Subsonic Wind Tunnel

Objectives

The primary objective of this experiment is to familiarize the student with measurement of aerodynamic forces in a wind tunnel. A six-component load cell will be used to measure forces and moments on a rectangular (two-dimensional) wing with a removable end plate. In addition, the student will become acquainted with the operation of a subsonic wind tunnel, including use of a Pitot-static probe and pressure transducer to measure wind speed.

Figure 1: The Low Turbulence Wind Tunnel (LTWT) in Montgomery Knight 104

Background

Aerodynamics of a Wing

Forces and Moments

In general, a body moving through a fluid is subject to three force and three moment components. The primary force components on an aerodynamic body, and in particular a wing, are lift and drag. Lift is the force component that is perpendicular to the oncoming flow direction, and drag is the component parallel to the flow (see Figure 2). Lift is generally considered positive when in a direction “upward” (opposing gravity), while drag is normally positive in the flow direction (i.e., tending to slow down the body). The third component is the called the side force. For objects symmetric about the side force axis and moving straight into the flow, there should be no side force. Side forces on a wing usually come about when an aircraft is turning or not flying into the wind.

The three moments are: the pitching, roll, and yaw moments (see Figure 2). The pitching moment is around an axis parallel to the direction of the side force, and would act to change the pitch or angle of attack (a) of the lift body. The roll moment is around an axis in the drag direction, while the yaw moment acts around the lift axis. For most wings flying into the wind, the pitching moment is the dominant component. It is customary to define the pitching moment about the aerodynamic center; the point at which the pitching moment does not vary with lift coefficient, i.e., angle of attack.

Figure 2. Some of the force and moment components on an airfoil. The angle of the attack (α) is shown between the chord line and the flow direction.

The magnitudes of the aerodynamic forces and moments depend primarily on the shape of the body, the speed and orientation of the body with respect to the fluid, and certain properties of the fluid. For a rectangular wing, lift primarily depends on the curvature (camber$^1$) of its cross-section, the angle of attack, the flow speed, and the density of air. The shape of the wing’s cross-section is known as an airfoil.

The wing used in this experiment, shown in Figure 3, is symmetrical and rectangular, with a removable end plate, shown in Figure 4. Its key dimensions are given in the below table. Aerodynamics forces on the airfoil will be measured as a function of wind speed, angle of attack, and presence of the end plate.

Dimension	Value (inches)
Span (wing)	27
Span (end plate)	0.08 (flat aluminum plate)
Chord (wing)	12.2
Chord (end plate)	22
Max thickness (wing)	3
Max thickness (end plate)	9 (3" offset of wing airfoil)

End plates, flat plates mounted perpendicularly on the end of the wing, are "wing tip devices" commonly used in racing cars and historically used in aviation. Conceptually, these plates present a physical barrier that reduce/prevent the formation of wing tip vortices, thus improving the wing's efficiency. Raymer$^2$describes end plates as increasing the "effective aspect ratio" of the wing, leading to lower induced drag, although its presence in the flow will of course increase profile drag. Theoretically, with a decrease in span-wise flow that would otherwise have created a tip vortex, the area of the wing's surface now contributing to lift will increase.

Figure 3: Rectangular wing used in AE2610 (shown with outer panel removed)

Figure 4: The wing installed in the wind tunnel test section, with and without end plate

$^1$The camber line of a wing’s cross-section connects the midway points between the upper and lower surfaces.

$^2$Raymer, Daniel P. "Aircraft Design: A Conceptual Approach", pages 266 and 298.

Aerodynamics Coefficients

Absolute aerodynamics forces on a wing are commonly non-dimensionalized and expressed as aerodynamic force coefficients. For example, the lift coefficient ($C_L$), drag coefficient ($C_D$) and pitching moment coefficient ($C_M$) are defined as:

$C_L = \frac{L}{q_\infty S}$ (1)

$C_D = \frac{D}{q_\infty S}$ (2)

$C_M = \frac{M}{q_\infty Sc}$ (3)

where $L$ is the lift force, $D$ is the drag force, $M$ is the pitching moment, $S$ is the plan form area of the wing, $c$ is the chord length (see Figure 1), and $q_\infty$is the dynamic pressure. For our rectangular wing, $S = c \times b$ , where $b$ is the span (tip-to-tip length) of the wing.

