1. Jan. 28, 2014W. Bertl, PSI BPIX Cooling Status W. Bertl, PSI

2. Jan. 28, 2014W. Bertl, PSI Time for design studies is now over ! We have entered the period of testing and verifying our design goals. 3 main objectives: 1)Cooling plant (construction, operation, safety requirements, control) 2)BPIX cooling loops (cooling performance, various detector operating conditions) 3)Sensor cooling (heat conduction to CO2 coolant at full LHC operation )

3. Cooling Plant Jan. 28, 2014W. Bertl, PSI Prototype built, tested and operated by Cern groups. (see D’Auria’s talk)

4. Cooling Loops Dec. 06, 2013W. Bertl, PSI A full scale replica of 50% of the final detector piping has been built by Uni ZH. representing –x and +x side of half of BPIX Pipe extension to the supply tube not shown on this picture

5. Cooling Loops Dec. 06, 2013W. Bertl, PSI One half of the replica is now at Cern. The second half currently at ETH ZH for cleaning. (will be shipped to Cern soon) Testing of its cooling performance under various thermal loads assuming full or only parts of the detector under power over a temperature range between - -30˚C and +20˚C. see Robert Becker’s talk

6. Sensor Cooling Jan. 28, 2014W. Bertl, PSI Sensor temperature depends on : pipe temperatureexpected: L1 = -15˚, L2 = -14.5˚, L3 = -15˚, L4 = -15.5˚ power dissipation per module: L1= 7W, L2= 3W, L3 = 2.4W, L4= 2.3W contact surface contact material see results of test measurement (next slide) For the worst case at layer 1 about 70% of thermal load is dissipated by the ROC, 30% by the sensor. Goal: Sensor temperature sufficiently low such that no thermal runaway and reasonable low leakage current (typ. < -2˚C)

7. Sensor Cooling Dec. 06, 2013W. Bertl, PSI The ROC is in direct contact with the CFK base plate (base strips) which are glued to the cooling pipes. The pipe position on the CFK is such that the hottest part of the ROC (peripheral logic) is closest to the pipes. This design is very similar to the present Bpix except that the future layer 1 will be assembled using base plates instead of strips which will improve the thermal conduction. Tests with dummy modules improving thermal conduction by various methods have been done using water cooling (see next slide). Extrapolation to expected future conditions and CO2 cooling see overnext slide.

8. Sensor ΔT test Jan. 28, 2014W. Bertl, PSI Test setup similar as planned for layers 2-4, however standard CFK used with ~10 times worse heat conductivity than foreseen for final assembly (will be CFK K1100) Max. power in the test: 2 W/module Note: Layer 1 will have a 3 times larger contact surface than in the test W

9. Expected Sensor Temp. Jan. 28, 2014W. Bertl, PSI LayerPower scaling Surface scaling h scaling all scaling ΔTΔTTsensor (@-20˚) 13.531*1.1710-5 21.511*1.513-1.5 31.211*1.210.5-4.4 41.1511*1.1510-5.5 * This is a very conservative assumption...based on P=2W test result. CO2 temperature at -20˚C (-30˚C is possible) Note: At -2 deg. sensor current expected just below 50% of PS-limit. For 500 fb-1 @ -2 deg. expect that PS-limit of 20mA is hit.

10. Verifying Sensor Cooling Jan. 28, 2014W. Bertl, PSI Proposed actions within next months: Uni ZH will build a simple cooling loop with sufficient space to mount several dummy modules. Dummy modules consisting of CFK base plate (strips), a ROC bump bonded to a sensor and heated with thermal pads on top of the sensor will be build at PSI. Mounting those modules to the test loop using different techniques (w/o conducting paste, special glues,...) also done at PSI The assembled loop is planned to be tested with 2-phase CO2 at nominal cooling temperatures using the small CO2 cooling plant in Aachen.

11. Conclusion Jan. 28, 2014W. Bertl, PSI  Design of the cooling layout has been guided by calculations using the most advanced model specially developed for 2-phase CO2 microchannels (Thome et al). Results were confirmed by many measurements. (Cern, Aachen)  Extensive testing will be possible using the full scale replica resembling 50% of the final setup together with the full scale cooling plant prototype recently completed at Cern. Testing could start beginning 2014.  Building this replica has demonstrated that the full cooling system can be built and could stand the required pressure with large safety margins.  Sensor temperature expected to be sufficiently far below runaway temperature. This will be confirmed using a small test setup, however with full scale modules and pipes, in 2014.  The planned installation of new concentric stainless steel cooling pipes between the cooling plant and PP1 will allow temperatures down to -30 degree (at the BPIX entrance) which gives another safety margin for the sensor temperature.

12. Back-up slides Jan. 28, 2014W. Bertl, PSI

13. Jan. 28, 2014W. Bertl, PSI T & P along a loop ST-conv. ST-central slot Detector ST-return

14. Safety Issues Jan. 28, 2014W. Bertl, PSI Complete BPIX piping will be pressure tested to 160 bar. (Successfully done with cooling replica) Safety issues of the cooling plant see talk by Paola. Module temperature: On each z-side 6 cooling loops enter/leave the barrel section. We have enough lines available at the control cable (connector foreseen in central slot of supply tube) to equip each loop with 2 temperature sensors, mounted at the first/last module of the loop. An abnormal temperature reading could be used to shut down all (or selected) power supplies by DCS.