WBS 1'1 EMU Chambers

1
WBS 1.1 EMU Chambers

• Cathode Strip Chambers
• Andrey Korytov
• L3 Manager

2
Outline

• System Overview
• CSC Group Organization
• RD - cost efficient design
• - performance results
• CSC Production Plan, Production Sites
• Schedule, Milestones
• US Deliverables
• Base Cost Analysis
• Contingency
• Summary

3
CMS Endcap Muon System
Large CSCs (3.4x1.5 m2) 72 ME2/2 chambers 72
ME3/2 chambers Small CSCs (1.8x1.1 m2) 72 ME1/2
chambers 72 ME1/3 chambers 20o CSCs (1.9x1.5
m2) 36 ME2/1 chambers 36 ME3/1 chambers
4
Performance requirements

• Operation very reliable (limited access)
• Offline Resolution 75 mm per chamber (ME1/2)
• 150 mm per chamber (others)
• Trigger 1 mm resolution per chamber
• fast (gt92 within 25 ns window)
• Rates 300-1000 Hz/cm2 (random hits)
• 100 Hz/cm2 (charged particles)
• no aging up to 0.1 C/cm (10 years of LHC)
• B-field non-uniform and up to 1 T

5
Cathode Strip Chambers

• Same chamber presision measurements trigger
• offline spatial resolution 50 mm
• trigger spatial resolution 1 mm in presence
• of electromagnetic debris (6-layer CSC)
• fast timing lt25 ns for 6-layer CSC
• Can work in difficult environment
• high rate capabilities (1 kHz/cm2)
• large (4 Tesla) and non-uniform B-field
• Also
• two coordinates from single plane
• strips can be shaped to measure f-coordinate
  alignment marks are easy
• no stringent control of gas mix, temperature, and
  pressure

6
CMS EMU CSCs

• trapezoidal chambers
• length up to 3.4 m
• width up to 1.5 m
• 6 planes per chamber
• 9.5 mm gas gap (per plane)
• 6.7 to 16.0 mm strip width
• strips run radially to measure f-coordinate
• 50 µm wires spaced by 3.2 mm
• 5 to 16 wires ganged in groups
• wires measure r-coordinate
• gas ArCO2CF4305020
• HV4.1 kV (Qcathode110 fC, Qanode140 fC)

7
EMU Chamber Parameters
8
CSC Group
institution involvement UC Davis -
Simulation Production (UCLA Site) UCLA - RD,
Design Production (UCLA and Fermilab Sites) UC
Riverside - RD Production (UCLA Site) Carnegie
Mellon - RD, Design Fermilab - RD,
Design Production (Fermilab Site) Florida -
RD, Design Production (UF and Fermilab
Sites) Livermore - RD Ohio State - RD,
Design Purdue - RD, Design Production
(Fermilab Site) Wisconsin - RD,
Design Production (Fermilab Site) PNPI -
St.Pitersburg - RD, Design Production (PNPI
Site) IHEP - Beijing - Design Production (IHEP
Site)
9
CSC Group Organization
10
CSC RD Goals

• CSC Cost Minimization
• Simple and Robust Design
• geometry to allow for relaxed tolerances
• minimal number of parts and simple assembly
• Identify Cost Driving Materials, minimize their
  cost
• find readily available commercial products
• Identify Cost Driving Labor, minimize cost
• automate labor intensive operations
• minimize in-house labor, i.e. place orders in
  industry
• testing in the production to minimize repairs
• Maintain Adequate Performance
• Reliable operation at Qcathodegt100 fC
• 75-150 mm off-line spatial resolution
• 1 mm trigger spatial resolution
• gt92 probability of correct bunch crossing
  assignment

11
RD CSC Prototypes
P0 (1995) - muon beam performance tests
T0 (1994) - design tests

6 planes 60x60 cm2 w16 mm s2.5 mm 2r30 mm

2 planes

"top" of large CSC (ME2/2) s(x/w), timing,
efficiencies, tails



P0 (1996) - muon beam performance tests
T1A (1995) - design tests
new winding technique s3.2 mm (no wire support,
timing!) HV segmentation, buttons RTV gas
sealing strips milled with 45-degree
cutter grinding was used to make gap bars
6 planes 60x60 cm2 w6 mm s3.4 mm 2r50 mm

