Sami Kiiskilä and Pasi Takala
Tampere University of Technology
Signal Processing Laboratory
P.O.Box 553, 33101 Tampere, Finland
Table of Contents
Heat Flow Around Printed Circuit Boards and Heatsinks in Low Gravity Conditions
Heat Flow Around Printed Circuit Boards and Heatsinks in Low Gravity Conditions
Abstract
The behaviour of thermal flow around heated electronic components during low gravity conditions was studied. Measurements were done aboard the ESA microgravity aircraft with a spatial thermal transducer array around a heated power resistor. Obtained data was then processed with computer to visualize the heat flow.
1. introduction
The purpose of our experiment was to find out how much extra cooling electronic equipment will require in microgravity conditions when compared to normal gravity conditions. In most cases cooling of electronics is achieved by air flow altough in some applications liquid cooling is also used. The air flow required by the equipment is usually created by convection flow or by a fan. Cooling by convection flow is based on the fact that gravity pulls colder air down and that makes the hotter air to move upwards. Since that phenomenon is absent in microgravity the equipment will cool only through conduction and radiation which might cause the equipment to overheat if special precautions are not taken.
2. The equipment
Our equipment consisted of a lexan box that included the measurement system. Block diagram of the system can be seen in figure 1. All numerical references in this chapter refer to figure 1accordingly. The power resistor (1) was used as the heat source. Around the power resistor was a matrix of NTC resistors (2), which were the temperature measuring elements. We had also included a halogen light (7), a video camera (8) and a fog generator (9) to take pictures of the fog moving around the heated resistor. The equipment contained a microprocessor board (3) with A/D-converters (4), to measure the voltage changes caused by temperature changes of the NTC resistors. The processor (5) then sent the measurement results to a portable PC-computer (6) which was used to store the measurements. The system included also a power regulator (10) for the lamp and the microprocessor board. The functions of different blocks were controlled by a switch panel (11) which had also a syringe containing fog liquid for the fog generator. The test chamber was also equipped fwith a fan (12) to equalize the air temperature in the chamber between the measurements..
Fig. 1. Block diagram of the measurement setup.
The whole system was designed to be on board the microgravity flight and the data collected in PC and on the videotape was to be analyzed afterwards. In figure 2 is showed the geometry of the NTC matrix.
Fig. 2. Power resistor and the NTC matrix.
3. operations on the flights
We took several mesurements with the NTC array during two parabolic flights. Data was collected by one second intervals and stored on the hard disk of the PC.
During first flight we turned the heater element on when the first microgravity period begun and collected data during it. The power was kept off during the next two parabolas to let the system cool down. This heating and cooling cycle was repeated five times and during cycles 2 and 3 we let fog in the test chamber before the first parabola of the cycle. On cycle 4 we released the fog during microgravity. During the last three parabolas we turned the heat on at the beginning of each parabola.
On the second flight we turned the heat on at the beginning of each microgravity period and collected the data during it. This way we got 13 series of measurement from the second flight. The remaining parabolas were used to make experiments with the fog generator. This time we didn't used the fog generator and NTC array simultaneously to prevent any interference that might occur. We also turned the heat on at the beginning of six recovery phases of the parabolas to get reference data of how the thermal flow behaves in 1.8G gravity.
Our equipment worked fine in microgravity, injection and recovery phases of the flights. Since our power regulator (10) in figure 1 produced some heat we had equipped it with a small fan to prevent it overheating in low gravity conditions. The support structure of video camera did bent a little every time when the gravity changed but that was expectable and didn't cause any trouble.
We took the Scopdex medication before the flights and Mr. Kiiskilä didn't had any motion sicness on any of the flights and was fully operational during the flights. The only noticeable physiological effect was a minor thirst which was propably caused by the medication. Mr. Takala however got sic after a few parabolas on the first flight and had to vomit which reduced his functionality. He was not on the second flight so we cannot tell if he might got sic on that too. The feeling of weightlessness was a tremendous experience and we recommend it to everyone who might get the possibility to try it.
4. results of the experimentS
Figure 4 shows the results of the first parabolic flight and figure 5 shows the reference results that were measured during normal gravity conditions. The results from the second flight are presented in figure 6 and the results from the 1.8G gravity are in figure 7.
Please notice that the thermal values shown in the figures are not in Celsius but they show the relative voltage change over the NTC resistors during one second interval. Because the NTC has a logarithmic response curve the actual Celsius value may be calculated if the curve for the NTC is known. Since we are interested only in the relative temperature change the actual values are not computed.
We got six data vectors from first flight and the vectors contained 14 samples each. Since one vector contained invalid data we exluded it from the calculations. The picture is obtained from the data by averaging data vectors component by component and then plotting the figures. The vector averaging is presented in equation (1).
(1)
Where x contais the values of all NTC:s in the matrix, m is the number of measurements done during one parabola at 1s intervals and n is the number of parabolas.
When plotting the picture we computed the points between actually mesured values by weighted averaging method presented in figure 3. Points A and B are values of two individual NTC resistors at different locations around the heater element, P is the pixel in the picture that is being calculated and α ισ τηε ανγλε ßετωεεν A ανδ Π.
Fig. 3. Weighted averaging method to plot the pixels.
From second flight we got 13 data vectors with 27 samples in each. Since the microgravity phase didn't lasted during all 27 seconds we used only 20 first samples from each vector to plot the picture. The data was processed similarily to the data from the first flight. The reference data was measured during normal gravity conditions and we used 10 vectors each containing 12 samples. The averaging was done in the same way as with the previous measurements. The last picture was plotted in the same method and the data was measured during the 1.8G phase of the parabolic flight and we used .
When comparing figures 4 and 5 it can be seen that the thermal flow in microgravity is almost nonexistent. Same results can be noticed with figures 5 and 6. From figures 5 and 7 can be seen how the thermal flow gets stronger when the gravity increases.
There is noticeable amount of negative thermal change in the figures 5-7 which is caused by the fact that when the heating element was switched on it takes a few seconds from the power resistor to heat and start to heat the air surrounding it. During that time the NTC resistors are still cooling down from previous experiment and that causes the negative thermal changes. Since we let the system to cool down two parabolas time before reapplying power on the first flight this phenomenon is smaller in figure 4 than in figures 5-7.
The other methor of visualizing thermal flow in microgravity by using fog and video camera didn't produced useable results. It was found out that altough some movement of the fog was observeable in some of the video shots the fog was most of the time uniformly distributed in the test chamber. To get results by this visualization method we should use either very light dust on the heater element and look at the movement of those praticles. It could also be possible to use fog if it is released from several small pipes around the heater element. The lightning for the video camera was sufficient but better results could be obtained if the light would have been formed to a thin plane of light prependicular to the camera and the power resistor instead of using uniform lightning.
5. conclusions
Results of thermal flow experiment in microgravity were presented. Two methods of visualizing the flow were tried. Usage of thermal detector matrix around heater was found more efficient than videotaping fog movement around the heater. It was shown that the convection movement of air greatly reduces in low gravity conditions when compared to normal gravity. If electronic equipments are to be used in low gravity the cooling air around them must be forced to move with a fan.
Fig. 4. Results from the first microgravity flight.
Fig. 5. Reference mesurements in normal gravity.
Fig. 6. Results from the second microgravity flight.
Fig. 7. Results from measurements in 1.8G gravity.
