Minutes of the Tracker Technical Meeting
April 11, 2001
In connection with the mass budget, some concern was expressed about converter foils. The existing thick foils have such poor tolerance on thickness that they could spoil our mass budget if they came out on the thick side. Ossie will continue to look for a vendor with better tolerance. Tom will email to Sandro information on the vendors used so far: Schwarzkof Technologies (thick) and CSM Industries (thin). CSM has long lead time for the thicker foils, which is why they were used only for the thin. Sandro is looking at the vendor of Agile foils, which are 100 +- 15microns. He said that their process for foil production changes at 0.5mm thickness. He is waiting to hear if they can improve the tolerance on the thickness.
Robert remarked that everybody should read Ossie’s report on the results of the SLAC thermal tests of Si bonded to aluminum with silicone dot patterns. (http://scipp.ucsc.edu/groups/glast/mechanical/Dot_Pattern_Report.pdf) The question was raised as to whether it should be repeated with primer. Gwelen remarked that the primer may be more important for the conductive silicone adhesive. It seems to lack bond strength without the primer, and the primer is reputed to enhance the conductivity. Tom said that they would be going with the 39 dot pattern.
Erik reported that he had received the face sheet material back from Italy the previous day. Hytec will ship it back out today. He will hand carry the 4-ply face sheets and carbon-carbon material to Italy on the 23rd. He had a question on the MCM mounting and the grounding screws. He is planning to thread directly into the CC for the grounding screws. They did pullout tests with 1.6 mm fasteners and got excellent results. A few threads could support 290 N. Lots of threads supported more than 600 N. Tom asked why we could not eliminate all the other inserts for the MCM, since it is light and will also be held by sticky tape. Parylene will coat the threads and prevent carbon dusting. Robert asked how the conductive attachment is to be made to the aluminum core. That is a detail that isn’t fully thought out yet.
Erik talked about the movement or memory of the Alcomp CC material. Hytec ran tests over the weekend. They loaded a part in a 3-point bend test. There was only a small amount of creep. Cycling reveals only a little bit of hysteresis. They also got the piece back from the machine shop, which reported much higher numbers for creep or hysteresis. There was some bowing evident. Hytec will evaluate it.
Erik discussed the MCM closeout side thickness. For the 5 mm thick design, the 100g mass goes to 75g with pockets cut in from the back. In the 3 mm design the mass was 66g. The pockets are cut deep enough to leave only 2mm thickness. The mass could be further reduced by cutting another pocket in the central area after removing the boss with the facesheet alignment feature. The feature has to remain in the initial prototypes because the facesheets are already cut, but in the future, if a pinhole for facesheet alignment is needed at all, it could be cut directly into the 5 mm thick closeout. Erik thinks that one won’t notice machining costs of the pockets in a production run. The back has to be machined flat in any case, so the setup has to be done. He estimates only an additional 15 minutes of machine time. The reasons for the thicker side: One can turn around the mortise and tenon joint, so that the backside of the closeout can lie flat on a surface. The locating pinhole can be cut into the main material, requiring no protruding boss. Again, this leaves a flat surface on the back side for machining. Sandro would like to see a drawing. Erik’s isometric drawing is attached. Tom remarked that we must watch the machine speed in backside and pocket machining to avoid bowing. BJ will double check on where the active area begins in order to set the requirement for depth of the pockets, to keep material out of the active area.
Sandro said that he would like to center the silicon on the tray and thereby make shorter wire bonds from the detectors to the pitch adapter. This gives 1 mm of room on each side. This is still enough room for bonds to be made to the bias circuit. It will be important to be sure not to get glue squeeze-out onto the bonding pads on the bias circuit.
Somebody mentioned the pockets in the non-MCM walls. Should they also be turned around, in order to minimize material under the detector active area? It was pointed out that we already are relying on some of those pockets for clearance of parts on the readout cables.
Sandro reported that production of dummy ladders is continuing in 2 places. 14 ladders are done. One wafer in one ladder is off by 40um, but all others okay. Next week the will work on encapsulation. Gwelen will send Sandro description of the encapsulation method that he tried for MCM chips. The general concept was to use a high viscosity material for a dam and a low-viscosity material for filling. Sandro needs a room temperature cure material. He said that they must decide between silicone versus epoxy. Epoxy makes the ladder strong for handling and gives better physical protection to wire bonds. Sandro asked what was the main point of the encapsulation. Robert said that it is primarily to prevent shorting by way of conductive fibers. Gwelen asked about the exposed bonding pads on the opposite end of the ladders. Sandro said that they could put a bead of glue over them.
Gwelen reported that the MCM design is back in the hands of SLAC again for a bit more rework. Gwelen is working on the mass estimate. He plasma etched some sample Kapton pieces, using Lick Observatory equipment. He found a tremendous improvement of wetting of the surface (by water) after etching. He and Alec made some tests with epoxy and glass beads, using an etched piece, a piece that simply was cleaned, and an untreated piece. They will make peel tests on those samples.
Tom said that for the PDR the parts list and the mass estimate are in progress. He has started on processes and procedures. The ICD is started but is rough. Sandro said that he doesn’t yet know if he would attend the May 30 PDR. Robert asked if people saw a need to delay by a month, which would probably be acceptable given that the instrument PDR is delayed by 2 months. The consensus was to go for the May 30 date, even though it will mean a hard push to get ready in the preceding few weeks.
Sandro said that he would like confirmation of definite travel plans and hotel needs for the upcoming trip to Italy.
Issues:
Action Items: