Bigger concerns were raised about the pressure drops in the vapour side of the cooling system, between the detector structure and the BPR, affecting the tem- perature of the coolant and questioning the possibility of having the design -25◦C or necessary -15◦C evaporation temperature in the cooling stave.
The original design plan of the evaporative cooling system [33] assumed 470 mbar pressure drop budget, over the coolant return pipes, from the detector structure (end of stave) to the back pressure regulator. This gave the possibility of having an evaporation pressure of 1.67 barabs in the stave cooling pipe (C3F8refrigerant’s temperature of -25◦C corresponding pressure), in case of the minimum possible back pressure at the distribution rack before the BPR of 1.2 barabs (Section 2.4).
After the installation of the piping system, especially after the necessary modi- fications for the Heater’s design and position (Subsection2.5.4), it became doubt- ful if it was possible to have such a small pressure drop budget over the modified piping structure.
Tests were performed on the installed in ATLAS SCT Barrel loops, to check the cooling circuit temperature (S1 and S2 sensors at the exit of the Stave)(Section 2.6) in case of minimum possible back pressure in the system.
The liquid inlet pressure on PR was set to 12 barabs and 14 barabs leading to the 13 barabs pressure before the capillaries at the detector (the difference being caused by the hydrostatic head for top and bottom quadrants). The system back pressure was varied by changing the value set on the BPR. The BPR was set to the minimum value of 1.2 barabs and to 1.5 barabs, 2.0 barabs and 3.0 barabs. SCT Barrel modules were turned ON (fully powered) with the power dissipated per module of ≈6 W, which is the power load for the unirradiated SCT Barrel module. Measurement of the cooling circuit temperature (average of S1 and S2 sensor values) as a function of the back pressure in the system (vapour pressure measured before the BPR) shown as an averaged of the SCT Barrel cooling loops
abs] [bar Pressure
1.5 2 2.5 3
C]o [Stave Temp.
-20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0
Q1 Q2 Q3 Q4
Figure 3.6: SCT barrel cooling circuit temperatures (averaged by quad- rant) as a function of vapour back pressure.
The small difference in temperature for top (Q1,Q2) and bottom (Q3,Q4) quadrants is caused by a small difference in massflow which depends on hydro- static pressure in inlet lines and set points (See Section 2.4).
The measured temperature of the SCT Barrel staves clearly show that with the existing system it is impossible to reach the design evaporation temperature of -25◦C or necessary -15◦C (for top quadrants) even in case of unirradiated modules. Therefore it is impossible to guarantee thermal stability of the inner detector, especially at the end of the operation period, when modules will be irradiated and work at “full power” (10.5 W).
To study this problem in detail it was decided to assemble a new test setup in the CERN SR1 Laboratory. This test structure duplicates the in-pit installation as much as possible and has been used to perform extensive tests to study the cooling performance of the ATLAS Inner Detector Evaporative Cooling System.
Laboratory Measurements, Analysis and Results.
As the main part of this PhD work, I had a leading role in the design and construction of a test station in the CERN SR1 laboratory. This was used for extensive studies of the thermal behavior of the ATLAS inner detector cooling system; the results from which were used to make predictions of the performance of the ATLAS ID cooling system at the end of life of Phase I of LHC operation.
This work included measurements of pressure drops in the system, measurements of temperature over the stave pipe, study of flow and other parameters. Based on results (presented below) we were convinced that the existing cooling system cannot reach the design evaporation temperature, therefore a new approach was proposed. I studied the thermal behavior of the evaporative cooling system with C3F8/C2F6 refrigerant mixture. To achieve this a machine to blend C2F6 and C3F8 was designed and produced. A sonar based mixture analyzer was designed and developed to enable online real time measurements of the blend ratio. I proved that with the correct blend mixture it is possible to achieve the necessary evaporation temperature (-15◦C) for the ATLAS inner detector and it should maintain the inner detector’s thermal stability even at the end of the 10 years
operation period. The work relating to the design of the test system and the results and interpretation from the C3F8 coolant measurements are presented in this chapter. The work performed on the blend station, sonar analyzer and C3F8/C2F6 blends are presented in the next chapter.