Thin airfoil theory predicts that ($C_L$) for a symmetric airfoil, i.e., where the chord and camber lines are the same (also known as a zero camber airfoil) that also has infinite span is given by

$C_L = 2\pi\alpha$ (4)

This expression is valid only for small and moderate angles of attack. At sufficiently large angles of attack, the flow over the wing no longer follows the surface, and the wing’s lift coefficient begins to drop, i.e., the wing stalls.

The dynamic pressure $q_\infty$ is given by

$q_\infty = \frac{1}{2} \rho_\infty v_\infty^2$ (5)

where $p_\infty$ is the density of the approaching flow and $v_\infty$ is its speed (relative to the body/wing). This is called the dynamic pressure because according to Bernoulli's equation (which is valid for a low speed, constant density flow) it represents the difference between the stagnation (or total) pressure and static pressure of a flow, i.e.,

$p_o - p = \frac{1}{2}\rho_\infty v_\infty^2$ (6)

The static pressure is the pressure one would measure if moving with the flow (or without requiring a change in the flow velocity), while stagnation pressure is the static pressure that the flow would achieve if it was slowed (in an ideal manner) to zero velocity. For an ideal gas, the density is given by

$\rho = \frac{p}{RT}$ (7)

Here $p$ is the (absolute) pressure, $T$ is the (absolute) temperature, and $R$ is the gas constant for the specific gas.

Sensors and Transducers

Load Cell

The device used to measure forces and moments experienced by the wing in this experiment is a six-component load cell. Being six-component, it is able to measure all three forces and all three moments. It achieves this via an array of strain gauges bonded to a machined block of hardened stainless steel. The shape and material composition of this block is cleverly designed to ensure that even complicated combinations of force and moment inputs result in distinct measurable strain outputs. Just like with the Tensile Testing lab, these strains are sensed via strain gauges, also placed strategically on the load cell body and using Wheatstone bridges to convert small changes in resistance to measurable changes in voltage. A factory calibration process relates known applied forces and moments to these voltage outputs, resulting in a mapping (called a "calibration matrix"). This calibration matrix is supplied to the customer who then uses it to transform strain gauge voltage outputs, collected via a data acquisition system, to resolved forces and moments. In our case, the load cell used is an ATI Gamma SI-130-10, which is mounted just below the wind tunnel floor, in between the wing and a stepper motor which is used to vary angle of attack. This setup, shown in Figure 5, sees the wing and stepper motor mounted vertically to eliminate the moment applied to the load cell as a result of the wing's mass. The load cell is mounted directly on the wing's quarter-chord line to remove the need to perform any offset correction. The stepper motor is pre-programmed to traverse from its initial position, having been zeroed, to an angle of attack of 20 degrees in 1 degree increments. At each angle, the stepper will dwell for 3 seconds and command the LabView VI to log the load cell measurements 1000 Samples per second.

Figure 5. Wing, load cell, and stepper motor assembly

Pitot-Static Probe

Utilizing the Bernouilli Equation (6) above, wind speed in the tunnel is measured using a differential pressure transducer, in this case a capacitance-type transducer called a Baratron. This transducer interprets the displacement of a diaphragm due to a pressure difference across the diaphragm as a change in capacitance in an electronic circuit, which is output as a DC voltage change. This voltage change is proportional to the differential pressure experienced, meaning that this voltage measured by a data acquisition system can be transformed to dynamic pressure via a simple linearly scaling. The dynamic pressure is measured directly by means of a Pitot‑static probe mounted in the freestream. The probe has one hole facing directly into the wind tunnel flow and another located so the air flows along the surface of the hole (see Figure 6). The pressure in the first hole is the stagnation pressure ( $p_o$ ) and the second experiences the static pressure ($p$). The two holes are connected via long lengths of tubing to the two sides of the Baratron shown in Figure 7, which has a known sensitivity of 1.016 mmHg/Volt.

If the data are to be meaningful, each test run must be carried out at a (nearly) constant value of wind tunnel dynamic pressure. Ideally, the magnitude of the aerodynamics force coefficients would be independent of the magnitude of the dynamic pressure. In reality, this assumption might not be completely accurate, so the dynamic pressure should be held constant during a particular data‑taking run so that the coefficients will not be influenced by changes in freestream conditions. The speed of the fan that drives the wind tunnel is nominally held constant by the fan’s motor control system. This should produce a nearly constant wind tunnel dynamic pressure. (If necessary, the dynamic pressure can be manually adjusted with the tunnel speed control.)

