ATLAS Pixel Detector Pixel Support Tube PRR Assembly

ATLAS Pixel Detector Pixel Support Tube PRR: Assembly, Production, Schedule February 2003 N. Hartman LBNL

Pixel Detector ATLAS Overview • Assembly Breakdown Structure (ABS) • Assembly Processes and steps – Assembly flow – Parts required – Tooling required • Production Information – Tooling status – Production plan • Resources • milestones – Schedule February 2003 N. Hartman LBNL

ATLAS Pixel Detector Assembly breakdown structure (ABS) Barrel PST assembly shown; b. s. = barrel shell. February 2003 N. Hartman LBNL

ATLAS Pixel Detector Assembly Steps and Procedures February 2003 N. Hartman LBNL

Pixel Detector Assembly Flow – How is the PST Assembled? ATLAS Barrel Shell Flat Rail V-Rail Barrel Shell with flanges 1 1 Backing Ring Mount Pad 1 barrel shell structural 2 4 1 Flange Asm barrel shell w/ mount pads 5 Flange Face 3 Hoop Stiffener barrel shell w/ SCT Mounts Insulator 1 Hoop Element 4 1 February 2003 SCT Mount Block Asm N. Hartman LBNL

Pixel Detector ATLAS STEP 1 – Flanges Assembled Backing ring Backing Ring Single piece skirt and face (L shape) Flange Cross Section • Process – – • Caul plate (Fabricated) Drill jig (Fabricated) QA Steps – – February 2003 Flange masked Adhesive applied (bondline controlled with glass beads) Backing ring weighted with caul plate Assembly drilled with hole jig Tooling required – – • Flange Face Weigh parts before and after for resin content Inspect datum hole positions after machining N. Hartman LBNL

ATLAS Pixel Detector Bond Jig Details Accurate holes on horizontal diameter Of flange provide references for all Bond operations. Gussets provide perpendicularity February 2003 Bars provide length Reference only (invar) N. Hartman LBNL

Pixel Detector ATLAS STEP 1 a – Flanges Bonded to Shell • Process – – – • Tooling required – – • Flange and jig masked Shell centered in bond gap with shims Flange sealed to bond jig Adhesive applied behind flange Adhesive pulled through joint with vacuum (from front) Bond Jig (Fabricated for prototype length) Barrel and Forward Length Spacer Bars (needed) QA Steps – – Remove from jig and weigh (resin control) Inspect parallelism and flatness of flanges after bonding February 2003 N. Hartman LBNL

Pixel Detector ATLAS STEP 2 – Hoop Stiffeners and Rails Bonded to Shell • Process – – • Tooling required – – • Hoop Stiffeners – Applied before rails – Shell masked – Hoops assembled in place, as single bonded units, 3 parts at a time – Glass beads used to control bond line – Compression strap used to fixture while curing Rails – Rails are affixed to ladder structure that runs through PST shell (with vacuum) and references to pins on flange plate – Adhesive applied to shell – Rail tool lowered into place None for stiffeners (self-fixturing) Vacuum rail jig attachment (needed) QA Steps – – Remove from jig and weigh (resin control) Inspect positions of rails at flange ends (to ensure alignment when PST sections are mated) February 2003 N. Hartman LBNL

Pixel Detector STEP 3 – Mount Pads Bonded to Shell/Flanges ATLAS • Process – – • Tooling required – • Machined mount pads affixed to bond jig extensions (referenced to accurate pins on flange plate) Shell and mount pad masked Adhesive applied to mount pad (bondline controlled by fixture) Mount pad/jig extension moved into place Bond jig extensions (fabricated) QA Steps – Remove from jig and weigh (resin control) February 2003 N. Hartman LBNL

Pixel Detector STEP 4 – Mount Block Locations Transferred from SCT to Mistress Gauge ATLAS Frame Gauge blocks • • SCT mistress gauge blocks register to PST Interface penetrations Carbon fiber tooling bar Frame is bonded to them and supported in SCT barrel while adhesive cures February 2003 N. Hartman LBNL

Pixel Detector STEP 4 a – Mount Block Locations Transferred from Mistress Gauge to PST Bond Jig ATLAS • • • February 2003 Blocks machined to Surveyed position of PST mounts (relative to IP_Nominal) Balls register Mistress gauge in proper orientation Bond in SCT_Interface Transfer Feature N. Hartman LBNL

