Week 5

April 28, 2016

In week 5 we continued to work through some minor problems to set ourselves up for success in the coming weeks.

The first order of business was establishing what was the appropriate amount of working fluid that should be placed in the pipe. In week 4, we tested our pipe filled 25% of the way with water, and it did not function as desired. There was not enough fluid in the pipe, so the water was condensing before it reached the top of the pipe, and therefore was not transferring heat all the way to the cool end of the pipe.

We knew we needed to add more water, but how much more? Over the gap between weeks 4 and 5 we did some research into prior experiments with heat pipes, as well as looked at some professionally crafted heat pipes to see how much of the way they were filled with fluid. What we encountered was a range of answers, all dependent on not only the heat pipe dimensions but on the materials, angle of operation, and functioning temperature. We saw analyses on heat pipes filled up to 85% of the way with water, while some others were as low as 25%.

Immediately, we thought that equations would be needed to describe our application and to choose appropriately from our results. However, we realized that this simply could not be done, since our heat pipe would not be subject to a constant temperature. Furthermore, we would not know what rate the hot air blower would increase in temperature until the experiment was conducted. Thus, we had to use a best estimate for the appropriate amount of working fluid.

Ideally, we would want to use exactly the right amount of water in the pipe for maximum heat transfer efficiency. But our analysis is not concerned with efficiency, only with wick comparison. Thus, if we had enough fluid in the pipe for it to operate as desired, as long as we kept this amount constant through all of the wick tests it would not matter if the amount of fluid indeed proved most effienct for the specific application. We therefore reasoned that 50% of the pipe would be filled with working fluid.

From there we removed the fluid from last trial, refilled the pipe, and went on to heat the water. However, something unusual happened. When the water was being heated in preparation to secure the cap, without any prior warning a large splash of water exited from the open end of the pipe.

We concluded that this must have been caused by heating only one spot on the bottom of the pipe, which caused a gas bubble to form with the hotter water trapped beneath the cooler water. After emptying and refilling the pipe, we again commenced heating, this time moving the heat source uniformly across the bottom half of the pipe. Once a bit of steam was seen exiting the top of the pipe, the cap was secured in place.

Since we tested our pipe the week prior, there were many other groups who were ahead of us in conducting tests of their pipes. Because of this, we did not conduct any tests in week 5. Therefore, we are still unsure if 50% full is the proper amount of working fluid. In week 6, we will test the pipe in its current state, and if that proves successful, we'll go on to test two other wick structures (three in total), reduced from a total of four due to time and resource constraints regarding the testing apparatus.

three different wick sizes will be compared in our analysis

In the meantime, we were able to cut and assemble the two other wicks. This way, as soon as the proper amount of working fluid is established, we can conduct our tests in quick succession to maximize the time we have to conduct our analysis. The aluminum wire screen was rather rigid and difficult to cut, but we made it work by again using the dowel as a support structure.

a wooden dowel has proven to be most useful in rolling the wicks

Our running hypothesis is that the smaller the wick (holes per inch), the better it will be at conducting heat. Wicks with small openings allow water to condense and be transported more quickly throughout the pipe. We believe this because we have come across the use of sintered metal wicks in a majority of professional applications. Sintered metal is designed to have very small openings which make it easy for water and other fluids to condense and use capillary action as well as gravity to return to the hot end of the pipe. Regarding the wick shown above, with 1/4 inch holes (4 holes per inch), we think that this will not be able to function as a wick. Because the hole size is so large, water will not be able to take advantage of capillary action, and there is not as much surface area upon which to condense. Because of these reasons, we ultimately believe that using this size wick will show no improvement in heat transfer over a heat pipe without a wick entirely.

It will definitely be interesting to tests these hypothesis in the coming weeks, not only to determine if they described the correct trends, but to see by what factor the rate of heat transfer is improved by reducing the size of the holes in the wick. We are going to keep the components as is, using copper for the pipe and water as the working fluid. Although some other pipe materials and working fluids may provide a more efficient transfer of heat, we are more concerned with the wick efficiency analysis in a copper-water heat pipe, the most common type for electronics cooling applications. By focusing on this heat pipe configuration, not only will we be able to conduct a proper analysis but we can produce findings that have relevancy to specific electronics applications.

We look forward to testing and further modifying our heat pipe as well as to beginning an analysis on the rate of heat transfer. We are currently on schedule for a timely delivery of the final report and analysis.

-- Alec, Tran, Matt, and Shjon

Week 4

April 21, 2016

In week 4 we made big strides in constructing the heat pipe. We started by using the pipe cutter to cut off a 2 ft. long piece of copper pipe to be used as the heat pipe. We chose to use this length because we felt that it would be sufficiently large to serve as a model or prototype for other heat pipes. We were not as much concerned with the pipe itself as we were with conducting a thorough wick analysis.

