Cooling tests of the MSGC barrel B1 rod prototype

Transcription

1 B1 note August, 98 Cooling tests of the MSGC barrel B1 rod prototype Jussi Alanen Helsinki Institute of Physics Cooling system of the MSGC barrel B1 rod prototype was studied experimentally. The measurements were done using water and C 6 F 14 -fluorocarbon as coolants with different coolant flows and heat loads. Work also consisted of putting up the test set-up and finishing the measurement software. Page 1 of 14

2 1. Cooling test setup The test set-up consisted of: 1. MSGC-barrel B1-rod with cooling plates 2. RMS Lauda cooling device (used with water) 3. Valve and connector unit to control the flow (used with water) 4. Uniphase cooling unit (used with C 6 F 14 -fluorocarbon) 5. Minco HK51R.1LB Kapton heaters 6. Selec APS M power supply for the heaters 7. Pt 100 and pt 1000 temperature sensors 8. RS Components differential pressure sensor 9. Connector box with internal current reference for the sensors 10. Kobold instruments flowmeter model SFL 11. Pulse counter Kobold type PFE B for the flowmeter 12. Gossen Konstanter LSP power supply for the sensors and the flowmeter 13. Insulated box made of polystyrene 14. PC with LabVIEW version National Instruments PCI-MIO-16E-4 data acquisition card Test object in the cooling tests was the MSGC-barrel B1-rod. The rod was equipped with a cooling pipe, 12 heat exchangers, and 6 cooling plates made of copper. Cooling plates were heated by Kapton heaters to simulate the heat generated by the hybrid s electronics. During the measurements the rod was located inside a thermally insulated box made of polystyrene panels to restrict heat exchange between the test object and the laboratory space. One of the cooling plates was equipped with three pt 100 temprature sensors. A cooling plate and location of Kapton heaters and temperature sensors attached to the plate is shown in figure 1. In addition the inlet and the outlet coolant temperatures and coolant temperature at the middle point of the rod were measured. Pt 1000 sensors were used in these measurements. Due to the difficulties in putting the sensors directly into the coolant or attaching them to a round pipe, an adaptor piece was manufactured. The adaptor was glued onto the pipe. The sensors were then attached to the flat surfaces of that adaptor and, for inlet and outlet, on the pressure sensor connectors.the ambient temperature in the insulated box was measured using a pt 1000 temperature sensor. The coolant flow was measured from the inlet pipe. This measurement was not automated but the values were recorded manually. The pressure drop between inlet and outlet pipe was measured after every change in heating power. This measurement was automated but due to the lack of measurement channels in the connector box as well as on the DAQ-board a continous measurement was impossible. Pressure sensor location as well as the locations of inlet and outlet coolant temperature sensors is shown in figure 2. The overall setup is shown in Figure 3. Page 2 of 14

3 Figure 1. Cooling plate and the location of the heater and the sensors. Figure 2. Pressure sensor and inlet and outlet coolant temperature sensors. Page 3 of 14

4 Figure 3. The test setup 1.2 Measurement system hardware The measurement system for cooling tests is based on a National Instruments data acquisition board installed on a personal computer. Temperature and pressure sensors are connected to the DAQ board via a connector box. The box has connectors for sensors, internal current reference and switches to change input channels mode between current excited temperature sensor mode and simple voltage read mode. Connectors are numbered similarly to the DAQ-board differential channels (i.e. box channel 0 is connected to the board channel 0 and so on). When measuring temperature with resistance-temperature-detectors (RTDs), such as pt 100 or pt 1000, the sensors are supplied with a constant excitation current and the resulting voltage over the sensors is read to the DAQ-board. For this the temperature sensors are connected to the four-pin connectors of the connector box. Current supply is provided by connector box s inner electronics powered by an outer voltage source of about 8 V. It is possible to change the excitation current between 0.1 ma and 1.0 ma by using a jumper inside the box. The box s inner current reference was not ideal for the purpose since the output was temperature dependent. This variation was monitored by using the reference resistance for one channel. The effect in the temperature reading over the constant resistance was 0.4 K during the measurements. Page 4 of 14

