ATOC 1070 Weather Lab 8

Wind Tunnel Experiments: Wind and Pressure with Real World Application

A. Introduction

Wind Tunnel experiments are used in meteorology and engineering sciences to investigate flow over and around obstacles, such as mountains, airplane wings, or vehicles, under controlled conditions. This experiment is designed to examine the relationships between wind speed and the pressure it exerts, the effect of high winds on vehicles, and the mechanism by which very high winds can destroy buildings.

B. Objectives

This laboratory exercise consists of three parts.

Part I. Use a wind tunnel to establish the wind-pressure relationships when air flows over a flat surface. A "pitot tube" sensor and a turbine anemometer are used to obtain both pressure and windspeed. The Excel spreadsheet will fit a curve to your data, and you will compare your equation describing the wind-pressure relationship with the theoretical equation.

Part II. Measure the effects on wind on model vehicles, and then scale the results upward to predict the effect of wind on full-size vehicles.

Part III. Finally, we will measure the distribution of pressure (force per unit area) when air flows over a model Quonset hut. This exercise will also give us some insight into what might happen when air flows over a building in a severe windstorm, such as a tornado or hurricane (check out this video of what actually to a building when a tornado passed over it). Extrapolation is necessary because our little wind tunnel cannot generate wind speeds comparable to those, say, in a tornado.

C. Procedure

Part I.

1. Setup: The sketch below shows the setup for this part of the experiment. In addition, you will use a "Turbo-meter" wind meter, or anemometer.

A Pitot tube (NOT a Pilot tube!) is a device that measures velocity by letting upstream air impact the end of the bent tube. The wind will exert a force (a.k.a. pressure) on the air inside the tip of the tube. This wind-caused pressure at the tip of the pitot tube is transmitted through the rest of the tube and connecting hose to a pressure measuring device called a manometer, described below. Intuitively, one would expect the magnitude of the pitot pressure to increase with windspeed. Theory predicts that pressure should increase as the square of the windspeed. (This is the Bernoulli effect, with which you may be familiar.) That is, the manometer attached to the Pitot tube should give a pressure reading

P = 0.5 * [air density] * V2

Using the Manometer: First, remove the black rubber caps, and zero the manometer by adjusting the thumbwheel at the right port. Zero pressure corresponds to room pressure. Attach the hose to the port so that pressure will force the fluid down the tube (but don't attach the hose so tightly that you can't easily remove it). As pressure rises, the fluid is forced down the manometer tube and into the reservoir (see sketch above). As the fluid in the reservoir rises, it exerts a back pressure on the tube. When the two pressures (air down the tube, liquid pushing back up) are equal, equilibrium is reached, and you can read the pressure in the tube. The scale is calibrated in standard scientific units for pressure, Pascals (Pa, or newtons/meter 2).

In this section, we expect the wind to cause a high pressure in the tube, simply because air is blowing into the tube, so attach the manometer hose to the left port, as shown.

2. Experiment

Run the Variac (Powerstat) settings from 20 to 50 (in steps of 5), record thevalues of pressure (with the manometer) and wind at the exit (using the anemometer). Be sure the anemometer is set for meters per second (m/s). Also, be sure to take the pressure reading before taking the wind measurement, as holding anything at the end of the tunnel changes the pressure inside the tunnel. Check that the fluid level in the manometer is stable reading before reading the pressure. If the pressure reaches 120 Pascals, stop the experiment (DISCONNECT THE HOSE). Although the wind tunnel will run a little faster, it gets really noisy (especially if everyone is trying for high speed at the same time); thus we will only reach about 14 m/s.

3. Data Analysis

Enter your data in the spreadsheet (Wind Tunnel.xls) provided on WXPAOS09 in the "1070 Labs", "Wind Tunnel" folder. The spreadsheet will create a graph of manometer pressure vs. turbine speed.

Note on math:
A mathematical relation y=aX n is called a power law relation. The "a" is a number (called a constant) and the relation says that y is obtained by raising x to the n'th power, then multiplying it by the number a. Familiar functions have n = 1 (a straight line) and n = 2 (a parabola). Other values of n gives curves of different steepness or curvature.