Figure 6. Schematic of Pitot-static probe in a wind tunnel.

Figure 7. The Baratron used to measure dynamic pressure in the Low Turbulence Wind Tunnel

FlowKinetics Manometer/Flow Speed Meter

Whereas the Baratron measures only dynamic pressure, the FlowKinetics 3DP1A device also measures local ambient pressure and temperature, enabling air density and thus flow speed to be calculated. As shown in Figure 8, the FlowKinetics has a selector dial for choosing the display of either pressure or velocity (with multiple options for units on each). The FlowKinetics is an independent reference in that it does not output any signals that are read during the lab; we will be using it only for a sanity check on flow speed magnitude and for the ambient pressure and temperature measurements.

Figure 8. FlowKinetics 3DP1A all-in-one manometer and flow speed meter

The Wind Tunnel

A wind tunnel is a duct or pipe through which air is drawn or blown. The Wright brothers designed and built a wind tunnel in 1901. The basic principle upon which the wind tunnel is based is that the forces on an airplane moving through air at a particular speed are the same as the forces on a fixed airplane with air moving past it at the same speed. Of course, the model in the wind tunnel is usually smaller than (but geometrically similar to) the full size device, so that it is necessary to know and apply the scaling laws in order to interpret the wind tunnel data in terms of a full scale vehicle. The wind tunnel used in these experiments is of the open‑return type (Figure 8). Air is drawn from the room into a large settling chamber (1) fitted with a honeycomb and several screens. The honeycomb is there to remove swirl imparted to the air by the fan. The screens break down large eddies in the flow and smooth the flow before it enters the test section. Following the settling chamber, the air accelerates through a contraction cone (2) where the area reduces (continui­ty requires that the velocity increase). The test (working) section (3) is of constant area (42" x 40"). The test section is fitted with one movable side wall so that small adjustments may be made to the area in order to account for boundary layer growth, thus keeping the stream-wise velocity and static pressure distributions constant. The air exhausts into the room and recirculates. The maximum velocity of this wind tunnel is ~35 mph, and the turbulent fluctuations in the freestream are typically less than 0.5% of the free­stream velocity. Thus, it is termed a “low turbulence wind tunnel.”

Figure 8. Schematic of the Georgia Tech Low Turbulence Wind Tunnel.

Procedure

1. Take a tour of the LTWT with the TAs, looking particularly at:

   1. The layout of the tunnel as shown in Figure 8

   2. The pitot-static probe(s) used by the FlowKinetics and Baratron

   3. The wing with removable end plate

   4. The load cell/stepper motor assembly

2. Take a safety briefing from the TAs

   1. Consult the power-on checklist for the tunnel operating procedure (also posted on the control room wall)

      1. This has changed SLIGHTLY for Fall 2024. Listen to your TAs!!

   2. Assign one or more visual observers to watch the wing during its sweeps and be ready to hit the emergency stop in the control room should anything untoward happen with the stepper motor.

   3. *READ ALL INSTRUCTIONS FULLY*

3. Help the TAs ensure everything is powered on:

   1. The wind tunnel itself (energize the master breaker near the tunnel inlet on the VFD)

   2. The ESP32 microcontroller commanding the stepper motor (USB cable should be plugged in to PC)

   3. The orange peripherals power cable (this powers the DAQ and the USB hub under the tunnel)

   4. The black emergency stop cable (this powers the stepper motor under the tunnel)

   5. The FlowKinetics manometer - follow the steps to configure it for measuring air (without calibrating the transducer or setting a correction factor other than 1.0). Make a manual note of the ambient room pressure, temperature, and air density.

   6. The large monitor - if the remote is missing use the physical button (on the underside of the front in the middle)

4. Prepare the LabView VI

   1. Open the Aero Loads VI application found inD:\AE2610\<semester year>\Aero Loads\. Drag the VI to the large TV screen so everyone can see unless the screen is already mirrored.

   2. When prompted, select a save file location in D:\AE2610Data\<semester>\<section> and an arbitrary filename (this file will not be saved).

