Review Of Wind Tunnels.docx

CHAPTER 2

LITERATURE REVIEW 2.1 Introduction

With the advent of high performance computers, the wind tunnel designs have remarkably improved. Improvement in the wind tunnel design leads to much controlled performance. In order to reduce the turbulence and improve flow uniformity, new shapes for contraction section and honey comb structures are developed. Over the past decades lots of research have been done to optimize the shape of the contraction section of the wind tunnel. Many researchers have developed new shapes for contraction section which are latter numerically and experimentally validated. Some researchers have tried different combination of honey shapes and size along with screens of different sizes to reduce turbulence. This section gives a brief on some of the works done on the contraction section, combination of honey comb and screen and complete methodology for wind tunnel design.

2.2 Literature review Prof. Dr. Ihsan Y. Hussain et.al. In their work” design, construction and testing of low speed wind tunnel with its measurement and inspection devices” has designed and constructed a low speed open circuit wind tunnel. In their work they have also describe the design calculation, simulation and construction of the wind tunnel. The proposed wind tunnel has a test section with cross-sectional area of (0.7x0.7m^2) and length of 1.5m. The maximum speed is about 70m/s with empty test section. The contraction ratio is about 8.16. A total of three screens were used to minimize the flow disturbance in the test

Wind tunnel

10

LITERATURE REVIEW

section. The fan used in the wind tunnel were powered by75hp AC motor[2].

Fig 2.1: wind tunnel parts

Fig 2.2: contraction wall shape

Fig 2.3: test section solution

Saiful A. Siddique in his work “Design , develop and CFD validation of a subsonic wind tunnel” redesigned a low speed open type wind tunnel which was supposed to be used experimental investigation, observation and demonstration of a compressible fluid flow phenomenon. The wind tunnel was designed to operate upto 85mph in 200x200mm^2 test section. A CFD model of the wind tunnel was developed and analyzed in commercial CFD software Fluent13. The meshed geometry was composed of 93,903 nodes of tetrahedral elements. RNG K-epsilon turbulence model with scalable wall function was used as turbulent model[3]. Design, Analysis and Renovation of existing Wind tunnel

11

LITERATURE REVIEW

Fig 2.4: assembly drawing of wind tunnel

Fig 2.5: total pressure distribution

Design, Analysis and Renovation of existing Wind tunnel

12

LITERATURE REVIEW

Fig 2.6: velocity magnitude in wind tunnel

Fig 2.7: comparison between CFD and experimental velocity results

Vinayak kulkarni et al, in their work ”simulation of honey comb screen combination for turbulence management in a subsonic wind tunnel”, addresses the design aspect of honey comb and screens for a open circuit wind tunnel installed at IITG, India. The effectiveness of the honey comb and honey comb-screen combination is studied by simulating the flow field by using a commercial CFD package ANSYS - CFX. Turbulence modeling is done by RNG K-epsilon turbulence model with scalable wall function. The simulation was done for honey comb of different length, cell shape and screens of different open area ratio and were found to be in a good agreement with experimental and theoretical result[4].

Fig 2.8: geometric parameters of honey comb structures

Design, Analysis and Renovation of existing Wind tunnel

13

LITERATURE REVIEW

Fig 2.9: geometric parameters of screens

A total of three different cross-section ( circular, square and hexagonal) combined with screens(up to 3) have been considered for the turbulent management. While developing the computational model it was observed that computationally it is very difficult to provide the realistic non-uniform condition at the entry of the settling chamber as experienced in the experiments. Such random inlet conditions would essentially simulate the realistic case in which air can enter the wind tunnel from any direction and at any level of turbulence.

Fig 2.10: geometry of computational domain for(a) honey comb alone and (b) honey comb with one screen ( c)square cell (d) circular cell ans ( e) regular hexagonal cell honeycomb.