T1B (1995) - design tests
"bottom" of precision CSC (ME1/2) s(x/w),
timing, efficiencies, tails, 1/2-strip finding
trigger hardware LCT in presence of e/m debris
2 planes (both T1A and T1B)
P1A (1996) - large scale design tests
2 planes 3.3 m long 1.2-0.8 m wide w16-10
mm s3.2 mm 2r50 mm
P1 (1996) - large scale complete chamber
6 planes 3.3 m long 1.2-0.8 m wide w16-10
mm s3.2 mm 2r50 mm
P2 (1997) - full scale 11 large chamber
6 planes 3.3 m long 1.5-0.8 m wide w16-8
mm s3.2 mm 2r50 mm
detailed performance studies cosmic ray
tests first complete set of final electronics
prototypes 1998 muon beam tests 1998 high rate
tests (muon beam gammas)
12
CSC Design Single Plane Parameters

• Large Gas Gap -- 10 mm (cf. 5-10 mm)
• relaxes panel flatness tolerances
• allows for wider strips (save on electronics
  channels)
• Thick Wires -- 50 mm (cf. 20-30 mm)
• very robust mechanically
• Large Wire Spacing -- 3.2 mm (cf. 2-2.5 mm)
• relaxes wire placement tolerances
• no intermediate wire supports
• fewer wires

13
CSC Design overall chamber
Design is simple (few parts), robust, and
suitable for mass production
14
CSC Design Cost Driving Materials

• Cost Driving Materials have been identified
• Panels
• Wires
• Gap Bars
• Wire Fixation Bars
• Significant cost reduction of these materials has
  been achieved as a result of an extensive RD
  (still they constitute 74 of the overall
  material cost)
• All material costs based on vendor quotes
• Materials tested on large prototypes

15
CSC Design Panels - material cost driver

• Panels from 3 bidders (out of 11 - see on the
  right) were ordered
• flatness, mechanical properties studied (CMS
  Note 1995-094)
• Plascore was chosen
• panels with FR4 skins (ready product)
• very stable
  polycarbonate core
• flat within our spec.s
• inexpensive
• willing to work on technology improvements
• EuroComposite and Teklam are fallback options
• A few batches of panels from Plascore have been
  ordered since 1995
• All large prototypes (and most of small ones)
  were made out of these panels
• and showed very reliable performance

Panel Manufacturers AAR Cadillac ACT Advanced
Composites EuroComposite Hexcel M.C.Gill Norfield
Oregon Composites Plascore Teklam Todco
16
CSC Design Wires - material cost driver

• Large MWPC Systems used Luma wire 587/km for
  50 mm wire
• Sylvania quote for similar wire 144/km (also
  used in MWPCs)
• Sylvania and Luma wires were purchased and
  thoroughly tested,
• thus, we could directly compare Sylvania wire
  against Lumas
• visual analysis under microscope good
• elasticity limit and breakage point good
• breakage under sparking good
• chamber performance good
• All large scale prototypes have Sylvania wire
• and have shown reliable operation

17
CSC Design Cost Driving Labor

• Strategy for
  minimizing labor cost
• robust and simple chamber design, suitable for
  mass production
• labor intensive operations are identified and
  automated/optimized
• strip milling vs. etching
• wire winding
• wire soldering
• gluing
• parts are purchased from commercial vendors to
  minimize in-house labor
• testing along chamber assembly to minimize
  repairs

18
CSC Design Strips - labor cost driver

• Lowest etching quote (two bidders CCT,
  Buckbee-Mears)
• 9,549 per chamber (9 panel sides)
• plus 104K initial investment is required
• Milling (all based on large scale prototypes)
• 1 hour (per plane with 80 strips) fast
• 25 mm precision good
• 45o cutter mills with almost etching
  quality good
• smooth vs. milled cathode CSC performance ident
  ical
• With all handling, cleaning and testing,
  milling will cost
• 1,900 per chamber, i.e. 5 times cheaper (2M
  of savings)
• All prototypes (except for the small P0) have
  had milled strips
• and showed very reliable operation

19
CSC Design Wire Winding - labor cost driver

• Transfer frames too labor intensive
• New winding machine to wind directly on panels
• two planes (1000 wires each _at_4 turns per min) -
  less than one shift
• minimum handling
• no asymmetric stresses on a panel
• Chamber prototype results
• uniform tension 5 (we need 10)
• wire spacing 100 µm (we need 200 µm)
• All prototypes showed very reliable operation

20
CSC Design Wire Soldering - labor cost driver

• Hand-soldering (P1A large scale prototype 1000
  wires per plane)
• 2 FTEs x 6 days per 6-plane chamber
• Robotic soldering machine was built at Fermilab
  
