LHCb FT SiPM Cooling workpackage

Mandate

As defined at the Heidelberg Workshop (April 2013) ( summary): PG advises, simulates. Imperial: engineering design and construction.

Tasks:

• Assess and measure heat load
• Demonstrate long operation, cycling, condensation
• Thermal cycling to demostrate mechanical robustness, absence of leaks etc

Events

• 17 October 2013 SiPM Cooling at the Workshop (Indico link)

Documents

• The draft specifications document (sorry, still a draft, still evolving, many pieces still to be incorporated!): PDF v. 1.1 (August 2013)
• CFD simulation Project Proposal (link to EDMS document).

Status summary:

05 July 2013

• Chilled air cooling lab has been set up. A CTU student is working full-time on the air cooling tests. 4 different options of a cooling pipe will be tested: 3 round tubes with and without inner structures, a rectangular multi-finned radiator (baseline). Currently, the compressed air conditioning is being upgraded.
• Full-scale mockup of a single end-cap, with 6 cm outer plug and ~13 cm support plate fragment has been constructed. It will be used mostly for the air-cooling tests.
• The simulation work was paused since mid June, upon completion of the heat load study for the generic end-cap design. The result: 7.5 W per module side. The work will be resumed, once the consensus on the end-cap configuration has been achieved. The intermediate report will be ready shortly.
• Preparation for the liquid cooling (2-phase) tests is ongoing. The tests will be performed at the TOTEM C3F8 -based facility. Mono-phase liquid cooling test setup is being designed (2-stage, with a -10C pre-cooler and a Peltier-based main cooler, down to -50C).
• Useful links with the industry have been established in May-June: Institute of Thermoelectricity (UA) - specialized Peltier coolers, Thermacore Europe (UK) - low-temperature heat-pipes, composite heat-spreaders.

07 August 2013

• The work on the air cooling tests continued in July. 4 different options of a cooling pipe have been tested with a full-scale end-cap mockup #1. The compressed air conditioning was upgraded by adding an industrial air dryer. The report is being prepared by the CTU team (expected date: August 10).
• 2nd generation mockup was designed for the upcoming 2-phase cooling tests, using the concept discussed within the end-cap working group. The parts have been made, but not assembled, yet.
• The simulation work is largely completed. The latest end-cap design has an estimated heat load of 5.7W per module side. In addition, promising results were obtained for "the cooling-though-flexPCB" option (to be discussed with the SiPM community). The summary report is under preparation.
• The liquid cooling (2-phase) test bench at the TOTEM C3F8 -based facility has been prepared. The initial measurements, using a special thermostatic enclosure with a controlled thermal load, began this week.
• Mono-phase liquid cooling test setup will be based on the the FPW-91 Julabo chiller, already delivered, to be installed at Point 8.
• Negotiations with Thermacore Europe Ltd: an offer to design and manufacture prototype low-temperature ammonia-filled heat-pipes was received. Due to a high price, this option had been put on hold. Solid metal heat spreader with external Peltier chiller is a viable option for a BGV-like module design (discussed with Massi and Fred, TBD with Guido)

04 September 2013

• 2-phase cooling tests started at the TOTEM test bench at H8, initially with a specially designed isothermal box containing a real cooling pipe (evaporator), a simulation of the end-piece and SiPM array, as well as a heater simulating a heat load of about 10W.
• Results on temperature uniformity obtained, with different pressure differentials.
• 2nd generation mockup , implementing the end-cap concept adopted in July, has been constructed and installed at the H8.
• The JULABO chiller for single-phase liquid tests installed at Point 8, in the 2885/2-201 lab
• A smaller new JULABO chiller F81ME /HL, intended for D-module cooling, will be ordered shortly

10 October 2013

• 2-phase cooling tests continued with the new mockup, completed by September 16.
• Two different insulation wall versions were tested. The total thermal load of the mockup was estimated as ~20W (including the external capillary and loosely insulated connection pipes), by observing the "dry-out" effect at the lowest temperatures at dummy SiPMs (down to -45...-48°C). Under these conditions, the C3F8 -based system had to be operated with a significant underpressure on the suction side (0.42 bara).
• The reports about summer tests have been prepared by the CTU group
• A new round of thermal analysis: evaluation of local parameters of a 2-phase cooling scheme (capillaries, mass flows, thermal cycle)
• First series of single-phase liquid cooling tests with the (borrowed) Julabo chiller: completed by October 1.
• Ordered:a JULABO chiller F81HL, intended for D-module cooling; the delivery is expected this week.
• Preparation of the SiPM cooling workshop of October 17

Thermal simulation

A CFD simulation Project Proposal (link to EDMS document). This project, joint with EN/CV, was intended to evaluate the heat load of the FT module end caps and give some insight for various prospective cooling options. Author: Enrico da Riva (CERN, EN/CV/PR).