We then cleaned both ends of the pipe and coated it in flux. The cap piece was placed on one end and soldered, and the adapter on the other end and soldered. Both pieces were soldered securely and the connections were waterproof.

We then went to insert the wick, and encountered a bit of a problem. Initially, we thought that we would simply roll the screen a few times over and slide it right into the pipe. However, this was more difficult than anticipated, since the screen rolled unevenly and had a tendency to crease if bent. In addition, the free ends of the screen had wires poking out in many directions, which made it difficult to simply slide the screen into the pipe.

We did manage to find a solution to this problem. We rolled the screen around a long wooden dowel, which helped to provide structure and prevent creasing. This gave us a tight, uniform roll of screen. By creasing a piece of the screen over the end of the dowel, we could then use the dowel as a sort of ramrod to push the screen role all the way in the pipe, removing the dowel at the end.

A wooden dowel was used to provide structure to the rolled screen

PTFE tape was then wrapped around the threaded part of the adapter. This would ensure a secure, watertight seal when the cap was tightened on top. A 1/2 ft. piece of copper pipe was cut off (from scrap pipe) and filled completely with water. This acted as a measuring device so we could fill about 1/4 of the heat pipe with water. We then capped the pipe and inverted it many times to ensure that the entire wick was moist. It is crucial that all parts of the wick are moist before building up pressure in the pipe to ensure that the vapor will condense properly during operation.

A piece of copper pipe was used as a measuring tool to fill the heat pipe

Now came the crucial part of the heat pipe construction. We removed the cap, and securing the pipe vertically, heated the bottom end. As soon as vapors were observed at the open end of the pipe, we quickly secured the cap and tightened it with a wrench. The heat pipe was now complete, but we would have to test it to check if it was working properly.

A test rig was assembled as follows. A plate stand was placed on the table top, and a rubber clamp attachment secured to the vertical support rod. The heat pipe was secured in the rubber clamp and tilted to make a roughly 45 degree angle with the horizontal. A hot air blower was placed on the table adjacent to the bottom end of the heat pipe. A glass shield was placed on the other side to prevent the hot air from directly blowing onto the wall behind. One probe of a temperature sensor was placed in contact with the non heated (higher) end of the heat pipe, and secured next to the pipe with electrical tape. The other sensor was left free to be periodically placed in contact with the heated (bottom) end of the heat pipe.

The hot air blower was then turned on and the waiting began. If the heat pipe was working properly, we would expect to see similar temperatures at both the heated and non-heated ends of the pipe. This would signal that the working fluid is effectively transferring heat away from the heated end.

However, our heat pipe was not working as intended. The heated end of the pipe reached roughly 500 degrees F, but the non-heated end never made it above 100 degrees F. But an important observation was made by moving the probe in contact with the heated end to different parts of the pipe. We recognized that when the heated end was roughly 500 degrees F, the middle of the pipe was around 300 degrees F. This meant that our heat pipe was working properly, but the working fluid was condensing well before it reached the top of the pipe.

We recognized that one of two things could be done to remedy this problem. Our first option was to shorten the pipe to roughly half of it's original size. That way the top of the pipe would become the point at which the fluid was condensing. This option would require a bit of work, since the cap and adapter were already soldered to either ends.

Our second option was to insert more working fluid into the pipe. That way, there would be more liquid to vaporize, and it would remain in the vapor stage long enough to reach the top of the pipe before condensing. This was clearly the better option, as we would only have to add more water and recreate the partial vacuum. Unfortunately, we ran out of time, and did not get a chance to test the pipe with additional water added. Over the course of the next week we will be doing research to try to discover if there is an equation that will allow us to find the exact percentage of the pipe that needs to be filled with working fluid for maximum heat transfer.

Hopefully by next lab period, with these changes instituted, we will have a working heat pipe prototype, at which point the analysis of different wicks can begin.

-- Alec, Tran, Matt, and Shjon

Week 3

April 14, 2016

Now that we had determined the roadmap ahead for our project, week 3 was the time to set everything up so that work could begin. To do that we needed to compile a comprehensive list of materials that would be needed in every aspect of the project. This included not only the pipe and wicks themselves but components such as tinning flux and solder that would be necessary to secure the components.