• using commercially available automated
  soldering head (Panasonic)
• - 3.5 sec per soldering joint (tested on P2
  full scale CSC),
• - high quality and uniformity of soldering
  joints
• Projected Time with all handling, wire cutting,
  etc
• 1 FTE x 4 days per 6-plane chamber,
• i.e. 3 times faster than hand soldering
• Estimated Savings at Fermilab Production Site
• 270K (labor) - 90K (machine investment)
  180K
• P2 protototye (full scale large CSC) works
  reliably since January 1998

21
Performance RD Operation Reliability

• P1A, P1, P2 large scale prototypes
• operation point at 4.1 kV and at 4.5 kV
  prototypes are still operational
• P1 has been under HV for more than a year
• P1A shipped twice, P1 shipped once no problems
  (a gt10g)
• Aging Studies
• 10 years of LHC at full luminocity gt
  accumulated charge on wires is 0.1 C/cm
• Use of CF4 gas is know to prevent aging gt no
  aging effects up to 10 C/cm
• CSCs made according to our design and out of
  the design materials
• 20 CF4 - chamber irradiated in excess of 13
  C/cm
• 10 CF4 - chamber irradiated in excess of 13
  C/cm (high gas flow) plus
• additional 0.7 C/cm accumulated at gas
  flow rate 0.1 V/day
• and 1.5 C/cm at gas flow of 1 V/day
  (nominal)
• No gas gain variations were observed. Dark
  current remained small.

22
Performance RD Off-line Spatial Resolution
solid points single plane resolution in cosmic
rays for top ( ), middle ( ), bottom ( )
parts of the trapezoidal CSC (P1
prototype) solid curves single plane resolution
Monte Carlo dashed curves 6-plane resolution as
extrapolated from single plane data
P1 - large scale prototype
Chamber planes are half-strip staggered and
expected six-plane resolution is 80 mm, i.e.
well within 150 mm spec.
23
Performance RD Trigger Spatial Resolution
s0.7 mm

• BEAM TEST RESULTS
• comparators find hits to within a 1/2-strip
  with 92 efficiency
• six-plane patterns (LCTs) are found with 99
  efficiency
• and 0.11(strip width) 0.7 mm resolution
• in presence of em debris accompanying muon
  behind the iron

24
Performance RD Trigger Bunch Tagging

• 6-plane patterns are found with
• 99.5 efficiency
• (smaller prototype beam tests)
• Time stamp 97 probability to be
• within 25 ns window
• for the 2nd earliest signal
• out of 6 hits in an LCT pattern
• (well within the required 92)

P1 - large scale prototype
25
CSC Design Summary

• Chamber Design is optimized
• Cost Drivers (materials and labor) are identified
  
• and their cost is well understood and
  minimized
• Mass Production Tooling is prototyped
• Full Scale Chamber Prototypes meet our
  requirements

26
CSC Production Plan

• Production is divided between five sites
• Fermilab site
• UC site
• UF site
• PNPI - St.Petersburg site
• IHEP - Beijing site
• optimal use of Fermilab infrastructure and
  university contributions
• use of the base program resources at Fermilab
  and universities
• - chamber construction and tooling experts
• - physicists with expertise in chambers and
  electronics
• use contributions from CMS collaborators
  (PNPI-St.Petersburg,
• IHEP-Beijing) assembly of smaller chambers
  (4M in US costs)
• one common chamber dsign
• all parts and critical tooling made in US

27
CSC Production Sites

• Fermilab site (Fermilab/university consortium)
• - procurement of chamber parts for all chambers
• - strip milling for all chambers
• - assembly and HV-training of large chambers (148
  ME23/2)
• - sample testing
• UC and UF sites
• - outfitting large chambers with electronics and
  services
• - system tests of large chambers
• - installation, commissioning, maintenance of
  large chambers
• PNPI and IHEP sites
• - assembly and tests of smaller chambers
• - installation, commissioning, maintenance of
  smaller chambers
• all parts and critical tooling are provided by
  US
• labor is covered by PNPI and IHEP--it is their
  contribution to CMS

