from ANNEX 02 to NCR/FM/INFN/RM2/003

1
Preventive Cuts (Pcut) Procedure

• originally thought to prevent delaminations in
  TVAC
• only performed on 8 tower0 trays on W-side
  (8layers)

• 26.5 cm/tile
• 4.24 m/layer
• 34 m total length of tower0 preventive cut

from ANNEX 02 to NCR/FM/INFN/RM2/003
2
Bubble Cuts (Bcut) Procedure
Small (lt2mm) cut far from bias lines (gt1mm) using
an edged scalpel Insulation test (OL) with an
ohm-meter between tray and shield plane Used
for tower0 trays and TowerA trays
Estimate of total bubble-cut for tower0
17layers x 2bubbles x 2mm lt 7cm
3
Tower0 trays summary chart
Tower Position Tray S/N Tray type W Pcut Bcut (side) HV Short (side) layers  Comments
18 001 TOP Sand-Blasted (SB) N FAIL TVAC   Y17 No Si
17 048 MID SB N Y(T/B)   X16-X17 No Si
16 047 MID SB N FAIL TVAC   Y15-Y16 No Si
15 004 MID Hand-Blasted (HS) Y Y(T/B) X14-X15 No Si
14 017 MID HS Y Y Y13-Y14 No Si
13 012 MID HS Y Y(T) Y(B) X12-X13 HV-short on bottom
12 011 MID HS Y Y(T/B) Y(B) Y11-Y12 HV short on bottom
11 008 MID HS Y Y(T/B) Y(B) X10-X11 HV short on bottom
10 006 MID HS Y Y(T) Y(B) Y9-Y10 HV short on bottom
9 046 MID SB N Y(T) N X8-X9 OK
8 007 MID HS Y Y(B) N Y7-Y8 OK
7 010 MID HS Y Y(B) N X6-X7 OK
6 009 HEAVY HS Y Y(T/B) N Y5-Y6 75 disc. ch.
5 008 HEAVY HS Y Y(T/B) N X4-X5 183 disc. ch.
4 035 HEAVY SB N Y(B) N Y3-Y4 249 disc. ch.
3 011 HEAVY SB N Y(T) N X2-X3 1319 disc. ch.
2 023 LIGHT   N Y(T/B) N Y1-Y2 OK
1 015 LIGHT   N Y(B) N X0-X1 21 disc. ch.
0 002 BOTTOM   N Y(T) N Y0 OK
4
Conclusions risk assessment

• 0 short on shield lines
• 4 HV short on trays Pcut and Hand-Sanded W
• 23 layers OK on HV line (6 Pcut, 15 Bcut)
• 27 layers OK on shield line (8 Pcut, 17 Bcut)

• ? all shorts due to W whiskers generated from
  tiles hand-sanding which entered long, deep Pcuts
  such whiskers were actually observed
• on towerA
• NO Pcuts
• W tiles are etched and primed, therefore very
  flat, clean and whiskers-free
• Bcuts are superficial (see procedure)
• Lets calculate anyway a conservative HV-short
  probability
• a HV short would cause 1 ladder loss (on tower0
  trays with HV-short we disconnected HV resistors
  on MCM boards for all 4 ladders to avoid current
  flow)

Prob(1ladder-loss/Bcut) 4 / 34 m x 2mm 0.23
10-3 , i.e. 1ladder-loss every 4250 Bcut or 1
ladder with HV-short every 2150 ladders (assuming
1 Bcut every 2 ladders)
5
Conclusions schedule constraints
List of available trays
type TrayID Bcut - bottom Bcut - top MCM pocket
M 078 10 1.14 small
M 082 11 11,12,10,14,15 ready
M 051 6 7,10 small
M 077     assembled
M 063     ready
M 050     assembled
M 053   7 assembled
M 066 6,10 7,11 ready
M 079     assembled
M 064 7,8,12 7,12 ready
M 081 7,10,14,15   ready
M 080 2,7,15 11,10 ready
M 076 11 10 small
M 062 10 6,7,9,10,1 small
M 002 11   ready
T 002 6,10 NA NCR
type TrayID Bcut - bottom Bcut - top MCM pocket
B 003 NA 10 ready
B 004 NA 6,10,11
H 012 6,10,11 6,7,15 ready
H 010 7,10,11 6,7,10,11 ready
H 020 6,7,11,12   ready
H 006 10,11 6,7,9,10,11 ready
H 005 10,11 11 ready
H 037 7   small
H 018   11,15 small
L 013 7 7,11,12,16 small
L 013 7 7,11,12,16 small
L 004 10 6,10 small
M 042     II test article
3 trays without Bcut and MCM pocket deep enough
to hold MCM ? GA flow will be slowed until new
trays are available without Bcut