5 The voltage read mode is used when reading pressure sensors. The voltage is read using the two-pin connectors of the connector box. Excitation voltage, supplied to the pressure sensors, should be about 10 V (maximum is 16 V). Sensor output voltage is 0 mv when no differential pressure is applied and 100 mv at the upper limit of the pressure range (15 or psi / about 1or 2 bar, depending on the type of sensor used). When changing between RTD and voltage read mode, channel switches on the connector box should be turned towards the connectors used in the measurement (4- pin with RTDs and 2-pin with voltage read). It is important to notice that if some of the four-pin connectors are not in use, when measuring with RTDs, the corresponding switches should be turned to the voltage read position. That short-circuits the excitation current line over the empty connectors so the other connectors are supplied with the current. 1.3 Measurement software The measurement program, named tempre.vi, is implemented in LabVIEW version 5 and bases on a program earlier developped by Hans Danielsson. Program is divided into subprograms pt100-2.vi, pt vi, and press-2.vi. These subprograms take care of reading pt 100, pt 1000 and pressure sensors. These programs also do the conversion from the voltage read to temperature / pressure. Possibility to enter calibration arrays for temperature sensors is also included in these programs. Calibration arrays are entered separately for each type of sensors in the front panel of corresponding subprogram. The main program, tempre.vi, writes the measured values into a text-type file. The file can be read to a spreadsheet program for analysis and presentation purposes. 1.4 Calibration of the sensors The temperature sensors were tied together so that they were in same tamprature and then they were corrected by entering a calibration array so that their temperature readings were the same. This allows measurement of temperature differences between sensors but not the absolute temperature values. It was asumed that in small temperature range (about K), the resistances have a similar enough temperature dependance, thus calibration of the slopes was not done. Flowmeter was calibrated for water using a measurement cylinder and a watch. The calibration factor for the pulse counter found to be 879. With this value the reading of the counter was directly the flow in cc/min. 2. Measurements The coolants used in the measurements were demineralized water and C 6 F 14 - fluorocarbon. Page 5 of 14

6 The current estimate for the maximum power is 1.1 W per heat exhanger. The loads tested were chosen to be 0.5 W, 1.0 W, 1.5 W and 2.0 W per heat exhanger, corresponding a power of 1, 2, 3 and 4 W per heater and a total power of 6, 12, 18 and W for the B1 rod. For the heaters resistance the nominal value of.1 : was used. However the mean resistance of the heaters was found to be.8 :. Due to this the real values for heating power, during the measurements with water as a coolant, were some 5 % higher than the nominal power indicated in the result figures or tables. With fluorocarbon the measured resistance of the heater circuit was used. At first the coolant fluid temperature was chosen to be 12 qc, as in preliminary calculations. This, however, caused some trouble since temperature difference between coolant and ambient was about 10 qc. So the water temperature was changed to be qc. The water warmed up 1 2 qc from this value before entering the rod. this was caused mainly by the extra pump used with Lauda cooling device. With fluorocarbon this kind of rise in temperature did not exist. With fluorocarbon the measurements were done with coolant temperatures 12 and qc. The room temperature during the measurements was about qc, fluctuating, however, by a few degrees during the day. The coolant flow rates used in the tests with water were 60, 90 and 150 ml/min. With fluorocarbon complete measurement series were done with flow rates of 0 and 0 ml/min. This different flow from that of water is due the fact that for the same temperature rise in coolant the fluorocarbon flow must be some 2.5 times higher than that with water. The pressure drop was also checked with flow of 0 ml/min. The exact flow rates during the measurements varied a little due to the difficulty of adjusting the flow. The measurement hybrid, on which the temperature sensors were mounted was moved to three different positions along the rod (close to coolant in / out, middle, far from coolant in / out) to study the effect of the hybrid position along the rod on the heat transfer efficiency. This test was done using water as coolant. 3. Results 3.1 Measurements with water The temperatures for different flow and power values are presented in the figures Same results are presented as tables in appendix A. Each value is an average of ten successive measurements after temperature stabilization. The Temperature differences between the middle of the cooling plate and the middle point of the coolant pipe is presented in figure 11. The variaton of the pressure drop between the coolant inlet and outlet is presented in the figure 12. Page 6 of 14