28
CSC Production
PNPI Site
parts and critical tooling (smaller chambers)
smaller CSCsElectronics, tested installation/comm
issioning
UC Site
large CSCs
large CSCsElectronics, tested installation/commis
sioning
Fermilab Site - panel production - large CSC
assembly
CERN
large CSCsElectronics, tested installation/commis
sioning
large CSCs
UF Site
Procurement
parts and critical tooling (smaller chambers)
etc.
smaller CSCsElectronics, tested installation/comm
issioning
frames
guard strips
gap bars
IHEP Site
wire fix bars
wire
panels
29
Fermilab Site Panel Milling
Location Lab 8 Operations Inspection
incoming panels (thickness, flatness,
etc.) Axxiom Machine drill holes, cut
trapezoids Gerber Machine mill strips, cathode
and anode artwork Certification milled panels
cleaned, milling quality certified Packaging pa
nels wrapped for storage shipping
30
Fermilab Site CSC Assembly
Location MP-9 Operations Gluing Station -
glue wire fixation bars - glue long and short
guard strips - glue gap frame
bars WindingSoldering Station - wind wires on
panels, glue wires - solder wires, cut wires
Pre-Assembly/Testing Station - solder Rs, Cs,
connectors, protection boards - test HV in
air wire tension/spacing (sample) Assembly
Station - stack panels and Al frame bars, test
HV - tighten bolts, RTV seal - wire HV
connectors HV-training Station - HV-training of
chambers Repair Station
31
Final Assembly System Tests Sites
Location UCLA and UF Operations Long Term
HV-Conditioning Final Assembly - services and
gadgets - electronics cards and enclosures (10
units per chamber) - cabling, marking cables
(100 different cables per chamber) Tests without
HV - cable connections (no mix-ups) - no dead
channels, oscillations, pickup noise
(ground/shield) - electronics-on-chamber (strip
calibration, thresholds, delays) Tests with HV -
HV is properly filtered - determine operation
point for each plane (gas gain, efficiency) -
minimum plateau with respect to Idark and Nnoise
(wires/strips) - tune anode LCT timing w.r.t.
to cosmic muons - time matching between anode
and cathode LCTs Detailed tests - time
permitting (not necessarily on each
chamber) Chamber repairs Database
32
CSC Project Schedule
33
CSC Production Schedule
Lab 8 MP 9 UC
Site UF Site PNPI IHEP 1999 22 6
/.5 yr 2000 98 17 8 8 2001 98 50 25 2
5 20 38 2002 98 50 25 25 26 50 2003 56/.5 yr
25/.5 yr 16/.5 yr 16/.5 yr 30 60
Panel Count by CSC type at Lab 8
ME23/2 ME23/1 ME1/23
TOTAL 1999 22 22 2000 42 20 36 98 2001 42 20
36 98 2002 42 20 36 98 2003 (1/2 yr) 16 40 56
34
Major CSC Milestones

• 01/01/98 P2 (ME23/2) 11 scale prototype
• 12/20/98 ME23/2 - final design
• 11/16/98 ME2/2 Preproduction Prototype
• 12/08/98 ME2/1 Prototype
• 01/15/99 Panel Production Design Review
• 03/16/99 Panel Production Site is ready, Start of
  panel production
• 12/21/99 22 ME23/2 CSC worth of panels made, Full
  Speed Production begins in Jan 2000
• 03/15/99 Chamber Assembly Design Review
• 07/01/99 Chamber Assembly Site is ready, Start of
  ME23/2 chamber assembly
• 12/21/99 6 ME23/2 CSCs assembled, Full Speed
  Production begins in Jan 2001
• 12/21/99 ME3/1 and ME1/2 Preproduction
  Prototypes, small chamber design review
• 12/20/00 PNPI and IHEP Sites ready, Start of
  Small Chamber Production in Jan 2001
• 12/21/99 FAST sites are ready, Start of detailed
  CSC tests with prototype electronics
• Full Speed Final Assembly and Testing begins in
  July 2000

35
WBS 1.1 Deliverables

• Design Deliverables
• Cathode Strip Chambers ME1/2, ME1/3, ME2/1,
  ME3/1, ME23/2
• CSC production tooling
• CSC testing equipment and procedures
• High Voltage System
• Production Deliverables
• 1444148 large ME23/2 chambers production of
  parts, assembly, outfitting with electronics,
  testing, commisssioning at CERN (Fermilab, UC, UF
  sites)
• All parts for smaller CSCs (assembly kits)
• 36238 ME2/1, 36238 ME3/1 (to be
  assembled/outfitted/tested at PNPI)
• 72274 ME1/2, 72274 ME1/3 (to be
  assembled/outfitted/tested at IHEP)
• Critical tooling and testing equipment for PNPI
  and IHEP sites
• High Voltage for all ME1/2, ME1/3, ME2/1,ME3/1,
  ME23/2

36
CSC Cost Estimate WBS L3
M12.835 (base, in FY97 ) 42 (Contingency)
M18.227
37
CSC Base Cost Estimate WBS L3
38
CSC Base Cost Estimate Categories
39
CSC Base Cost Estimate Details
CSC Base Cost M12.835
40
CSC Cost material cost database