Pixel Detector ATLAS STEP 4 b – Mount Blocks Bonded to Insulators • Process – Both parts masked – Adhesive applied to mount block – Glass beads used to control bondline Fiberglass Insulative shim • Tooling required – None, self-fixturing • QA Steps – Weigh (resin control) – Measure thickness Mount block February 2003 N. Hartman LBNL

Pixel Detector ATLAS STEP 4 c – Mount Block Assy’s Bonded to Shell/flanges Backside of bond surface • Process – – – • Inside of mount pad and mount block masked Mount block inserted through mount pad, bolted to jig Adhesive injected between mount pad and mount block bond surfaces Tooling required – – • Attachment surface to jig Bond Jig attachment pieces (Needed) Mount block interfaces bonded into jig attachment with SCT mistress gauge in place (needed) QA Steps – – Remove from jig and weigh (resin control) Survey positions of mount locations to SCT February 2003 N. Hartman LBNL

Pixel Detector STEP 5 – Pixel Mount Blocks Bonded to PST ATLAS Pixel Mount Ramp • Process – – – • Tooling required – • Outside of mount pad and pixel mount block masked Mount pad affixed to jig that transfers position from SCT mount block jig Adhesive injected between Pixel mount block and mount pad SCT mount block tooling – used for both pixel and SCT mount blocks, as they are defined relative to each other (needed) QA Steps – – Remove from jig and weigh (resin control) Survey positions of mount locations (including planarity) February 2003 N. Hartman LBNL

ATLAS Pixel Detector Sample Assembly Follower Raw material data (out time, date, remaining mass) are kept in log book at freezer. February 2003 N. Hartman LBNL

ATLAS Pixel Detector Production Planning February 2003 N. Hartman LBNL

Pixel Detector ATLAS Status of Tooling Items • Existing • – Flange Tools • • • Face piece Backing ring Assembly (bonding) Machining fixtures Flange bond plates to PST – Mount Pad • Layup tool • Machining fixture • Mount pad bond parts to PST – Rails • Full length tool Needed – Mandrel (Procure) • Cylinder • Cart • Stand – Mount bonding (In house) • SCT Mistress gauge • Bond jig attachments • Pixel mount bond parts – Rails (In house) • Bonding tool (vac fixture) – Hoop Stiffeners (In house) • Layup tool Main Bond Jig Parts are already fabricated February 2003 N. Hartman LBNL

Pixel Detector ATLAS Procurements/Responsibilities • Tooling – Only mandrel and carts/stands will be outsourced • Engineering design by outside vendor – complete • Awaiting quote for fabrication • Estimated delivery is 8 weeks – All other tooling fabricated in-house • Material – All material from Bryte technologies – ex 1515 cyanate ester resin • Usual delivery time is 3 -4 weeks • Material to be ordered by March 14 – Heater panels from outside vendor • Composite Parts – All parts made in-house, by LBNL technical staff February 2003 N. Hartman LBNL

Pixel Detector ATLAS Mandrel and Autoclave Carts Production Mandrel and cart have been designed and quoted by outside vendor - ~32 K USD February 2003 N. Hartman LBNL

Pixel Detector ATLAS Production Resources • Manpower – 3 technicians now trained in composites – Schedule assumes ½ FTE technician over production period • Facilities – Autoclave (1. 5 m id x 3 m long) – CNC ultra high-precision machining – Coordinate measuring machine – Cold storage (-40 Celsius, 3 m x 5 m) – Dedicated Layup facility (10 m x 10 m) February 2003 N. Hartman LBNL

Pixel Detector ATLAS Important Milestones • Production Readiness Review – February 18, 2003 – CERN • Full Size Mandrel and Material Ordered – March 1 -14, 2003 • Production Material Delivered – April 14, 2003 • Mandrel Delivered – May 1, 2003 • Production Begins – May 1, 2003 • Barrel PST Completed – September 14, 2003 • Forward PST’s Completed – June 1, 2004 • Pixel System Dry-fit (PST, BP Support, service panels, pixel frame) – July, 2004 • Fit-up with SCT (transfer jig bonded) – September, 2004 • SCT/PST Installed – January, 2005 February 2003 N. Hartman LBNL

ATLAS Pixel Detector Full Schedule February 2003 N. Hartman LBNL

Pixel Detector ATLAS Potential Risks • Prototyping – Problems with rail riders – not yet designed (pixel and beampipe/services) • Production – Delays or problems in mandrel machining – Delay in mount transfer from SCT – Problems in schedule or performance of pixel package dry fit February 2003 N. Hartman LBNL