Projected Budget

Category	Projected Cost	Manufacturer	Location
0.5 in x 5 ft Copper Type M Pipe	$6.76	Mueller	Home Depot
Compression pipe cap (2)	$13.84	SharkBite	Home Depot
Metal Pipe Cutting Tool	$9.97	SharkBite	Home Depot
Heat Insulation Tape	$7.28	Nashua	Home Depot
Dual Temperature Heat Gun	$21.70	Genesis	Home Depot
Temperature Sensor	$26.99	Nicety	Amazon
Aluminum Screen (3)	$33.00	Saint-Gobain	Home Depot
Estimated Shipping	$0.00
TOTAL	$119.54

We also had to begin to consider what materials could be borrowed or rented to try and minimize costs. After talking to the faculty at the Drexel machine shop, we learned that we had access to a wide range of tools and machines, the most important of which being a soldering iron.

We also learned that we could test the heat pipe in lab a few weeks before the end of the term, which will provide time necessary to compile the analysis in addition to the final report. When we test the heat pipe using the test rig, a temperature sensor and insulating tape will be provided by the class advisor, allowing us to collect data without having to purchase those materials.

Through this we were able to eliminate the need to purchase the heat gun and temperature sensor. After purchasing all materials, we also realized we could borrow a pipe cutting tool, so this price will be subtracted from the total cost of materials when it is returned. We decided to make a small design change to the heat pipe, opting to use threaded adapters and caps instead of SharkBite compression caps. Soldering on a threaded adapter will be more secure and reliable over multiple tests.

Another key component to minimization of cost was analyzing shipping rates. After looking at quite a few sites, the shipping rate would only fall to a reasonable price if a) we were ordering hundreds of dollars worth of material or b) we were willing to wait three weeks for shipping. Both of those options were unreasonable, so we decided on paying a bit extra for some materials but eliminating shipping costs by picking up all of the materials from a local hardware retailer.

Because of the tools and materials we were able to borrow and the cost we were able to save from eliminating shipping, we decided to purchase and test four rolls of screen instead of three. In addition, a 5 ft. long copper pipe was purchased, which will be cut into two 2 ft. long pieces and will be used to make two heat pipes. This way the group can use one pipe for testing of the wick (to keep conditions uniform throughout all tests) and then attempt to replicate the findings using the second heat pipe.
While this decision partly arose from the fact that due to a sale the 5 ft. pipe cost just pennies more than the 2 ft. pipe, replicating our results will only serve to bolster the analysis and any conclusions we may reach.

After all materials were purchased, the total came out to be $78.58. The cost of each material is listed in the updated budget below. This total, which is much less than the one detailed in the above budget, is due to the fact that many of the tools and equipment were able to be borrowed. However, the cost some materials, such as solder, an extra set of caps and fittings, Teflon tape, soldering flux, and an extra mesh screen were added into the budget after the decision was made to construct two heat pipes and test four wicks instead of three.

In addition, we are still deciding whether or not to purchase an infrared temperature sensor. The one provided to us in the testing rig may not be compatible with the data collecting software with which we are familiar, and may not be as accurate as we need it to be to establish the correct relationship between the heat transfer variables. If we do eventually choose to purchase this sensor, it will cost $14.63 including shipping (five business days), which will add roughly $8 to the total cost of materials after factoring in the money regained from returning the pipe cutting tool.

Below is a rendered CAD design (to scale) of how the components are going to fit together inside of the heat pipe. The end not shown has water in the bottom and is closed with a soldered cap. The rolled wick will run the length of the pipe and no further; it is shown protruding from the pipe for clarity.

Now that the plan is in place and the materials are in hand, construction of the heat pipe is able to begin. The current goal is to have the entire heat pipe completed by the end of week 5 so that testing can begin starting week 6.

-- Alec, Tran, Matt, and Shjon

Week 2

April 7, 2016

Week 2 was the true planning week for the project. While the first week offered a good start by focusing our group on copper/water heat pipes and their use in electronics cooling, the rest of the project has completely changed.

From the outset, we wanted to conduct a materials comparison as a challenge to not only observe physical properties but to describe them with an equation related to each materials effectiveness in transferring heat.

Initially, we thought to compare working fluids, since they are the most easily obtained and replaced. In addition, we would have loads of data on surface tension, viscosity, heat of vaporization, and other crucial properties at our disposal. However, after looking at various fluids in the comparable temperature range of water, which included acetone, methanol, and toluene, we decided that it would be too dangerous to be heating and vaporizing these materials in the workshop environment.

Our group then switched the focus to the pipe metal itself. We could easily compare a few metals such as aluminum, copper, and steel to see which was most effective at transferring heat. The major advantage here is that we already knew what the results of our analysis would be. By using each materials properties to calculate its Merit Number, which describes its effectiveness at certain temperatures, we could easily see that copper was the best metal to use. However, when it came time to compile a materials list, we had trouble finding steel and aluminum pipes that had the correct threading to be secured with an endcap. In addition, those materials could not easily be soldered, thus we decided to change course yet again.

