We finally were able to get a window this Saturday where we could test our system for a few hours without the threat of rain, and the test did not disappoint!
Kyle, Travis and I got to the lab at 930 this morning and began setting up the circuit. By 10 we were ready to go and turned on the two transformers set up in parallel. We immediately tested voltages around the system and were confident that the system was working the way we expected.
We proceeded to use our thermometer to test surface temperatures at 7 different locations every 30 mins. We also got temperature readings every 30 mins at spots on the normal asphalt, and a darker shade of asphalt so as to have control temperatures to compare our results too. We noticed a slight heating but it was hard to tell what was really going on.
Charlie came by and helped us take some more readings and then Nelson Gibson, a former student of our mentor showed up around 1215 with his infrared camera and took pictures of our system after it had been on for a little over 2 hours.
The pictures showed us that yes, our system is creating heat along the wire paths. Unfortunately, it also showed that our triplet wires were not producing much heat, and that the heat was not spreading very far from the wires. Nevertheless, the information these pictures will provide will be very valuable and will help us decide where we go from here. We would like to thank Nelson very much for donating his time and services to our team!
After the test was concluded, the team members on the site decided we did not want to completly rip up the system as it had provided us with some good data and might still be of use to us later. We opted to rip up the bottom 4th of the system as it was in bad shape, and then cover the remainder of the system with a layer of sealcoat. The parking space is now in good enough shape that we are confident people can park on it and walk on it without fear of tripping or puncturing a tire.
Today's test was a big success and has given us lots of good information on how to proceed in the future. Next semester we aim to lay down a better version of the system based on the information we learned throughout this semester so we can test during the winter months next school year.
Check out the Team SnowMelt website for thermal images from the tests we ran today. Kyle should be posting them soon!
~Jon
Saturday, May 9, 2009
Friday, May 8, 2009
We Have Heat!
Well at least a little...
After several weeks of frustration with the outdoor test grid, we believe that we finally had some success! After about 5min of power, the grid heated up the pavement from a temp of 22 C to 25-27C even 30C on top of the wires. Finally some good results coming at the perfect time, the day before we are scheduled to preform our long duration testing.
In our original testing with the outdoor grid, we had been using an estimated 20 gauge copper wire to connect the two transformers in our single transformer unit in parallel as well as the transformer unit to the buses. Because of the high gauge of the wire, there was a large voltage drop in the wire connected due to the high current. This caused us to lose 4V from the expected voltage at the top of the buses. We were able to measure the voltage across a wire branch near the bottom of the circuit at ~10V. This means we are losing another 4V in the buses. These loses were unexpected in the theoretical calculations and have thrown our resulting power and heat generation numbers off.
With the above setup I can estimate the following:
Considering the loss in the buses, I can estimate a 11V drop in a NiCr wire near the middle of the grid
V = IR, I = V/R = 11/8 = 1.375
P = (I^2)(R) = (1.8906)*(8) = 15.125 W
Today, we hooked two transformer units in parallel, so that we had 18V at 100A. We also used 8 gauge wire for all the transformer connections. With this setup we had a 16V drop between the buses, so we gained 2V. The remaining 2V are still being lost in the transformer to transformer connects. We also still have a 4V loss within the two buses at the end of the grid.
Preforming the same calculations:
Considering the loss in the buses, I can estimate a 13V drop in a NiCr wire near the middle of the grid.
V = IR, I = V/R = 13/8 = 1.625
P = (I^2)(R) = (2.6406)(8) = 21.125W
So we were able to gain 6W by using the larger gauge of wire. However the question arises is this really enough of a power difference for us to suddenly be able to observe such a obvious temperature increase. In todays quick trial, it was much warmer outside. The initial temperature of the pavement was about 22C as compared to 10-11C in past trials. Our small scale lab tests suggest that this should not be an issue, but in the large-scale I wonder if when it is colder, we lose more heat to the ground.
Another possible path to take to increase the power would be to connect the two transformer units in series. This would theoretically give us 36V at 50A. We know that ideally our .5 ohm grid would draw 72A with a 36V drop across it. However we know that there will be losses inside the connections within the transformers. The situation still needs to be more carefully considered so that we are sure we do not damage the transformers.
Thats all for now, expect another post later this weekend detailing the results of our long duration testing
~Kyle
After several weeks of frustration with the outdoor test grid, we believe that we finally had some success! After about 5min of power, the grid heated up the pavement from a temp of 22 C to 25-27C even 30C on top of the wires. Finally some good results coming at the perfect time, the day before we are scheduled to preform our long duration testing.
In our original testing with the outdoor grid, we had been using an estimated 20 gauge copper wire to connect the two transformers in our single transformer unit in parallel as well as the transformer unit to the buses. Because of the high gauge of the wire, there was a large voltage drop in the wire connected due to the high current. This caused us to lose 4V from the expected voltage at the top of the buses. We were able to measure the voltage across a wire branch near the bottom of the circuit at ~10V. This means we are losing another 4V in the buses. These loses were unexpected in the theoretical calculations and have thrown our resulting power and heat generation numbers off.
With the above setup I can estimate the following:
Considering the loss in the buses, I can estimate a 11V drop in a NiCr wire near the middle of the grid
V = IR, I = V/R = 11/8 = 1.375
P = (I^2)(R) = (1.8906)*(8) = 15.125 W
Today, we hooked two transformer units in parallel, so that we had 18V at 100A. We also used 8 gauge wire for all the transformer connections. With this setup we had a 16V drop between the buses, so we gained 2V. The remaining 2V are still being lost in the transformer to transformer connects. We also still have a 4V loss within the two buses at the end of the grid.
Preforming the same calculations:
Considering the loss in the buses, I can estimate a 13V drop in a NiCr wire near the middle of the grid.
V = IR, I = V/R = 13/8 = 1.625
P = (I^2)(R) = (2.6406)(8) = 21.125W
So we were able to gain 6W by using the larger gauge of wire. However the question arises is this really enough of a power difference for us to suddenly be able to observe such a obvious temperature increase. In todays quick trial, it was much warmer outside. The initial temperature of the pavement was about 22C as compared to 10-11C in past trials. Our small scale lab tests suggest that this should not be an issue, but in the large-scale I wonder if when it is colder, we lose more heat to the ground.
Another possible path to take to increase the power would be to connect the two transformer units in series. This would theoretically give us 36V at 50A. We know that ideally our .5 ohm grid would draw 72A with a 36V drop across it. However we know that there will be losses inside the connections within the transformers. The situation still needs to be more carefully considered so that we are sure we do not damage the transformers.
Thats all for now, expect another post later this weekend detailing the results of our long duration testing
~Kyle
Subscribe to:
Posts (Atom)