From: KEKVAX::UNNO "Y. Nobu Unno, unno@kekvax.kek.jp" 26-SEP-1997 11:50:41.63 To: ATLAS-SCT@ALTAS-LB.CERN.CH CC: UNNO Subj: Minutes of module meetins of September 1997 Dear collaborators, This is the minutes of the Module designs meetings held on Sep 4th and 5th, and follow-ups. If you have comments/corrections, please let us know. Regards, Nobu & Tony *********************************************************************** Minutes for the SCT modules meetings at CERN, September 4th and 5th (and follow up on Sept 10th) The main meetings, on the 4th and 5th September, consisted of presentations and discussions. The presentations were : Review of the constituent proposals from UCSC meeting - Carter Report of forward "informal" meeting on 2nd Sep. - Moorhead Summary of (detail) issues to be addressed - Carter RAL module design update - Apsimon KEK module design update - Terada A copy of the materials presented in the meetings is available from the ATLAS secretariat. Apart from the main meetings on 4th and 5th, there were informal meetings. One involved a small set of representatives of the forward modules on 2nd September and the other was a follow-up technical meeting of the barrel module work, on 10th September. A summary of the September 2nd meeting and the ad-hoc discussions concerning some forward module activities was presented by Gareth Moorhead. A brief note describing these is appended to these minutes; and it is expected that the forward community would be considering a future programme plan in discussions at CERN during the ATLAS week that followed. The remainder of the sessions addressed mainly with Barrel module issues, and below their main contents are outlined with reference to the items in the "Summary of issues to be addressed" [proposed decisions are given where appropriate]: * Stereo formation : +-20 mrad in the module, and the barrels to have fixing locations that provide the appropriate rotations for alternating 40mr layers. * Cabling interface : the methods of attachment(including a glue contact) were discussed, and the need to be consistent with barrel mounting brackets was accepted. The length of the module pig-tail cable was discussed and provisionally set to be slightly less than 30 mm from the hybrid (the connector outer edge on the pig-tail will be 60 mm from the centre-line of the detectors). * PG : optimizing its uses both thermally and mechanically were discussed, with a suggested plan to purchase PG sheets of PG700 and of PG1300. ==> The situation concerning PG was clarified and modified at a subsequent discussion with a representative from Advanced Ceramics. It was agreed that PG1700 is the only guaranteed product for the module baseboard construction, and both thermal and electrical modules will use PG1700 (barrel in 380um thick for baseline operation at up to 300V, and most probably 500um thick to produce an optimised module capable of 500V operation). PG1700 pieces will be ordered of size 100mm*100mm for barrel and 150mm*75mm for forward. PG1300 of BF-Goodrich is being investigated as an alternative. * Glue : it was agreed to define the positions and shapes of areas of contact in detail, and to use Araldite 2011 with boron nitride filler. There would be 3mm of stay-clear around wafer edges, and no adhesive around baseboard edges. * Cooling contact with pipe : method and area to be optimised Present design uses an oval pipe (Barrel) of a dimension of 2 mm (height) x 9 mm (width) (round end of 1 mm radius). The length of the "ear" of the module to be 60 mm. * Module handling : pickup and positioning features to be incorporated. Two fixing points (holes) in the cooling "ear" and one point in the non-cooling ear. * Barrel hybrid : the beramic diagnostic version of Carl to be the baseline version for electrical modules, with kapton beryllium and kapton PG being worked on as possible future options. Thermal modules to have "final" dimension hybrids, and to be designed if possible to have dismountable hybrids so that the different low-mass options could also be studied with a given basic module. The cooling contact area will be kept the same area as that for the final design for both "(diagonostic) electrical" and "thermal" modules, but the fixing points of the "(diagonostic) electrical" will have a wider spacing. * Thermal and Electrical interplay : compromises for overall performance were discussed, and current barrel module is to have beryllia stiffening plates on both edges for mechanical rigidity, and to aid electrical isolation at the cooling contact. * Mechanical integrity : It was proposed to have the PG1700 baseboards coated with 20um of polyimide on each surface to both improve mechanical strength and guarantee electrical isolation at the edges. Thermal cycling tests of such coated pieces are underway. * Thermo-mechanical modules and measurements : It was agreed to use the KEK irradiated mechanical grade detectors (8 detectors) in constructing the fundamental barrel thermo-modules with PG1700 baseboards(380um thick for 300V operation, and [most probably] 500um thick for 500V operation). Optimization details (mainly PG baseboard geometry) for both types to be determined after Taka's runaway studies are complete. Distortion measurements are to be made, initially by visual means, and hybrids of Beo, Be and PG will, if possible be interchanged on these modules (or on separate mechanical modules). Optimization - by Oct. Module fabrication - by Nov. Thermal runaway, thermo-distortion measurements - by Dec. * Safety factor on the thermal runaway : To design to an implicit safety factor of 2 against the thermal runaway specification in the module design appeared as intuitively