From:	SMTP%"A.A.Carter@qmw.ac.uk"  1-NOV-1997 06:42:15.63
To:	ATLAS-SCT@atlas-lb.cern.ch
CC:	
Subj:	Thermal Conductivity Measurements of TPG and PG


_______________________________________________________________________________
|                                                                             |
|            Thermal Conductivity Measurements of TPG and PG                  |
|            _______________________________________________                  |
|                   GA Beck, AA Carter and J Morris                           |
|          Queen Mary & Westfield College, University of London               |
|                                                                             |
|   In preparation for construction of thermal and electrical SCT modules we  |
| have carried out thermal conductivity measurements on a limited set of TPG  |
| and PG samples that relate to candidate materials for both module           | 
| baseboards and possible hybrid substrates.                                  |
|   The thermal conductivity measurements have been made with TPG/PG samples  |
| in the form of 10mm wide x 80mm long strips. Each strip was measured by     |
| clamping it in turn to a 0.3mm thick copper foil base, equipped with a      |
| resistive heat source at one end and four thermistors at 15mm intervals     |
| along its length. Thermal grease was necessary near the ends of the sample  |
| to ensure good thermal contact with the copper.  The cold end of both strips|
| was clamped in a heat sink, and the strips were surrounded by 3mm of foam   |
| insulation and by a copper guard, maintained close to the strip temperature |
| at each end.                                                                |
|   Measurements were made with a temperature difference of about 10 degrees  |
| between the ends of the strip.The error in the derived thermal conductivity,|
| k, due to uncertainties in the temperature gradient, cross-sectional area   |
| and heat input, was +-4%.  The error in correcting for conduction by the    |
| parallel copper foil was negligible (this procedure was tested using a      |
| sample of aluminium of known conductivity). Corrections (less than 2%) were |
| made for heat loss through the insulation and guard structure.              |
|   The samples of particular interest were:                                  |
|  (1) TPG: a sample of material currently under investigation as a prototype |
|          silicon module baseboard. The thermal conductivity of this sample  |
|          has a clear negative temperature coefficient.                      |
|  (2) PG700:  This material is mechanically superior. It is misleadingly     |
|          named, since our measurements confirm k is well below 700W/mK (as  |
|          indicated also by the manufacturer).                               |
|  (3) PG1300: A material with properties some way between the above two.     |
|    In summary, our results show:                                            |
|                                                                             |
|              Sample:  Thickness    k at 20C     Temp.Coeff.                 |
|                         (mm)       (W/mK)       (1/k)*dk/dT                 |
|        ______________________________________________________               |
|        | (1)  TPG    | 0.40   |  1705 (+-70) | -0.4(+-0.2)% |               |
|        ______________________________________________________               |
|        | (2)  PG700  | 0.25   |   330 (+-16) |   0 (+-1) %  |               |
|        ______________________________________________________               |
|        | (3)  PG1300 | 0.53   |  1360 (+-55) | -0.4(+-0.3)% |               |
|        ______________________________________________________               |
|                                                                             |
|   A more complete write-up of this work will become an ATLAS SCT note,      |
|   and will include details of the apparatus and an overview of the current  | 
|   understanding of TPG/PG thermal properties, including the affects of      |
|   encapsulation and review the availability of such materials that could    |
|   be appropriate for use in SCT modules.                                    |
|                                                                             |
|_____________________________________________________________________________|