The curve fit equation on your graph shows that P = a V n. Theoretically, n = 2. However, suppose that while you measured pressure accurately, your measured values of windspeed were higher than the true windspeed in the tunnel. You would end up with a value of n that is less than 2. The smaller value of n would compensate for the larger value of V in the equation (while keeping the value of P the same). Conversely, if your measured windspeed were lower than the true windspeed, you would end up with an n value greater than 2.
Now, suppose that your wind measurements are accurate, but your measurements of pressure are not. If your measured pressures are higher than they should be, n will become greater than two (because you raise V to a greater power to get a larger value of P). And, finally, lower than true pressure readings will results in your value of n being less than 2.
Armed with this knowledge, you may tackle the first question:

Question 1:
The curve fit equation on your graph shows that P = a V n. Is your experimental value of n greater or less than the expected value of 2? Refer to the previous paragraph to explain what measuring errors (pressure readings too high, etc.) could cause your value of n to be higher or lower than 2. What problems with the setup or experimental procedures or setup could have caused these measuring errors? Be specific!

Part II.

Now you will measure the effect on wind on a model vehicle, and by a method of extrapolation called "scaling" (not to be confused with scaling fish or scaling climbing walls), estimate the effects of high winds on real vehicles.

1. Setup

Remove the pitot tube from the wind tunnel. Place a "Hot Wheels" (or other brand) model car about 6 inches from the exit end of the wind tunnel. The model car should be placed broadside to the flow of the wind.

2. Experiment

Slowly turn up the power on the Variac until the model car blows out of the tunnel or overturns. The car may come out the end of the wind tunnel, so be prepared to catch it. Without changing the Variac setting, measure the wind speed at the end of the tunnel (in m/sec) and record this on your spreadsheet. Then turn off the Variac.

3. Data Analysis

While blowing model cars out of wind tunnels can be entertaining, there are more useful bits of information that can be found from this part of the experiment. One can actually predict how strong the wind should be to blow a real car off the road. To do this, your measured wind speed needs to be adjusted for the small size of the model by a technique called scaling. It works like this:
* A real car may weigh 1,200 kg, or 40,000 times more than a 30 gram model. Therefore, the force of the wind required to move a real car must be 40,000 times greater than the force needed to move a model.
* The force exerted on an object is pressure times area, where the area is the area of the side of car that is broadside to the wind. Typically, a real car is 50 times longer (and 50 times taller) than a model, so it has 2,500 times as much area.
* For a real car with 2,500 times as much area to have 40,000 times as much force pushing on it, the pressure (force per unit area) must be 16 times greater. Since pressure is proportional to (wind speed) squared, the wind speed must be 4 times greater.
* Therefore, you may multiply your wind speeds measured for the model car by 4 to arrive at the wind speed that will move or turn over a full-size car of similar design and shape.

Question 2:
a.) What kind of vehicle was your model (car, van, truck, etc.)?
b.) What was the wind speed (m/sec) you measured when the model blew out of the tunnel?
c.) Based on your experiment, what wind speed (m/sec) would move or turn over a full-size vehicle?
d.) What wind speed in miles per hour (multiply m/sec by 2.3) would overturn a real vehicle?
e.) What kind of storm could produce this speed of wind (check the tables in Ahrens, pages 267 and 295, or the Saffir-Simpson hurricane wind scale in the Hurricane Lab)?
f.) Does your predicted wind speed (for the full-size vehicle) seem reasonable, or does it seem high or low, and why?

After everyone has computed the answer to part "d" of Question 2, your TA will write a list of vehicle types and wind speeds determined by all the members of your class. With this list, you can answer the next question:

Question 3:
Which two or three vehicles require the highest wind speed to be overturned? Which two or three have the lowest wind speed? Why do these vehicles have different wind speeds? If you were driving a windy highway, such as Highway 93 south of Boulder or the Peak-to-Peak Highwayt west of Boulder (where winds can exceed 100 m.p.h.), which vehicle would you want to be driving? Why?

Part III.

