine air velocity in the testing section for each rotational speed. Using an inverter to control the electric current frequency and thus the fan motor’s rotational speed, the desiderated air velocity in the testing section is obtained.Fan ow Rtas) Figure 20, Wind Tunne! — Fan matching. By applying this matching procedure, a 5.5 kW axial fan at 1430 rpm was chosen for the cease study (see Fig. 21). Figure 21. Wind Tunnel Axial Fan Fig. 20 shows the intersection points at maximum fan rotational speeds (D), with air flow rates greater than the designed ones (7.5 mis and 35 m/s at the testing section) which happens in the closed testing section and the open one (Depew Fig. 20). Thus, this axial fan ‘meets all the wind tunnel requirements with an efficiency greater than 70%,From 0 - 1430 rpm, testing seetion air velocities from 0 — 37 mis are achievable in the closed configuration and from 0-32 ms in the open. 7.0. WIND TUNNEL CONSTRUCTION ‘The present test case has been built from a mechanical construction company (SET nti S.r1) in the industrial area of Augusta (Syracuse ~ Italy), All the designed wind tunnel components have been built and joined to obtain the closed-loop wind tunnel in the sequence shown in Figure | Figure 22 shows the wind tunnel under construction Figure 22. Wind Tunnel under construction ‘CONCLUSION This chapter deals with the design procedure of a low-velocity wind tunnel for airfoil aerodynamics analyses and micro wind turbine studies. In particular, it focuses on a closed- loop system with a square test section, A test-case design of a wind tunnel with a ‘500 x 500 mm test section of 30 ms along the axis was used as an aid to the design procedure description. The procedure described relates to the design/choice of square testing chambers, diffusers (generally two: one adjacent to the testing section and one adjacent to the fan to slow the flow), comers (with turning vanes) to guide the flow around 90°, axial fans, a settling31 chamber with a honeycomb, a series of ever-finer mesh screens, and nozzles to accelerate flow and provide constant velocity throughout the testing chamber. ‘The design procedure consists of the following main steps: Defining the test section dimensions and desired flow velocity by test type; Wind tunnel component design by test section eriteria, Wind tunnel component pressure loss calculation; Determining pressure loss throughout the wind tunnel circuit as a function of the possible flow velocity inthe testing section in both open and closed configurations; 5. Matching wind tunnel components to commercial fans, and energy considerations. From fundamental choices about wind tunnel design (defining the testing section dimensions and desired flow velocity), the design of all the wind tunnel components is described on the basis of a case study. Particular attention was paid to the nozzle design. This wind tunnel component is the ‘most critical since it strongly affects the quality of the flow (velocity, turbulence level, velocity uniformity) in the test section. Thus, detailed design criteria using Bell-Metha polynomials are shown and specific data forthe test case are provided. Following the components design, determining pressure losses in the wind tunnel cireui is described and energy considerations lead to a proper fan choice by means of a matching procedure. The matching procedure helps study wind tunnel behaviour at different flow velocities in the testing section. In this test case, a testing section flow velocity in the range from 0 m/s to SO m/s was used to determine the wind tunnel’s internal characteristics (pressure loss as a function of testing section flow velocity or fan flow rate). The matching procedure was used to determine the possible wind tunnel running condition, The matching procedure is also useful for establishing the fan speed controller (inverter) positions for the desired flow velocity. In conclusion, the wind tunnel build data confirm the efficacy and usefulness of the design procedure. REFERENCES [1] Barlow J. B., Rae W. H. Jr Pope A. Low Speed Wind Twmel Testing: 3" Ed. Wiley- Interscience: New York, 1999. [2] Metha R. D., Bradshaw P. Design Rules for Small Low Speed Wind Tunnels Journal of Royal Aeronautical Society 1979, Vol. 73. [3] Eckert W., Mort K. W. - Pope J. - deradynamic Design Guidelines and Computer Program for Estimation of Subsonic Wind Tunnel Performance National Aeronautics and Space Administration NASA TN D-8243, Washington, D.C., October, 1976. [4] Bell J. H., Metha R. D. Contraetion Design for Small Low-Speed Wind Tunnels NASA- (€R-182747, April 1988,32 [51 {6 a 8] (91 Wattendort F. L. Factors Influencing the Energy Ratio of Retwn Flow Wind Tunnels Fifth International Congress for Applied Mechanics, Cambridge, September 12-16, 1938, Idel’chick I. E. Handbook of Hydraulic Resistance; The Israel Program for Scientific ‘Translation, Tel Aviv, 1966, AEC-TR-6630. ‘Shames I, H. Mechanics of Fluids; 3° Ed, MeGraw Hill, New York, 1992, Prandtl L. Attaining a Steady Stream in Wind Tunnel NACA TM 726, Oct. 1933. Zal, P. Performance and Test Section Flow Characteristics of the National Full-Scale Aerodynamics Complex 80- by 120-Foot Wind Tunnel. NASA'TM 103920, 1993.