Design, Analysis and Renovation of existing Wind tunnel

14

LITERATURE REVIEW

Fig 2.11: geometry used to create practical situation at honey comb inlet

Kahraman Albayrak in his work “ Design of a low speed axisymmetric wind tunnel contraction” presents a method to design a three dimensional contraction based on the expansion of the stokes stream function in a series in terms of the distance in the longitudinal direction and radial distance from the axis of contraction. This method of design requires a suitable function for variation of velocity along the axis of the contraction. In his work the variation of velocity along the axis of contraction is selected such that the velocity increases gradually in the upstream direction and converges rapidly to a value of exit velocity in the downstream direction. In this way the velocity peak on the wall of the contraction is decreased without increasing the length of the contraction[5]. The function used for variation of velocity along the axis of the contraction is achieved by modifying the function for u (x,0) given by Cohen and Ritchie;

u( x,0)  a  b1 tanh( k1 x)  b2 exp( k2 x 2 ) Where, a  4.2,

b1  3.2

b2  0.8 k1 

1 2

Design, Analysis and Renovation of existing Wind tunnel

15

LITERATURE REVIEW

k2  0.3 And the area ratio is given as;

 b1  1   a  a  b1  AR     b1  a  b1  1    a The velocity variation along the wall shows a considerable peak at the inlet regions and the length of the contraction is large. However, the shape the shape of the velocity variation might be improved and the length of the length of the contraction might be decrease by shifting the exponential right and the selecting the constants properly in the equation given by Cohen and Ritchie. The new modified function was given as; u ( x,0)  a  b1 tanh( k1 x)  b2 exp( k 2 ( x  x0 ) 2 )

Where;

a  5.5

b1  4.5 b2  0.1 k1  1 k2  16 / 9 x0  0.6 The main features of the new design are compared with that of the design given by Cohen and Ritchie and following conclusion were made; 1. The present design yields considerably shorter contractions. The ratio of the length to entrance diameter of the contraction is low. Even though the area ratio of the present contraction is high, the present contraction length is shorter. 2. At the inlet regions of the contraction designed by Cohen and Ritchie, the velocity drops on the wall about 70% of the upstream value, whereas for the present design, the

Design, Analysis and Renovation of existing Wind tunnel

16

LITERATURE REVIEW

minimum value of the velocity on the contraction wall at the inlet is 95% of the entrance velocity. Anish Kumar Singh et.al. in their work “ Design and CFD Analysis of contraction wall profile of open circuit blow down type wind tunnel” designed the contraction wall profile of small speed open circuit blow down type wind tunnel based on the design rules given by R.D. Mehta. They have validated the design using commercial CFD code Fluent 6.3. They have identified several desirable characteristics that a contraction wall profile must posses. This characteristic includes, a wall profile having first and second derivative equal to zero at the inlet and outlet, inlet and outlet profile radii roughly proportional to the area i.e, the inlet radius is greater than the outlet radius. With this characteristics most favorable condition of flow uniformity, thin boundary layer and negligible losses is hoped to achieve[6]. A fifth degree polynomial developed by Bell and Mehta (1988) is most widely used for design of low speed wind tunnel contraction profile which is given as;





h   10 3  15 4  6 5 H i  H o   H i





X . L

Where H i and H o are the height of the contraction wall at the inlet and outlet respectively from the datum at the axis of symmetry. Using a transfer function to transfer function to transform Bell and Mehta’s polynomial to arbitrary inlet and outlet heights, while incorporating the change in shape provided by raising the polynomial to a power less than unity. The following transfer function is provided by Brassard and Dr. Mohsen Ferchichi (2005) as





h   10 3  15 4  6 5 H i

1

1

 Ho

 H 

1 

i

Where  is some of  defined for 0    1 . It was shown that any function  chosen for the above equation, normalized to vary between 0 and 1, will result in smooth function, maintaining first and second derivative at inlet and outlet of the resulting contraction profile.