• all intermediate reports are available at the Project's DFS site, under Presentations
• 2013-08-01: Addendum to 22/07: cooling through flex PCB.
• 24/07/2013: latest end cap simulation results ("July" geometry). 5.7 W/module
• 2013-06-10: 3D with Alu + edge insulation; multiport cooling pipe. 7.1 W/module
• 2013-05-30: Alu jacket; liquid and air cooling estimates. 7.1 W/module
• 2013-05-27a: First 3D results, more on liquid coolants. 3D≈2D
• 2013-05-14: End caprefinements, liquid_coolants. 6.3...8.5 W/module
• 2013-04-22: 2nd_round results: try to minimize the heat load. ~6 W/module
• 2013-03-13: 1st round results: heat load with a simplified geometry. ~10 W/module

Bottom line: in the ideal case (with no inlet/outlet tubes, perfect thermal contacts) the heat load per module should be less than 10W. The T drop between the cooling pipe and the Si die strongly depends on the stiffener material. For FR4: ~7° (~5.7 W/module); for carbon: ~3° (~6.3 W/module)

Cooling options

• Liquid cooling (mono-phase or 2-phase) - CERN, EN-CV group
• C6K coolant
• Air cooling - CERN, CTU
• Hybrid options
  • Air, with pre-cooling - CERN, CTU
  • Liquid/thermoelectric - CERN, ITE
  • Heat pipe/thermoelectric - CERN, Thermacore (, ITE)
  • integrated TE cooling if SiPMs (a micro-Peltier module in the SiPM package) - proposed by ITE, currently this option is not pursued.

Schedule

(Overview of main activities, transient activities are not mentioned). Completed activities are marked in green, ongoing – in brown, scheduled – in black, critical path -- red).

May

• 1^st round of thermal CFD analysis: head load and temperature profiles, mostly 2D (PG, EdR)
• A feasibility study of the single-phase liquid cooling option (infrastructure) (PG, Julabo)
• Order materials for the cooling mock-up⁽¹⁾, design of the 1^st mockup version (PG)
• Set up cooperation with the CTU⁽²⁾ group (air and evaporative cooling options) (PG, RL, VV)
• Set up cooperation with ITE⁽³⁾ on the Peltier cooling option (PG, ITE) – by end-May
• Prepare the air cooling lab facilities (PG, GD)

June

• Contacts with Thermacore Europe: "heat pipe" cooling option - weeks 1,3
• 2^nd round of CFD simulations: 3D, air cooling, heat load evaluation (EdR, PG) – weeks 1-3
• Start of the air cooling tests – preparation phase, calibrations etc (PG, VV, MD) – June 17
• Construction of the 1^st mock-up (GD, PG) – by end-June

July

• Draft report on the thermal simulation project (EdR, PG) – week 1-2, pending
• 3rd round of CFD simulations: new end-cap design, 2-phase cooling (EdR, PG) – weeks 2-4
• Air cooling tests, main phase (ME, PG) – week 1-4
• Liquid cooling tests (PG) – weeks 2-4, shifted to September
• Design and construct the 2nd version of the cooling mockup (PG, TS, GD) – weeks 3-4
• Preparation phase for the evaporative cooling option tests at H8 (VV, MD, GD) – weeks 1-4
• Intermediate report on the air and liquid cooling options tests (PG) – end of July

August

• Petr is away (vacations) – 2-10 August, TBC
• Final air cooling measurements (ME) – weeks 1-2
• Liquid cooling measurements with Julabo chiller(PG) – weeks 3-4, shifted to September
• Start of 2-phase measurements at H8 (MD) – weeks 1 (or 2)
• Possible follow-up of the joint ITE project on Peltier cooling (PG, ITE) – about end-August – mid Septenber
• Draft summary report on the SiPM cooling, choice of the cooling option for the Demo (PG) – end August- week 1 of September

September

• Final 2-phase cooling measurements (MD, PG) – weeks 1-2
• Final report by the CTU group (VV, MD, ME) – weeks 2-3
• Integration with the Demo module (PG, TS, all) – weeks 3-4 - postpopned
• Work on the TDR section about the SiPM cooling (PG) – postponed
• Liquid cooling measurements with Julabo chiller(PG) – weeks 4
• Possible follow-up of the heat-pipe joint project with Thermacore Europe - on hold

⁽¹⁾ A full-scale model of the FT module end cap, intended to test various SiPM cooling options
⁽²⁾ Czech Technical University, Prague. Group leader: Vaclav (Vic) Vacek
⁽³⁾ Institute of Thermoelectricity, Chernovtsy, Ukraine. Director: Akad. Prof. L.I.Anatychuk

Acronyms

TBC – to be confirmed

CTU – Czech Technical University (Prague)
EdR – Enrico da Riva (EN-CV-PJ)
GD – Gerard Decreuse
ITE – Inst. Of Thermoelectricity (Chernovtsy)
MD – Martin Doubek (CTU)
ME – Miroslav Erben (CTU)
PG – Petr Gorbounov
RL – Rolf Lindner
VV – Vic Vaceck (CTU)
TS – Trevor Savidge (IC)

Attached Documents

• SiPM_cooling_specs_1.2.pdf: Draft specs for SiPM cooling

• NW-CTU-for-LHCb-Aircooling-report_2013.pdf: CTU Long write-up: SiPM cooling (vortex tubes)

• NW_Final_ReportLHCbEvap_02.pdf: CTU Long write-up: SiPM cooling (C3F8 evaporative)

• Summary_2013.pdf: A summary of SiPM cooling tests in 2013

• Proceedings_03.07_submitted.pdf: Presentation at TIPP'2014 (SIPM cooling...)

• F.Tuma_c6K_as.....pdf: F. Tuma (3M Company) "Fluoroketone C2F5C (O)CF(CF3)2 as a Heat Transfer Fluid for Passive and Pumped 2-Phase Applications",full PDF version, difficult to find

• Novec_Memo.pdf: Project: Novec 649 as a replacement for C16F14 in liquid cooling systems