We finally decided upon comparing wicks. The wick is an internal structure that facilitates the transport of condensed vapors back to the bottom of the heat pipe. While they are not necessary for vertical applications (gravity does all of the work), they are needed in horizontal and near-horizontal applications. Many electronics require heat pipes to be installed in this orientation (for example in laptops), so it would be perfectly applicable to our area of focus.

Expensive wicks can be made out of channeled grooves and sintered metal, but we wanted to focus on metal mesh. Metal mesh wicks are cheap, easily inserted and removed, and can be purchased in small quantities, all important factors for our analysis. For our analysis we will be using aluminum screen mesh. This mesh can simply be rolled up and inserted into the heat pipe as long as it is thoroughly wetted and presses up against the walls of the heat pipe. It is important that the screen mesh be wet, else it will not be able to facilitate the movement of condensed vapors back to the bottom of the pipe. This will create a lack of working fluid, causing the heat pipe to not work properly.

Credit: www.celsiainc.com

Three common wick structures used in heat pipes

However, the main reason we wanted to conduct our analysis on wicks is because we didn't know which would be most effective. When conducting literature study we encountered many tests and experiments detailing the efficiencies of sintered metal wicks and grooved channels, which are the newest, most cutting edge wick technologies. But in all of our study not once did we find testing results on wire mesh wicks.

This is what makes our analysis truly beneficial. Unlike what we expected to do for comparing metals and working fluids, we are not studying some known factor with some end conclusion in mind. We honestly do not know which wick size will turn out to have the fastest rate of heat transfer. This allows us to follow the scientific method by making educated guesses and examining results with nothing to compare them to. Our analysis will then be unique and useful to people who choose to forego expensive commercial technologies in favor of the DIY method of improving and modifying electronic devices.

The next step is to purchase all materials and begin construction of the heat pipe, which should begin by next week. The updated project proposal detailing how we plan to compare the different wicks can be found under the project proposals tab below the original proposal. We remain on schedule as we head into week 3.

-- Alec, Tran, Matt, and Shjon

Week 1

March 31, 2016

Week 1 of the term marked the beginning of our freshman design project. After forming groups, we decided to begin research into heat pipes and their many forms and applications. While each group member was currently studying a different discipline of engineering, and had a varied base of knowledge and expertise, none of us had any knowledge on heat pipes and how they worked.

We embraced the challenge of starting from a weak knowledge base and building up our understanding through research and collaboration. A majority of the week 1 lab was spent learning about how heat pipes worked, what they were made of, and their many common applications. We were quite surprised at not only their simplicity but how common they were in our daily lives, in everything from laptops to air conditioners. It was amazing that so simple a device, one that required no moving parts or external power, could be still so widely used in the many products we encounter.

After building a sufficient knowledge base, we had to consider the application of our final product. We learned that copper/water heat pipes - copper pipes with water as the working fluid - were not only the cheapest and safest to work with but were the most effective for the temperature ranges in which we would be conducting our analysis. The most common application for copper/water heat pipes is in electronics cooling. Small pipes can be used in laptops to move hot air to the fans where it can exit the device. Larger ones are often used in server farms where the heat needs to be transferred away from the servers to a heat sink (such as water) located elsewhere. One can easily see the usefulness of a heat transport device like a heat pipe when the location of a cooler or other heat sink so close to electronic devices would simply not be feasible.

Because of the minute size and complexity required for prototyping a heat pipe for a laptop or other small device, we decided to construct a larger one (about 3 ft in length) that we would be able to more easily analyze towards the end of the course. In a way, the heat pipe we construct can act as a proof of concept of the effectiveness of a copper/water heat pipe itself, which can be bent and scaled as needed for different applications. It is for this reason that we decided to insert a wick into the interior of the pipe to facilitate condensation. Even though vertically oriented heat pipes can operate on gravity alone, the wick is there to show that the same pipe could be used in nonvertical applications and still work to the same or similar effectiveness as analyzed in the lab.

Periodically throughout the next ten weeks, as well as in the final report and presentation, we will take time to discuss how a certain feature of our prototype can be easily integrated into different electronics applications. While we are simply testing the effectiveness of the materials alone, we want to continue to develop our understanding of how a basic heat pipe can be easily modified to serve a very specific purpose.

Going forward, we hope to continue to further our understanding of the many uses and applications of heat pipes. We will continue to keep you updated on our weekly progress, as well as any other notable findings or events as they might occur on our journey to prototype a copper/water heat pipe. We will also begin to post pictures to document our progress along the way.

-- Alec, Tran, Matt, and Shjon