Design, Analysis and Renovation of existing Wind tunnel

17

LITERATURE REVIEW

Fig 2.12: Contraction wall profile when  =1

Fig 2.13: Contraction wall profile when  =0.5

Design, Analysis and Renovation of existing Wind tunnel

18

LITERATURE REVIEW

Fig 2.14: Contraction wall profile when  =  2

Fig 2.15: Contraction wall profile when  = sin 

Finally for the design of proposed wind tunnel  is chosen as sine function of  , it generate large radius at inlet, a smaller radius at outlet while maintaining a longer transition of the outlet radius to the test section.

Design, Analysis and Renovation of existing Wind tunnel

19

LITERATURE REVIEW

Fuh-min Fang et al, in their work Experimental and analytical evaluation of flow in a square-to-square wind tunnel contraction” investigated the flow characteristics within a square-to-square contraction, numerically and experimentally so as to gain additional insight into the contraction design. The problem of the contraction design is to search for a optimum shape with minimum nozzle length for a desirable flow quality at the nozzle end. When the length is reduced, the contraction cost less and fits into smaller space. In addition, the boundary layer will generally be thinner due to the combined effects of the decreased length of the boundary layer development and increased favorable pressure gradient in the contraction. However the possibility of the flow separation increases. Morel proposed the use of combination of two matched cubics, each having its apex at one end of the contraction, as a basic shape of 2-D and axisymmetric contractions design. When the dimensions of the upstream and downstream sections and nozzle length is fixed, the location of the matched points becomes the only parameter to determine the entire wall shape. The geometry of the square-to-square nozzle prototype is shown in the figure 2.16[7] .

Fig 2.16 : Description of the contraction prototype

The size of the inlet and exit cross-sections are, 2.40m x2.40m and 0.80m x 0.80m respectively. The nozzle length (L) is 3.6m. The location of the matched point is selected as 1.80m. The shape of the contraction is the result of a combination of two matched cubics, 3  1 x  y  (h1  h2 ) 1  2     h2  X m  L  

Design, Analysis and Renovation of existing Wind tunnel

x  xm

20

LITERATURE REVIEW

y

h1  h2  1  x 3  h   1  X m 2  L  2

x  xm

Where h1 and h2 are the half widths of the inlet and exit section of the nozzle respectively.

Fig 2.17: Schematic of the contraction (half plane) The experimental setup consists of three rows of pressure taps set along the center-lines of the top, bottom and one of the sidewalls of the contraction prototype. In addition hot wire anemometry is used to measure the mean velocity (x component). For a typical velocity measurement, signals are recorded within a period of 2 min with a sampling rate of 2 samples/s. The measurements and numerical simulations are performed under the condition that the velocity at the nozzle exit section is 15m/s. The results obtained are shown in the following figure 2.18 , 2.19 , 2.20 .

Design, Analysis and Renovation of existing Wind tunnel

21

LITERATURE REVIEW

Fig 2.18: Sectional velocity profiles (x/L=0.32, 0.51, 0.68 and 1.0): (a) calculated; (b) measured

Fig 2.19: Variation of velocity non uniformity

Design, Analysis and Renovation of existing Wind tunnel

22

LITERATURE REVIEW

Fig 2.20: Comparison of pressure distribution along wall center-lines

Kaven Ghorbanian et al. In their work “ experimental investigation on turbulence intensity reduction in subsonic wind tunnel” performed experiments to study the impact of trip wires on the turbulence intensity in a few low speed wind tunnels. They used different sizes of trip wires at different positions in the contraction section of the four different wind tunnels and measurements ere made at various free stream velocities.it was found that on using the trip wire the adverse pressure gradient moves toward the inlet of the contraction section. Consequently the intensity in the test section is reduced and flow uniformity is improved considerably. The Fig 2.21 Shows the position of the trip wires

Fig 2.21: Schematic of the contraction and position of trip wires.

Design, Analysis and Renovation of existing Wind tunnel

23

LITERATURE REVIEW

Table 2.1 : summary of the investigated experiment

Design, Analysis and Renovation of existing Wind tunnel

24