For questions contact Phia at phia@stanford.edu 

Background Information

When assembling the inner tracker (ITk) testing is performed at each step of the way to ensure that the sensitive part of the detectors, "modules," work. When in final assembly onto the outer quarter shell the ring modules on the endcap will need to be tested to ensure no loss of functionality in the loading process. However, because of the size of the quarter shell assembly, this cannot be done in a cooling chamber so these tests will be done without coolant. A big concern here is the heat output of the modules raising the temperature past the safe temperature range for the modules. 

The current solution is to do the testing in low power mode where most of the functionality, but not all, can be checked. To test the feasibility of this proposal, a mock ring with dummy heaters to simulate the power from the modules is used to test the heat output. The heat output across the ring can be read via 'RTDs,' or resistance temperature detectors. 

Additionally, on each ring, there is an outer type 0 or ``rigid" ring that provides power to the ring but traps heat and limits external cooling. In this test, due to practical limitations of the availability of a metal type-0 ring, a plastic 3D-printed type-0 ring is used instead. The difference in thermal properties must be acknowledged when interpreting the results. Additionally, due to the height of the RTDs on the heaters, the ring is placed further from the heaters, at a height of 0.17 inches rather than the planned 2.4 mm (0.094 in). 

Terms

RTD = resistance temperature detector

Readout via Raspberry Pi & Grafana

Heaters powered in serial power (SP) chains

low power (LP) mode

Setup

Images: The smaller fan used (left) and the larger fan (right) are shown in the same position relative to the ring. 

RTD Position

The position of the modules or dummy heaters can be found on the confluence webpage Dummy Heaters. The 'T' or 'S' denotes triplets or single modules at the inner ring 'L0' whereas the 'Q' denotes the quad modules at 'L1.' The 'U' is to indicate the cooling tubes up position and the 'D' for cooling tubes down. 

 Powering:

The modules, as well as the heaters, are powered serially in 'serial power' (SP) chains with L0 separate from L1 and the up and down sides of the ring separate as well. This is done to reduce the nonsensitive material within the detector. When doing the low-power testing, while it would take longer, it is possible to separate each side L0 and L1 to reduce the heat at any given time.

**missing Q20 due to broken power connection

***too fragile a connection to heater 12

The normal configuration would have 9 triplets and 10 quads.

We verified with a thermal gun that the power is distributed ~evenly and all the modules were checked beforehand to verify that they are roughly the same resistance and that the connections are good. More documentation can be found here: https://docs.google.com/spreadsheets/d/1s-4BwTW910EQzInhqSd7pgCI3uvk52Jl/edit?gid=2064913736#gid=2064913736 under the '19-0 July/August 2024 testing' tab.  

RTD readout

RTD readout code that was used in the pi has been copied here: https://github.com/phiamorton/SLAC-ATLAS-ITk-summer-project/blob/a693582498f9b6a66e0cd34a3b4696bee8363c64/RTD%20readout (if you actually want to edit how the RTDs are read out you must go into the pi and edit 'rtd-rpi/python/trd_test.py' I have simply copied it here for your viewing convenience). 

Using the pi: 

You must be in the clean room environment to access the pi

The pi is named pi@dcs-pi-lp and the password is dcs-pi-lp

To run the readout go to rtd-rpi/python and run rtd_test.py 'python rtd_test.py'

Grafana Interface: 

username ='atlas'

password = 'AtlasItkDataBase'

database = 'lab'

You can find the interface here http://atlascr.slac.stanford.edu:3000/d/UZDO0trIz/rtd-dummy-ring-test?orgId=1&refresh=5s or under the 'RTD dummy ring test' within the ATLAS clean room Grafana. 

Results

The results can be seen in this PowerPoint presentation and linked under 'presentations,' I would have inserted images directly with a caption but Confluence would not let me. 

Comparison to prior results

We determined that we cannot compare our results to the results from the ITk Outer Endcap group at LNF https://indico.cern.ch/event/1353986/contributions/5866686/attachments/2823320/4931112/LP%20mode%20testing%20for%20integration%20and%20LLS.pdf. The reason is that we are powering only one side of the ring at a time whereas they are powering both sides at once, making their temperatures hotter. This is likely why our temperature shows a plateau around 50 C even without fans whereas theirs does not. 

Additionally, they had a much larger variation in the resistances of the heaters they used. Their fans are cooling from the top of an enclosure whereas our cooling has been tested in the open with fans at a variety of angles. 

Presentations

7/18/24 Introduction talk:

8/22/24 update:

8/27/24 preliminary results:

8/28/24 final results:

Equipment 

Link to the smaller fan used https://www.amazon.com/Intel-E97379-003-Connector-Aluminum-Heatsink/dp/B01LWMGRX6 

”82.9 CFM 10%”  https://www.panacompu.com/panama/en/product-information/intel-e97379-003-air-cpu-cooler-120mm-fan-1000-3000rpm-no-rgb-32dba.

The Honeywell HT 900 fan has no documentation for the lower speed, but the “Maximum Cubic Feet per Minute (CFM): 185 CFM” (https://newboroughvillage.co.uk/uk5/honeywell+turbo+fan). 

the pi hat https://www.robotshop.com/products/sequent-microsystems-rtd-data-acquisition-8-layer-stackable-hat-raspberry-pi
And the connectors https://www.digikey.com/short/m5987tbp