In this section, you will measure the pressure distribution over a model of a simple "Quonset Hut", a rather basic design of pre-fabricated building used by the military in the 1940's. For many years after World War II, hundreds of students lived in surplus Quonset huts along Boulder Creek. If you can't live without seeing a picture of a Quonset hut, click here . You might consider this a small-scale model of a simple barn or other round-roofed building. If you've ever wondered how airplanes stay up in the sky, you can apply the results of this lab to airplane wings to answer this question.

1. Setup

Re-zero the manometer. Install the model hut in the wind tunnel. Replace the bent Pitot tube with a straight tube and attach it to the manometer. We want to get a picture of how the pressure varies along the surface of the model. Position the probe tip very close to the surface, but don't completely block the opening. Note: if the probe is close enough to the hut, the wind speed is essentially zero. Thus, the probe measures pressure, instead of wind speed.

Your setup should look like this:

2. Experiment


Set the blower powerstat at 30 and measure the exit wind speed with the turbine anemometer. On the Excel spreadsheet you used for Part I, find the Part III material. Using the given series of angles, find the pressure for each angle around the cylinder. The locations of these angles are marked on the "hut", with 0 degrees at the floor of the hut on the leading (windward) side, 15 degrees the first mark above the floor, 90 degrees at the top of the hut, and 180 degrees on the floor on the downwind side.
Place the manometer probe close (but not directly on) to the surface at the proper place on the hut. Start the tube at 0 degrees on the windward side of the hut (between the hut and the blower), with the hose going into the left port of the manometer (toward the 0 Pa mark). Take measurements (move up the hut) with this setup until the pressure reaches 0 Pa (near the top of the hut).

Turn off the blower, switch the hose to the other port, and rezero the manometer.
Take measurements down to the floor on the leeward side of the hut. After changing the hose, be sure to enter the pressures as negative numbers. The hose reversal is very important, otherwise manometer fluid is sucked up into the hose or completely out of the glass (making a mess!).

3. Data Analysis.
Record the data in the spreadsheet. The spreadsheet will make a plot of pressure vs angle over the hut. The prevailing air pressure in the room is about 850 millibars, or 85,000 Pa. Thus, the air pressures you measured represent small differences from the prevailing atmospheric pressure at different points on the outside surface of the hut. If you assume that the air pressure inside of the hut is the prevailing atmospheric pressure, you can estimate the forces these small pressure differences are exerting on the roof of the hut. Roughly, 1 Pa equals 0.021 pounds per square foot (p.s.f.). If the pressure difference on the outside surface of the hut is positive, there is a pressure force pushing inwards on the roof; if the measure pressure is negative, the force is pushing outwards.

Questions 4 and 5:

4. Draw a little diagram of the hut, with arrows indicating direction of the wind and the location on the roof of the greatest pressure forces (also indicate the direction of the force). Which part of the hut would you expect to fly off first if very high winds, such as a tornado, moved over the hut? Why?

5. Now, you will extrapolate your data to estimate the forces pushing on the roof of a hut in very high winds.
a.) What was your measured wind speed (m/sec) for this part of the experiment?
b.) What is this speed in miles per hour (multiply m/sec by 2.3 to obtain the speed in m.p.h.)?
c.) Multiply the speed (m.p.h.) by 10 - what do you get? What kind of storm can produce this kind of wind?
d.) What was the greatest negative pressure (in Pascals) that you measured on the hut?
e.) At this higher speed, the pressures on the hut will be multiplied by 10 2, or 100. Under these high wind conditions, what will be the greatest negative pressure (in Pascals) near the top of the hut?
f.) Multiply the above pressure by 0.021 to get the high wind pressure in pounds per square foot (p.s.f.).
g.) Finally, take a 10 by 10 foot section of roof. Using the above pressure (in p.s.f.), what is the force (pressure times area) exerted on this section of roof?
h.) Do you think this force is more or less than the weight of a 10 by 10 foot section of roof? What might this force do to the roof?

D. Write-up

See syllabus for lab report format.

Wind Tunnel Lab grading:

Procedure and Objective: 15%
Activity (spreadsheet with data and graphs): 20%
Each of 5 questions worth 10%: total 50%.
Conclusion 15%
Total: 100%