SLAC Phase II Secondary Collimators

1
SLAC Phase II Secondary Collimators
US LHC Accelerator Research Program
BNL - FNAL- LBNL - SLAC

• 6 October 2005
• LARP Collaboration Meeting-St. Charles, IL
• Tom MarkiewiczSLAC

2
Task1 Studies of a Rotatable Metallic
Collimator for Possible Use in LHC Phase II
Collimation System

• If we ALLOW (rare) ASYNCH. BEAM ABORTS to DAMAGE
  METAL JAWS, is it possible to build a ROTATING
  COLLIMATOR
• that we can cool to lt10kW, keeping TltTFRACTURE
  and PH2Olt1 atm.
• that has reasonable collimation system efficiency
• that satisfies mechanical space accuracy
  requirements

• Scope
• Tracking studies to understand efficiency and
  loss maps of any proposed configuration
  (SixTrack)
• Energy deposition studies to understand heat load
  under defined normal conditions damage extent
  in accident (FLUKA MARS)
• Engineering studies for cooling deformation
• Construct 2 prototypes with eventual beam test at
  LHC in 2008
• After technical choice by CERN, engineering
  support
• Commissioning support after installation by CERN

3
Task1 Timescale Manpower
FY 2004 Introduction to project FY 2005 Phase
II CDR and set up of a collimator lab at SLAC FY
2006 Design, construction testing of RC1 FY
2007 Design, construction no-beam testing of
RC2 FY 2008 Ship, Install, Beam Tests of RC2 in
LHC May-Oct 2008 run FY 2009 Final drawing
package for CERN FY 2010 Await production
installation by CERN FY 2011 Commissioning
support RC1Mechanical Prototype RC2 Beam Test
Prototype
Meeting/advice Tor Raubenheimer Andrei Seryi Joe
Frisch
Active Manpower Eric Doyle-Engineering Lew
Keller-FLUKA Yunhai Cai-Tracking Tom Markiewicz-
Integration
Future Effort Controls Engineer Designer
Engineer2 Postdoc1
Planned hires Mech. Engineer2 Postdoc1
4
2005-06-01 DOE Review of Collimation Program A

• "The activity in collimation is impressive, with
  the work being approached in a very professional
  manner. It is a critical problem, and solving it
  will have great impact on the ability of the LHC
  to reach design luminosity. Even the task sheets
  for this project were done very professionally,
  lending confidence in a well-managed and
  well-focused activity. (The synergy with the ILC
  is clearly evident here.)"

5
Status of Phase II Collimator Conceptual
Designat 2005-06-01 DOE Review

• Adequate software in place and MANY studies have
  been done but
• We do NOT yet have a conceptual design for the
  1st of the 11 collimators needed (per beam) in
  IR7
• 28 of 30 Phase II collimators will not have a
  heating problem
• No magic design or material which could
  simultaneously provide good efficiency with
  combination of energy absorption, thermal
  conductivity, thermal expansion to maintain 25
  um flatness tolerance over length of jaw during
  1hr/12min (90/450kW) beam lifetime transients for
  nominal jaw length (1m) and gap setting (7s)
• Focus on
• 150mm O.D. 25mm wall Cu jaws with helical cooling
  tubes
• 150mm O.D. solid Cu jaws cooled with axial flow
  over 36 of arc

6
Study of Material for Secondary Collimators

• High Z materials improve system efficiency
• Copper being considered because its high thermal
  conductivity
• Available length is about 1 meter
• Achievable efficiency is about 3.5x10-4 at
  10 s
• As Sixtrack program adds absorbers/tertiary
  collimator we expect x10 improvement

Yunhai Cai
Copper
Similar result was obtained by Ralph Abmann
7
Energy Deposition in Metal Phase II Secondary
Collimators w/ Carbon Phase I Collimators Open
Calculated in FLUKA using CERN-provided input file
8
Power absorbed in one TCSH1 jaw at 10s when 80
(5) of 450kW of primary beam interacts in TCPV
(TCSH1)
9
Helical and axial cooling channels
10
ANSYS thermal and structural results for full ID
cooling and limited cooling arc showing 64 less
distortion with limited cooling

