Cooling Working Group (CWG) Minutes #6

Date of the meeting:
16/04/1997
Place of the meeting:
40 R-B10
Present:
P.Bonneau, M.Bosteels, G.Benincasa, A.Carraro, G.Dumont, P.Fontaine, J.Godlewski, C.Gregory, P.Lecocq (Cie DATE), R.MacKenzie, P.Petagna, W.Van Sprolant

Agenda:

  • 1-Approval of the previous set of minutes.
  • 2-First considerations on cooling in power racks.
  • 3-Technologies and products for cooling in industrial electronic (Cie DATE).
  • 4-AOB.

    • Short Summary:
    The cooling system for the electronics racks will cover several hundred units and an industrial-type design study will have to be conducted relating to both technology and costs in order to minimise the prices. New more effective and pleasant systems do exist.
    The company DATE set out its expertise and experience in the cooling field and, in particular, with electronic boards. The various techniques (natural or forced convection, heat sink, conduction, cold plate with or without gap pad, etc.) were compared and their use explained.


    Detailed Minutes:

    As the minutes of meeting # 5 had been sent out only the day before, it was agreed to defer their discussion to the next meeting.

    M. Bosteels set out the broad outlines of what was to be the cooling system for the electronics racks in the LHC experiments.
    They would be located in 2 distinct areas which gave rise to different problems:
    - In the main cavern itself, hence in a closed area which was inaccessible with the beam running and subject to specific restrictions connected with radiation or magnetic fields, which would make it difficult to use standard fans or circulation pumps.
    - In the control rooms where the restrictions arose more from human causes, i.e. several hundred rotating fans gave rise to no negligible noise pollution for the people who sometimes "lived" in these rooms.
    By way of example, for ATLAS there were to be about 250 racks in the control room and another 250 or thereabouts in the main cavern.
    The cooling circuit for a rack could be divided into 3 parts:
    - The generation of cold and its conveyance to the primary exchanger in the areas (cooling sets, chilled water, etc.).
    - The secondary circuit between the primary exchanger and the (secondary) exchangers in the rack.
    In view of the safety and structural requirements, the most interesting configuration would be one secondary circuit with its control and monitoring unit for a group of 10 or 15 racks.
    - The transmission of heat between the secondary exchangers and the electronic boards within the rack.
    Traditionally, this exchange took place vertically by forced convection using finned exchangers fitted horizontally between the crates fitted with fans. The limiting factors here were, at stated above, noise, sensitivity to the strong magnetic field and also the dissipation capacity per unit area over the electronic boards. In addition, there was considerable heat dissipation into the environment, requiring the chambers to be air-conditioned.
    Another method involved arranging small cooling plates directly against the boards (direct cooling) in the crate and transferring the heat by conduction.
    An initial report on the tests made by ECP in this matter would be presented at the next CWG meeting.
    Finally, one of the most important items to be considered was the quantity of equipment involved; thousand of connectors represented thousands of CHF; a great deal of work had to be done on standardisation studies and drawing up a call for tenders to avoid prohibitive costs of the final systems.
    P. Lecocq presented the subject. DATE was a small firm specialising in the manufacture of thermal and thermodynamic equipment. It was capable of designing and producing any type of equipment intended for heat dissipation and cooling, e.g. heat ducts, cooling sets, heat sinks of all kinds, cryo-surgical equipment, exchangers, heat screens, furnaces, ovens, etc.
    Its design office possessed powerful software for the design, simulation and thermal analysis of components. Its workshops had all the tooling needed for handling metals and plastics -- a foundry, engineering, sheet-metalworking and welding equipment, a furnace, etc. -- and specific inspection facilities: survey, thermal, hydraulic, leak-testing, etc.
    With regard to electronics cooling, on the basis of its experience DATE had been able to draw up the (non-exhaustive) table below of the various methods depending on the mode and the fluid: Table 6-1

    The performances of the various methods, based on a fin-type heat sink, could be compared using the Table 6-2

    P. Lecocq then gave a practical explanation of the various techniques applicable to electronic board cooling and the results obtained. Many samples and manufacturing examples were shown. The choice depended, of course, on the heat flux released by the board and the desired operating temperature. The whole is summarised in the graph below (the cold source is at 20°C and the heat flux assumed to be uniformly distributed throughout): Figure 6-1

    The graph clearly showed that the heat seal (gap pad) between the cold plate and the board was very important. All commercially available products were based on more or less filled silicone to improve the final conductivity, which varied from 1 to 2 [W/m.K] for paste or semi-rigid products and rose to 10 for rigid duplicate mouldings.
    The problem of the permissible temperature on an electronic component and where it should be measured was also raised. 80°C was generally accepted on the component casing, but the temperature inside the component was unknown.
    DATE's full design study on electronic board cooling and a file on its product range were available.
    The next meeting is to be held on Wednesday 21/05/97, Building 40, room R-A10 from 10 a.m. to 12 noon.
    Agenda:

    CERN/P.BONNEAU/30/05/97