   3. Hit the red STOP button in the top left of the VI.

   4. Select the corresponding Stepper Controller COM Port (consult Windows 10 Device Manager if required; it was COM3 as of 10/26/2024).

   5. Hit the white GO button in the top left of the VI.

   6. When prompted, select a save file location in D:\AE2610Data\<semester>\<section> and a unique filename (this file WILL be saved). Manually type in the file extension as ".csv"

   7. Verify that the Stepper Controller is successfully communicating with LabView by visually checking that Stepper Motion Status and Stepper Controller Time boxes are populated, with the time increasing. In the absence of successful comms, follow this debug sequence, re-trying after each attempt:

      1. Ensure the stepper controller is plugged in to both the PC's and controller's USB port and is powered up (red LED can be observed on the ESP32 stepper controller under the tunnel).

      2. Ensure the correct COM port is selected in the VI (double check against Windows Device Manager). Note that the VI only checks the COM port when it is first activated, any changes in the COM dropdown will be ignored after that, so you will need to stop and restart the VI for any COM changes.

      3. Power cycle the stepper controller by unplugging then re-inserting the USB cable from the PC and the orange peripherals cable. Then close and re-open the LabView VI, re-trying the opening sequence.

      4. Restart the PC, power cycle the stepper controller, and try everything again.

      5. Have a TA call Venkat Putcha

   8. Input ambient pressure and temperature from the FlowKinetics into the Air Density Calculation part of the VI. Input 287 for the Specific Gas Constant. This should yield a realistic air density (note that the graph will not populate until this is done, otherwise it will be trying (unsuccessfully) to plot NaNs.

   9. Zero the Baratron by clicking Tare Baratron.

   10. Zero the load cell by clicking Bias on the load cell control in the VI.

5. Zero the wing. Recognizing that the wing is symmetrical, and thus will produce 0 lift at 0 angle of attack, zero the yaw position:

   1. Ensure the stepper motor power is powered OFF.

   2. Utilizing an observer at the wind tunnel outlet, rotate the wing by hand until it is as close to parallel with the wind tunnel walls as possible. It will not be perfect at this time but that is not an issue as we will fine-adjust momentarily.

   3. Return to the control room and close all doors to the tunnel.

   4. Power on the stepper motor.

   5. Following the power-on checklist, get the tunnel up to full speed. Make a manual note of the wind speed. Make the following two changes to the power-on checklist for Fall 2024:

      1. Perform the initial safety test with the dial set to 20%

      2. Set max speed with the dial set to 50%. There is a physical limiter on the dial disallowing going over this power setting.

   6. Switch the LabView VI to Jog Modeand adjust the yaw position until the observed lift is 0.

   7. Click Set Home when you are happy the observed lift is 0.

   8. Power down the wind tunnel and wait for the fan to stop spinning (this may take a minute).

6. Perform the experiment

   1. Ensure the airspeed reading on the VI is close to 0 m/s. Tare the Baratron again if necessary.

   2. You will perform a total of 4 sweeps. Two different wind speeds, with and without the end plate. It doesn't matter what order you do the sweeps in so if the end plate is still installed from the previous session, do that one first, and vice versa.

   3. Power up the wind tunnel such that the speed on FlowKinetics stabilizes at 16 m/s.

      1. This will be roughly 46-48% on the dial. Keep in mind the wind speed does fluctuate so make sure the fluctuations center around 16 m/s!

      2. With Jog Mode active, command the turntable from 0 to +22 degrees in 1 degree increments. At each position, activate the logging button for a few seconds to capture load cell data.

      3. Note down and discuss anything interesting that you see or capture with your group and TA!

   4. Return the wing to its zero position by clicking Go Home. If the loads are not zero after returning home, you may need to re-zero the angle of attack and set a new home position as you did in "Step 5.5-5.7: Zero the Wing" above

   5. Decrease the wind speed of the tunnel such that the FlowKinetics reading stabilizes at 12 m/s

      1. This will be roughly 34-37% on the dial. Keep in mind the wind speed does fluctuate so make sure the fluctuations center around 12 m/s!

      2. With Jog Mode active, command the turntable from 0 to +20 degrees in 1 degree increments. At each position, activate the logging button for a few seconds to capture load cell data.