being a sensible target. It is proposed to specifically define the safety factor as: runaway at twice the heat flux at the anticipated upper bound of fluence, while ensuring -10 deg.C at the module cooling contact. Generally the factor 2 was accepted but it was agreed to provide further justification for this value. * Engineering drawings : RAL (D.McPhail) will make engineering drawings of the agreed barrel modules : one for "final" hybrid and one for "diagonostic" hybrid. ___________________________________________________________________________ Some notes from an ad-hoc meeting on Sep. 2 at CERN (Tim Jones, Hans-Gunther Moser, Gareth Moorhead, Tony Carter, Nobu Unno) An overview of the Forward Module Programme ___________________________________________ At the UCSC SCT meeting there were updates on the FORWARD MODULE programme, in areas of design concepts and the study of thermal runaway, with a concensus emerging in both : The TDR design forms the basis of the current approach, with a central spine baseboard that forms both the mechanical structure for the wafer assembly and the extraction path for subsequent detector heating. Details on the far-end cooling contact and location are still being considered, as well as an optimised geometry in the inter-wafer connection area. The double-sided end-tap hybrid design is well established, and detailed work is underway on optimising (and maximising) the cooling contact area, and providing a design that thermally isolates the module wafers from the heat dissipated by front-end chips on the hybrid, but which allows the thermally conducting spine to transmit heat efficiently to the cooling contact on the hybrid. Quartz pitch adaptors provide the mechanical and bonding link between the hybrid and the wafers. Thermal studies, both simulations of temperature profiles and runaway scenarios are continuing, as well as the preparation of thermal modules. The need to improve the thermal interfacing between the spine and the cooling point, and to reduce material if possible, has led to TPG being the proposed spine material (but a programme of work in this area is needed before establishing a final design). Some specific studies now underway include: In the UK and Munich, forward groups are now constructing, and will test, both thermo-mechanical and electrical modules based on the TDR design. This will allow components already in-hand to be used; electrical modules will be of the outer ring type, with thermal modules, in principle, being of each ring type. These thermal modules, capable of two-point cooling, will be tested as single units for module-specific activities and will also be mounted and tested on prototype wheel sectors at RAL and possibly NIKHEF. Electrical modules will be operated as single units and subsequently in multi-module configurations, and will incorporate the new electronics, with the next generation of the double-sided hybrid being compatible with both CAFE/ABC and ABCD. The aim is to have complete sets of thermo-mechanical and electrical data characterizing these modules by December. The recent development of modules incorporating TPG, based on the TDR module concept, has been pursued at MPI/NIKHEF and in the UK. Preliminary ideas were presented at the UCSC SCT week, where designs were shown that improve the thermal conductivity along the length of the module and need only a single cooling contact at the hybrid end. The details of these designs are in the process of refinement with the aim of developing a unified design within the next few weeks. In parallel with the current prototyping of TDR-type thermal modules, it is expected that first thermal prototypes of the TPG-type module will be constructed and tested. The timescale for the construction of the first of these modules is ~ 2 months. It is unlikely that TPG modules will be fully developed by December this year. The hybrid for the TPG modules will be the same as that for the TDR modules except for a difference in geometry that relates to the profile at the interface to the silicon and hence does not affect the thick-film part. Sufficient BeO has been ordered to allow the manufacture of several hybrids of both the TDR and TPG-type modules. If the TPG prototyping is successful it will be used to either eliminate the second cooling contact or to reduce the radiation length for two-point cooling depending on the engineering merits of the two cooling contact concepts. This programme requires that a suitable mounting block can be designed to provide the required cooling capability. Several groups (KEK, Australia) continue to investigate the feasibility of applying a modified version of the barrel centre-tap design in the forward region, looking at possible layouts and disk structures with reduced material. A second single-sided centre-tap electrical prototype constructed at Melbourne is currently being tested at H8. Further modules including thermal/mechanical prototypes will be constructed as parts become available. ================== RFC 822 Headers ================== Return-Path: AAC@vxcern.cern.ch Received: by kekvax.kek.jp (UCX V4.0-022J, OpenVMS V6.2 VAX); Fri, 26 Sep 1997 04:47:24 +0900 Received: from dxmint.cern.ch by kekux.kek.jp (8.8.5/rewrite from KEK-Master/1.0c/97.2.21) id EAA10353; Fri, 26 Sep 1997 04:46:32 +0900 From: AAC@vxcern.cern.ch Received: from vscrna.cern.ch (vscrna.cern.ch [137.138.28.123]) by dxmint.cern.ch with SMTP id VAA31570 for ; Thu, 25 Sep 1997 21:46:00 +0200 (MET DST) Date: Thu, 25 Sep 1997 21:45:39 +0200 To: UNNO@KEKVAX.KEK.JP CC: AAC@vxcern.cern.ch Message-Id: <970925214539.21c0b8ae@vxcern.cern.ch> Subject: draft minutes of modules meetings of September 1997