A Study of Turbulence Created by Stationary Objects in a Wind Tunnel

Renee Wald
IB Physics II

Table of Contents

Introduction

[ Table of Contents ]

One of the hardest scientific problem is that of calculating turbulence. In 1952, physicist Horace Lamb said, "I am an old man now, and when I die there are two matters on which I hope for enlightenment. One is quantum electrodynamics, and the other is the turbulent motion of fluids. And about the former I am rather optimistic" (Moin 63). Turbulence is all around and even inside us, yet it isn't fully understood by top specialists in fluid dynamics. Many things can create turbulence. Anything that disturbs the smooth, laminar flow of a fluid will create eddies: turbulence. However, the questions scientists ask are not about the causes of turbulence; instead, they relate to calculating and controlling it.

An understanding of turbulence, however limited, is key in building almost anything having to do with the flow of fluids, such as airplane wings and engines, boat designs, artificial hearts, weather forecasting, and pipelines (Cipra 1361, Moin 63, New theory 951). Turbulence increases the pressures on all of those structures, and a design that could limit or control the effects of turbulence would have a tremendous effect on their efficiency. This is one of the reasons Richard Feynman called turbulence "the most important unsolved problem of classical physics" (Moin 63). However, turbulence can't be completely controlled until it is understood, and that isn't as easy as it sounds.

In recent years, understanding turbulence has become much easier with the development of technology. One of the technological advances that helps with the understanding of turbulence is computer simulation (Moin 63). Recently, computers have become advanced enough to simulate the effects of turbulence. These simulations decrease designers' reliance on wind tunnel testing, which is expensive and can be unreliable (Cipra 1361, Moin 63). Greater understanding of turbulence has also been achieved through structures like the Superpipe at Princeton, with which experiments are conducted exploring turbulence at high pressures (Cipra 1362, New theory 951). More detailed images of turbulence provided by new photography techniques are also improving scientists' understanding of turbulence (Cipra 1362, Wright 1609). However, none of these developments have led to a full understanding of turbulence.

Although there are disagreements in the interpretation of data from the Superpipe (New theory 951), and designers still use "unreliable" wind tunnels to verify computer-formed aircraft designs (Moin 63), the new developments have allowed greater control of turbulence. Researchers have recently demonstrated that riblets, small v-shaped grooves, on an aircraft's wings control eddies and reduce drag (Moin 67-68). This and other drag-reducing shape changes would not have been possible without the increased understanding of turbulence provided by new technologies.

These technologies, although extremely helpful to scientists and designers, are not available to a high-school student. However, the method previously used by the designers, namely the wind tunnel, is available. I used that method to observe turbulence for myself. I wished to find out the effect of shape on turbulence. I assumed that more streamlined objects would create less turbulence and I set out to test this hypothesis.

Setup

[ Table of Contents ]

My first job in my study of turbulence was to make a wind tunnel. I found an article by Shawn Carlson that described how to make a wind tunnel inexpensively, using the materials from the garage (Carlson 106). Unfortunately, Carlson's idea of inexpensive wasn't quite the same as mine, and my garage, apparently, wasn't as well equipped as his. I was able to make substitutions and alterations in his design, and I made a wind tunnel as inexpensively as I had hoped I could.

The wind tunnel was fashioned with a long cardboard box as the central tube, and a household fan created the wind. The fan was attached to the box with duct tape, and a plastic sleeve, created from a garbage bag, kept the connection airtight. The assembly was secured to a sheet of plywood, with the central tube being supported by another cardboard box. The airflow in the experimenting area was kept smooth by window screening on one side and a cardboard grid on the other. A clear plastic window in one side of the box allowed me to see my experiments, while a hook assemble suspended from the top of the box held the objects on which I was experimenting.

Experiment

[ Table of Contents ]

I soon found that the wind tunnel alone did nothing for my experiment. I needed to be able to see the airflow and the turbulence and to do that, I needed smoke. My first method of making smoke was to use incense. However, the incense made only a small stream of smoke, and it had the added disadvantage of giving me an extremely distracting headache. Deciding that incense was a bad idea didn't take long, and soon I was experimenting with a controlled fire using fallen leaves as fuel.

It took a while before I found a method that would produce smoke for any length of time. The leaves burned slowly and produced enough smoke that they were an ideal fuel for my purposes, but they were difficult to start burning. The smoldering leaves would not light any other leaves, and I was forced to use newspaper to keep the fire going. Since newspaper burns quickly and flames up nicely when it's first lit, I needed someone with me to watch the fire during all of my experiments.

The last problem I had with my experiment setup was difficulty seeing the smoke. Even with the greater quantity of smoke created by the burning leaves, it was hard to see what was going on. I was forced to turn off all of the lights in the garage and use a flashlight in order to see the experiment. I could see the turbulence better when I focused the beam of the flashlight, but this made it impossible to look at more than a tiny area. For this reason, it took a lot longer to examine the turbulence on each shape than I had expected it to.

When I finally got everything set up right, I examined the turbulence on seven different shapes. The first object I observed was a golf ball, which is like a sphere with indented "dimples" covering it. I wanted to compare the turbulence of a golf ball with that of a normal sphere because of a diagram in one of the articles I read (Moin 66). The dimples are supposed to change the airflow around the ball and create more pressure behind it than a regular sphere would have. My observations with the golf ball and a ping pong ball (a normal sphere the same size as the golf ball) confirmed this. With the golf ball, the airflow stayed close to the surface and the only turbulence was in its wake.

The ping pong ball, however, had a turbulent boundary layer surrounding it, and its turbulent wake was longer than that of the golf ball.

My next object was a cassette tape, which, I discovered, acted like a rectangular disk. When I turned it so the narrow side went with the airflow, it didn't create any turbulence.

Eddies were created when I turned it the other way, though. There was turbulence both in front of the cassette, and in its wake.

The circular disk (a lid) that I observed created the same type of turbulence as the cassette.

Next, I observed a cylinder. The turbulence of the cylinder was almost all created by the corners on the top and bottom. The eddies created by these corners spilled out and formed a large boundary layer on top and under the cylinder as well as contributing to the wake.

Following the cylinder I observed a rectangular box, which had the same turbulent pattern as the rectangular disk, with the exception of a lengthened wake.

My final object to observe was a cone, which I observed in three different positions. The first position was point up, in which position the only turbulence was along the bottom, flat, side.

Next I placed the cone point first, which created a very little turbulence in its wake.

Then I placed it point back, and in this position the turbulence was all in a boundary layer at its front.

I also measured the wind speed of my wind tunnel by a method I found in Carlson's article (Carlson 107). I tied a ping pong ball to a protractor and found the wind speed with this calculation.

Conclusion

[ Table of Contents ]

Based on these observations, I have concluded that turbulence is created along the flat edges and in the wakes of objects. The most turbulence was found on the disks and the rectangular objects, which were composed solely of flat surfaces. These findings agree with my hypothesis that more streamlined shapes would create less turbulence.

Works Cited

[ Table of Contents ]

Warning: the Science Magazine links require a subscription.

Links

[ Table of Contents ]
The Wind Tunnel Web Page
Contains the history of wind tunnels, ways to build them, flow simulation software, and a page of links.
Stanford's Gallery of Turbulent Flows
This site has animations and simulations of many types of turbulent flow, with detailed descriptions.
Turbulence at Los Alamos
This site has two-dimensional simulations of turbulence, with easily understood descriptions.
Computational Fluid Science Group
Essays and visualizations of different types of turbulent flow.
ONERA
This french design center has diagrams of the wind tunnels they use.