      3. Note down and discuss anything interesting that you see or capture with your group and TA!

   6. Power down the wind tunnel. Wait for the fan blades to stop spinning before exiting the control room (this may take a few minutes)

   7. Open the tunnel side panel. If the end plate is installed, remove it. If the end plate is not installed, install it. This is achieved using the provided hex key and the 6 screws.

   8. Zero the load cell by clicking Bias on the load cell control.

   9. Return the tunnel to full speed, following the power-on checklist. Keep the following changes in mind:

      1. Perform the initial safety test with the dial set to 20%

      2. Set max speed with the dial set to 50%. There is a physical limiter on the dial disallowing going over this power setting.

   10. Zero the wing again using the procedure above if necessary

   11. Repeat Step 6.3-6.5 above to capture 2 new data sets with this new configuration.

   12. Power down the wind tunnel when you have finished gathering all 4 data sets.

7. Data check

   1. Stop the VI.

   2. Open the file you selected as your save file in MS Excel. If you didn't give it a file extension when you selected the save file you will need to add it now by right-clicking on the file, clicking Rename, and then adding ".xls" to the end of the filename.

   3. Generate quick scatter plots of Fx, Fy, Tz, and Baratron Voltage against Angle-of-attack to sanity check your logged data. If any data is missing or looks wildly corrupt, repeat the relevant sweeps.

   4. Once successful data has been acquired, ensure it is emailed to your group or uploaded to Canvas before leaving the lab.

8. Assist the TAs in shutting down the experiment

   1. Power down the tunnel master breaker

   2. Unplug the ESP32 stepper controller USB cable from the PC

   3. De-energize the stepper motor and peripherals by flipping the switch on the power strip in the control room

   4. Turn off FlowKinetics

Data to be Taken

1. From the FlowKinetics device: ambient air density, temperature, and pressure.

2. Tared forces, moments, and baratron voltage of the wing at 20/22 angles of attack and two wind tunnel speeds, with the end plate.

3. Tared forces, moments, and baratron voltage of the wing at 20/22 angles of attack and two wind tunnel speeds, without the end plate.

Data Reduction

Tip - Reducing the data from this lab, and indeed in many other assignments in your time at Georgia Tech, can be much smoother/quicker/more accurate if you write scripts to do the manual work for you. Go to the AE3610 MATLAB Tips and Tricks page for some pointers and a how-to guide for reducing binned data (this lab's data is binned by angle of attack and wind speed).

1. For each batch of force/moment and Baratron voltage data collected at each angle of attack, calculate its mean and variance.

2. Convert the mean forces/moments in the balance reference frame to the wind frame (i.e., convert to lift and drag). To do this you will need to know the following facts:

   1. The wing's centerline is perfectly aligned along the y-axis of the load cell, with the front of the wing being in the negative-y direction

   2. The load cell's y-axis is oriented such that positive-y is pointing down the wind tunnel, and the x-axis is perpendicular to the y-axis, with positive-x pointing towards the control room

   3. The origin of the load cell is located on the quarter-chord line of the wing

3. Convert the mean Baratron voltage to wind speed using the known relationship with dynamic pressure discussed above, and the manually recorded ambient air density.

4. Express all forces and moments in non-dimensionalized coefficient form. The chord to be used is the chord of the wing from the wing leading edge to the flap trailing edge as shown in Table 1 above, and the dynamic pressure to be used is that calculated in the previous step.

5. Uncertainty Information needed for Uncertainty Analysis:

   1. Baratron Voltage: +/-0.002 V

   2. Load: +/-0.05 N

   3. Torque: +/-0.001 N-m

Results Needed for FORMAL Report (also see instructions on Canvas)

1. A table containing wing geometry ‑ chord of wing, span, planform, thickness ratio.

   ​

2. Plot a figure showing lift coefficient (ordinate) versus angle of attack (CL vs. $\alpha$) for the wing at the two wind speeds, with and without end plates.

   ​

3. Plot a figure showing drag coefficient (ordinate) versus angle of attack (CD vs. $\alpha$) for the wing at the two wind speeds, with and without end plates.

   ​

4. Plot a figure with drag coefficient as abscissa and lift coeffi­cient as ordinate (CL vs. CD – also known as a drag polar) for the two wind speeds, with and without end plates.

   ​

5. Plot a figure with moment coefficient (ordinate) versus angle of attack (CM vs. $\alpha$) for the wing at the two wind speeds, with and without end plates.