7 37 Figure 4. middle. Flow rate 154 ml/min. Coolant temperature qc 37 Figure 5. middle. Flow rate 63 ml/min. Coolant temperature qc Page 7 of 14

8 38 37 Figure 6. close to coolant inlet/outlet. Flow rate 155 ml/min. Coolant temperature qc 37 Hybrid, outlet edge Figure 7. close to coolant inlet/outlet. Flow rate 61 ml/min. Coolant temperature qc Page 8 of 14

9 37 Figure 8. close to coolant inlet/outlet. Flow rate 91 ml/min. Coolant temperature qc 37 Hybrid, outlet edge Colant in Figure 9. far from coolant inlet/outlet. Flow rate 153 ml/min. Coolant temperature qc Page 9 of 14

10 37 Hybrid,outlet edge Figure 10. 'T (hybrid middle - coolant middle) [ C] far from coolant inlet/outlet. Flow rate 62 ml/min. Coolant temperature qc. far from inlet, 62 m l/m in close to inlet, 61 m l/m in middle, 63 ml/min close to inlet, 91 m l/m in close to inlet, 155 ml/min far from inlet, 153 ml/min middle, 154 ml/min Figure 11. The temperature differences between the middle of the cooling plate and the middle point of the coolant pipe Page 10 of 14

11 Pressure drop [mbar] Flow [ml/min] Figure 12. Variaton of pressure drop between coolant inlet and outlet as a function of coolant flow. 3.2 Measurements with fluorocarbon The temperatures for different flow and power values are presented in the figures Each value is an average of ten successive measurements after temperature stabilization. Same results are pesented as tables in appendix B. The temperature differences between the middle of the cooling plate and the middle of the coolant pipe for different flows and coolant temperatures are shown in figure 17. The variaton of the pressure drop between the fluorocarbon inlet and outlet is presented in the figure Figure 13. middle. Flow rate 0 ml/min. Coolant temperature qc Page 11 of 14

12 Figure 14. middle. Flow rate 0 ml/min. Coolant temperature qc Figure 15. middle. Flow rate 0 ml/min. Coolant temperature 13 qc Page 12 of 14

13 Figure 16. middle. Flow rate 0 ml/min. Coolant temperature 13 qc 'T (hybrid middle - coolant middle) [ C] middle, 0 ml/min, coolant T=12 C middle, 0 ml/min, coolant T=12 C middle, 0 ml/min, coolant T= Figure 17. The temperature differences between the middle of the cooling plate and the middle point of the coolant pipe. Page 13 of 14

14 Pressure drop [mbar] Flow [ml/min] Figure 18. Variaton of pressure drop between coolant inlet and outlet as a function of coolant flow. 4. Discussion Pt 1000 sensors were found more suitable than pt 100 sensors for this measurement. There was much more fluctuations in the pt 100 sensor readings than when measuring with pt 1000s. It would also help in optimization of the excitation current to have only one type of sensors in use. Also some additional measurement channels would be needed in order to measure pressure continuously and to measure the ambient temperature outside the insulated box. 5. Conclusions The measurement results seem to be very linear. The measured temperature differences between middle of hybrid and middle of coolant pipe show little dependance on the hybrid position along the rod, coolant flow or coolant temperature. The pressure drop was quite linear with both coolants. With fluorocarbon the pressure drop increased more rapidly than with water when increasing the flow. Page 14 of 14