ATLASRPC FLOW CELLS EPDT Gas Team R Guida

ATLAS-RPC FLOW CELLS EP-DT Gas Team R. Guida, M. Corbetta 1

OUTLINE ■ Flow oscillations in different Racks ■ Flow oscillations in RPC racks in ATLAS/CMS ■ Flow leaks ■ Problems with racks ON/OFF ■ Valves Reliability ■ Pump Pressure Test ■ Valves Test (US/UX) – Rack 61 ■ Buffer Test – Rack 61 ■ Pipes Modification – Rack 61 ■ Conclusions EP-DT Gas Team, R. Guida, M. Corbetta 2

FLOW OSCILLATIONS Average Sigma per Rack ■ A sample of 12 I/O couples of flow cells per each gas rack was analysed, looking at data from Nov 2018 ■ Output cells show larger oscillations, with Rack 1 and Rack 2 being the worst cases (7%) 7. 0 6. 0 Sigma [%] ■ Flow measured from Input cells have small oscillations, within 1% 8. 0 Value with good PID settings 5. 0 4. 0 3. 0 2. 0 1. 0 0. 0 1 61 2 62 3 63 4 64 5 65 Rack Input Output Racks with modification EP-DT Gas Team, R. Guida, M. Corbetta 3

FLOW OSCILLATIONS ■ The difference in Input/Output oscillations is significant in ATLAS RPC racks ■ CMS RPC show smaller oscillations and negligible difference between the sigma of Input and Output channels * I/O with same scale, shifted to show the trend CMS - RPC EP-DT Gas Team, R. Guida, M. Corbetta ATLAS - RPC 4

FLOW OSCILLATIONS CMS RPC - Flow Cells 8. 0 8 7. 0 7 6. 0 6 5. 0 5 4. 0 IN 3. 0 OUT Sigma % ATLAS RPC - Flow Cells 4 2. 0 2 1. 0 1 0. 0 IN 3 OUT 0 1 2 3 Racks EP-DT Gas Team, R. Guida, M. Corbetta 4 5 1 2 3 4 5 6 Racks 5

AVERAGE FLOW LEAKS ■ Within the oscillations error, racks show an average leak of the 15 -20%, except for Rack 65 with 30% ■ The leaks are also reported from experiment people, visible in their flowmeters mounted on the chambers 35 Leak [%] ■ The same flow cells were also analysed in terms of flow leaks Average Leak per Rack 30 25 20 15 10 5 0 1 61 2 62 3 63 4 64 655 Racks EP-DT Gas Team, R. Guida, M. Corbetta 6

RACKS SWITCH ON/OFF Pressure Variation ■ It is reported from ATLAS people that when Racks are switched off/on, new leaks are appearing in the chambers ■ With the current procedure, the pressure variation created is of about 9 mbar (PT 6124 in Rack 61) ■ We try to use a new procedure, to limit the impact on the chambers NO Leak New Leak Intervention EP-DT Gas Team, R. Guida, M. Corbetta 7

RACKS SWITCH ON/OFF ■ With the addition of capillary pipes to the compressed air valve, the opening/closing operation is smoother and the pressure variation was reduced from 9 mbar to 3 mbar EP-DT Gas Team, R. Guida, M. Corbetta module input pressure NO Leak detector pressure NO Leak New Leak 8

RACKS SWITCH ON/OFF ■ Using the new procedure allows to reduce the pressure variation (0 -1 mbar): NO Leak New Leak detector pressure NO Leak module input pressure New Leak EP-DT Gas Team, R. Guida, M. Corbetta 9

VALVES RELIABILITY Problems with valves in the past: ■ ATLAS MDT, few years ago > electronic chassis stopped working, valve closed ■ RPC 904, old valve recuperated for Closed Loop Test > leak of 20 -50 l/h in joint around axe (between piloting part and valve), air inside system with loss of gas EP-DT Gas Team, R. Guida, M. Corbetta 10

