Minutes Cooling Working Group (CWG) #3

Date of the meeting:
19/02/97
Place of the meeting:
40 R-A10
Present:
P. Bonneau, M. Bosteels, M. Bozzo, A. Carraro, G. Dumont, R. Gregory, S. Jääskeläien, R. MacKenzie, T. Niinikoski, A. Onnela, B. Pirollet, A. Placci, R. Principe, S. Sergueev, W. van Sprolant. J. Ylöstalo, E. Zoubarev.
Agenda:

Short Summary:

Details of the various sources of the heat from the ALICE and CMS (tracker) sub-detectors were given with their power and specified temperature.
The choice of cooling systems had already been made for certain detectors among the range presented at the previous meeting: evaporative, pressurised or under-pressure liquids and binary ice.
The choice of exchangers and pipework was more uncertain.
A very accurate system of temperature acquisition by Pt 100 which was applicable to the cold parts of the experiments was presented. In more general terms, temperature acquisition in the detectors would involve several thousand sensors and could not be dealt with in the conventional manner.

Detailed Minutes:

It was regretted that there had been no specific diagram for the binary ice circuit, whereas there was one for the evaporative systems. Those were obviously only theoretical circuit diagrams and binary ice could theoretically be used in a liquid circuit.
Still with regard to binary ice, it was pointed out that if the use of an evaporative system in a large physics detector had yet to be demonstrated, the same applied to a binary ice system. Evaporative systems were in use for cooling the electronics of a super-computer, while binary ice systems existed in industrial cooling applications like cold chambers.

Only ALICE and CMS were presented.

C. Gregory first of all announced that ALICE had received official approval. He then presented a summary table of all the heat sources in the experiment, with everything relating to the magnet -- ~4 MW for information, but that was not covered by the CWG -- and the various detectors and their associated electronics.

Unit
Power [kW]
Operating temp. [°C]
Present assumption on system
Counting rooms
800
15
.
Racks bunker
??
15
.
Racks UX25
??
15
.
Inner Tracking System: Pixels, Si Strip
4.5
20
.
Inner Tracking System:
Silicon Drift Detector
2.3
20±0.1
Evaporative cooling
Time Projection Chamber
25
25
Lower pressure
Pestov
150
20
.
Photon Spectrometer
5
-25
Binary ice
Muon Tracking Chambers
20
20
.
It should be noted that, unlike the case for ATLAS and CMS, the silicon detectors were to operate at ambient temperature, owing to the much lower radiation level expected in ALICE.
A presentation was then made of some of the ancillary systems in which the runs of the gas and cooling pipework had been simulated, showing that they essentially obeyed the rules discussed at meeting #2. There nevertheless remained some delicate points, like the passage through the magnet doors and access in the tracker.
A. Onnela began by pointing out that he was speaking on behalf of the CMS Tracker and that the teams were working on other detectors, especially in the calorimeter. M. Bozzo had been appointed to try to co-ordinate work on cooling in CMS.
The characteristics presented were:
.
Silicons + Pixels
MSGCs
Quantity of heat to be removed and from where
~10 kW (Si) + 4 kW(Pixels)
-70% in FE-electr.
-10% in Si wafers
-20% in cables
~10 kW barrel + 14 kW Fwd
-80% in FE-electr.
-20% in cables
Operating temperature [°C]
max. -5
18
Coolant temperature [°C]
~ -20
~13
Gradient admissible for the detectors
±2
±5 (may be less)
Thermal stabilisation requirements
A few degrees, because:
avoid deforming structures, optical modulators
A few degrees, because:
avoid deforming structures, optical modulators
Need for thermal screens
-Needed towards MSGCs
(-gas flow towards beam-pipe)
-Between Si and Pixels while mounting Pixels?
-Needed towards Si
-Towards Preshower and ECAL?
Present assumption on the systems
-Water + Ethanol for Si
Overpressure
-Binary ice for Pixel
(liquid as back-up)
Overpressure
-N2 flushing
-Demineralized water
Overpressure
(underpressure would be used if possible)
-N2 flushing
Pipe material
-Al for Si barrel
-Stainless steel for Si fwd
-Al for Pixel
-Stainless steel for barrel
-Al for fwd
Size and position of collectors
?
?
Any other important point
Need for insulating supply pipes (avoid condensation)
.
That table gave rise to a few questions and items in suspense and A. Onnela had drawn up a summary table:
Another matter for concern was the succession of detectors operating at different temperatures since there again appeared to be regions below ambient temperature after the Si/Pixel at -10°C and the MSGCs at 18°C. .
A. Onnela then presented the MCGC drawers with their stainless steel cooling pipe and the layout of prototype B1 with the distribution manifolds. Each drawer was supplied in parallel and the load losses were balanced by the design of the manifolds.

The presentation could not be made owing to the absence of H.P. Gugerli.

T. Niinikoski and J. Ylöstalo presented a temperature measurement system using RTDs (Resistance Temperature Devices) initially developed for cryogenics. A technical note was available (CERN-PPE/96-56).
Resistance temperature measurement had several advantages: the low cost and small size of the sensors, great sensitivity and rapid response.
The main resistance sensor types included metallic, metal oxide and semiconductor elements, in the geometry of bulk, wire or thin film. The most popular was of pure platinum calibrated at 100 ohm for =°C (Pt 100) and, according to the tests made, it was the most suitable (with the RhFe sensors) at the magnetic fields and radiation levels in the LHC detectors.
Its greatest drawback was self heating: with a Pt 100 and an industrial converter (5 mA), the ÆT was 0.125 K in water and 1 K in air.
Without going into excessive detail, the acquisition system developed was applicable to any kind of RTD and its characteristics were:
There was one remark to be made on the final item: low cost to the cryogenicists but very expensive to people working in cooling for electronics.
By and large, finding the temperatures needed for thermalisation and alarms in the detectors was a field requiring closer examination. There was a need for several hundred sensors, for example, in the trackers and it was clearly impossible to connect them in the conventional way. They would use either the transmission system of the detector itself (was provision made fore that by the electronic engineers?) or the sensors would have to be "intelligent" and use a databus.
That discussion in fact covered all the problems of slow control and should be conducted at specific meetings.

In the discussion on the material of the cooling pipes (Al vs. stainless steel), it was proposed to consider the cupro-nickel alloy. It was apparently possible to obtain finished products (pipes) in the same sizes as Al and stainless steel made of that alloy which was highly corrosion resistant. More information was needed, especially concerning the radiation length, in order to complete the picture.

As A. Onnela had stressed in his open questions, the selection of the right material was a real problem and it was currently clear that an R&D programme had to be started, particularly in order to determine the permissible minimum thicknesses in relation to the material, the fluid and corrosion and to find realistic connecting (and disconnection) methods (welding, brazing, adhesive securing, crimping, unions, etc.).

All the characteristics of the water/glycol (ethylene and propylene) mixtures between -40 and 100°C were available (thanks to R. MacKenzie).

The next meeting would be held on Thursday 6/03/97, Building 40, Room 2-A01 from 10 a.m. to 12 noon.
Agenda:

CERN/P.BONNEAU/30/05/97