• I. Parameters, Parts for all CSCs

41
CSC Cost material cost database

• II. Quotes are linked to the part list

42
CSC Base Cost Estimate Materials
43
CSC Base Cost Estimate Labor

• construction of the full scale prototypes (P1,
  P2) is the basis for the estimate
• major and most of minor tooling protypes have
  been used in the full scale chamber prototypes
  assembly
• all operations have been timed
• operations at individual production stations are
  outlined in space and time
• 80 efficiency is assumed (Lab 8, Lab 6
  experience),
• i.e. manpower assumed is sufficent to make 1 CSC
  every 4 days, while average yeald is assumed to
  be 1 CSC every 5 days
• tooling capability is 1 CSC every 3 days in one
  shift
• soft production ramp-up is built in
• standard Fermilab and University labor rates are
  assumed

44
CSC Cost Panel Production Labor
45
CSC Cost CSC Assembly Labor
46
CSC Cost Labor at peak of production
Fermilab 1 PhD supervisor (senior) Lab
8 1 PhD supervisor 0.5 tech. supervisor
(P.Deering) 2 CNC operators (1 Axxiom, 1 Gerber)
1 tech (cleaning/inspection/wrapping) 0.4
techs (2 techs one day a week for
handling/shipping/etc.) MP 9 1 PhD
supervisor 1 tech. supervisor (K.Kephart) 7
techs (primary assembly stations) 1 specialist
(testing/HV-training expert) 2 PhDs
(repairs/tests/etc.) FAST Sites 1 PhD
supervisor (senior) 1 PhD lab physicist (full
time in the lab) 1 tech (outfitting with
services, electronics, handling, etc.) 1 expert
(electronics/chamber problem debugging) 2 FTE
students (1 FTE assembly, 0.5 cathode work, 0.5
anode work)
47
CSC Manpower Profile
48
CSC Cost Obligation Profile
49
Contingency MFxJF
Materials
maturity factor (MF) 1.0 - purchased 1.1 -
catalog item, P.O. 1.2 - vendor quote 1.3 -
complete design (TDR) 1.4 - indirect estimate,
request for info w/sketches 1.5 - estimate
based on conceptual design judgment factor
(JF) 1.0 - standard, simple, we buy now at this
price 1.1 - extracted from very similar
quote 1.2 - chain of vendors - not
100 prototyped 1.3 - single vendor 1.4
- lack of knowledge (not prototyped)
contingency on not purchased materials cannot
be less than 20
50
Contingency MFxJF
Labor and Engineering
maturity factor (MF) 1.25 - vendor quote
and/or complete design judgment factors (JF)
1.0 - coordination, management 1.1 - labor
directly verified on prototypes 1.2 - not 100
prototyped 1.4 - lack of knowledge (not
prototyped)
51
Contingency Examples

• FR4 sheets 1.2 (vendor quote) x 1.3 (single
  vendor) 1.56
• ME3/1 Gap Bars
• 1.2 (vendor quote for ME2/1) x 1.1 (similar
  chambers) 1.32
• Shipping 1.5 (conceptual design) x 1.2 (not 100
  prototyped) 1.80
• Winding 1.25 x 1.1 (labor directly verified on
  prototypes) 1.38
• Repairs 1.25 x 1.4 (lack of knowledge, not
  prototyped) 1.75
• Technical coordination of overseas sites
• 1.5 (conceptual design) x 1.4 (not prototyped)
  2.10

52
Contingency by Category

• overall CSC contingency 42
• material cost is based on large scale
  prototypes and quotes 36
• direct labor estimate is based on prototype
  construction 50
• production and testing equipment/tools
  50
• equipment/tools maintenance 110
• production supervision 25
• base program and university contribution labor
  assumes
• at least 50 contingency available
• foreign collaborators are responsible for
  contingency
• (2M in US costs) on their contribution

53
Summary

• CSC design - optimized and validated on large
  scale prototypes
• - prototype performance meets our
  specifications
• Cost - materials from the planned vendors are
  verified on prototypes
• - material cost is based on vendor quotes
• - labor evaluation is based on experience with
• full scale chambers construction using final
  tooling prototypes
• - labor cost calculated from standard
  Fermilab/University rates
• Organization - production plan has been
  developed
• - production sites and site managers are
  identified
• - agreement on responsibilities and
  deliverables between
• institutions to be involved in the
  production have been reached
• Schedule - Project File outlines detailed
  schedule through 2004
• - all past CSC milestones have been
  accomplished on time