PUMP PRESSURE TEST ■ Pump pressure is normally set to -5 bar ■ It was tried to vary this pressure, down to about -7 mbar Pump ■ No effect was seen on the flow cells trend Flow Cells EP-DT Gas Team, R. Guida, M. Corbetta 11

RACK 61 - VALVE TEST Sigma Trend - Rack 61 14. 0 ■ Rack 61 was tested with a new valve in US, responding to the PT 100, while leaving open the valve in UX ■ When at 50%, the PID regulating the US valve was very unstable ■ With 65% and 100%, the PID could stabilize to an equilibrium and the overall oscillations of the channels has diminished ■ *Line 10 showed problem in the past lin 2 OUT line 3 OUT 10. 0 SIGMA [%] ■ UX valve: set to 100%, 65% and 50% line 1 OUT 12. 0 line 4 OUT line 5 OUT 8. 0 line 6 OUT line 7 OUT 6. 0 line 8 OUT 4. 0 line 9 OUT line 10 OUT 2. 0 line 11 OUT line 12 OUT 0. 0 1 PT 100 UX Nov 2018 Saver=6. 8 EP-DT Gas Team, R. Guida, M. Corbetta 2 PT 100 US (UX 100%) Saver=4. 7 3 PT 100 US (UX 65%) 4 Saver=4. 5 12

RACK 61 - BUFFER TEST ■ To limit pressure fluctuations an extra buffer with a flexible membrane was installed on Rack 61 ■ The presence of the buffer makes the system less stable, it was kept for few hours to see the possible effect ■ No variation was measured on the oscillation width of Output flows from flow cells is Rack 61 13

RACK 61 - BUFFER TEST Sigma % - Rack 61 14. 00 12. 00 Sigma % 10. 00 8. 00 6. 00 4. 00 2. 00 0. 00 1 2 3 4 5 NO buffer 6 7 Channel 8 9 10 11 buffer Buffer EP-DT Gas Team, R. Guida, M. Corbetta 14 12

RACK 61 – MODIFICATION ■ Some pipes modification were done on Rack 61, trying to limit the oscillations in the flow measured by flow cells ■ Test A : Fast PID, Simple Impedance (as Rack 64/65) Test B 1: Fast PID, 19 pipes, ID 4 mm, L 3 cm Test B 2: Fast PID, 19 pipes, ID 4 mm, L 10 cm Test B 3: Slow PID, 19 pipes, ID 4 mm, L 10 cm ■ Test A showed a significant decrease of the oscillations in all channels, with Sigma going from 4. 5% down to 3% ■ No significant different was seen in the sigma of the flow oscillations in the three different periods of Test B EP-DT Gas Team, R. Guida, M. Corbetta 15

RACK 61 – MODIFICATION Test A Sigma % - Rack 61 Modification 12. 00 10. 00 Sigma % Sold = 4. 5% Snew = 3% 8. 00 6. 00 4. 00 2. 00 0. 00 1 2 3 4 Normal EP-DT Gas Team, R. Guida, M. Corbetta 5 6 7 Channel 8 9 10 Simple Impedance 16 11 12

RACK 61 – MODIFICATION Test B 1 EP-DT Gas Team, R. Guida, M. Corbetta Test B 2 Test B 3 17

CONCLUSION ■ Manipulations on Rack 64 and Rack 65 significantly improved the width of Output flow oscillations (up to 5%) ■ Stopping and restarting the Racks seems to increase the leaks in the chambers, the modification of the valve and new ON/OFF procedure can eliminate the problem ■ PID values set correctly contribute improving Output flow stability (-1%) ■ The buffer did not help improving the oscillation, but made the system unstable ■ Modifications to Rack 61 (as Rack 64/65) allowed to reduce Output flow oscillations (from 4. 5% to 3%) EP-DT Gas Team, R. Guida, M. Corbetta 18