360 cooling of I.D.
36 cooling arc
Note transverse gradient causes bending
Note axial gradient
61C
89C
Note more swelling than bending
support
dx221 mm Spec 25mm
dx79 mm
64 less distortion
support
11
Material Comparison SS Transient Thermal
Deflection ANSYS simulation results for 150mm OD
x 100mm ID x 1.2m L

• Notes
• BeCu is a made-up alloy with 6 Cu. We believe
  it could be made if warranted
• 2219 Al is an alloy containing 6 Cu
• Cu/Be is a bimetallic jaw consisting of a 5mm Cu
  outer layer and a 20mm Be inner layer
• Cu 5 mm is a thin walled Cu jaw
• Super Invar loses its low CTE above 200C, so the
  152um deflection is not valid
• Heat flux to water of 106W/m2 or greater is in
  regime of possible film boiling

12
Material evaluations compared to Cu based on
150mm x 1200mm x25mm wall model
Material Reasons for rejection in favor of Copper
BeCu (6 Cu-loaded Be) Beryllium is forbidden by CERN management, low cleaning efficiency due to few particles absorbed fabrication difficulty
Super Invar Poor thermal conductivity ? high temperature (866C), desirable properties (low thermal expansion coefficient) disappear at 200C
Inconel 718 Poor thermal conductivity ? high temperature (Tmp 1400C lt 1520C transient peak) very high deflection (1039um SS, 1509um transient)
Titanium Poor thermal conductivity ? deflection 2.7 x Cu (591um, SS)
Aluminum Relatively poor cleaning efficiency, water channel fabrication difficulty
Tungsten High temperature on water side (240C - 30bar to suppress boiling) high power density - can't transfer without boiling
13
Directions under investigation negotiation at
time of 2005-06-01 DOE Review

• Redefine secondary hybrid system to treat 1st
  collimator as special
• Break 1st secondary into two (unequal?) lengths
  of perhaps different materials
• Grooved expansion slots to limit deformation
• Adjust gaps of the first carbon metal secondary
  to reduce heat load while maintaining efficiency
  with remainder of secondary system
• Keep 1st C-C secondary collimators jaws at 7s
  and leave out 1st metal secondary collimator
• Relax deformation tolerance relaxed to if jaws
  expand AWAY from beam
• Begin to deal with LHC infrastructure
  operational constraints
• 45mm jaw gap at injection incompatible with NLC
  inspired circumferentially mounted gap adjustor
• Look into adopting Phase I adjustment mechanism
• Spatial constraints of LHC beam pipes tunnel a
  challenge
• Jaw dimensions, tank dimension

14
IR7 Collimator Layout
Primary Collimators
Beam Direction
Hard Hit Secondary Collimators
15
Possible Path to Immediate RC1 Prototype Leave
TCS1 Carbon-Carbon, Remainder Cu
Inefficiency   1C-10Cu All Cu
Horizontal 2.84x10-4  3.72x10-4 
Vertical 3.63x10-4   4.36x10-4 
Skew   4.57x10-4     3.85x10-4 
16
Concentrating E_dep in Front Part of Jaw
17
Deflections referred to shaft
Effect of shortened jaws, wider gap, jaw support
scheme
Deflections referred to edge

• Notes
• Solid jaws 150mm OD x 24mm ID. Basic jaw
  120cm L
• Aperture transition from 10s to 7s
• 7s cases based on CERN ray files for interactions
  in TCPV
• pre-radiator not used Phase I carbon jaw to
  concentrate energy toward front of Phase II jaw.

basic jaw 120cm L
18
Circumferential grooves reduce bending deflection
by interrupting continuity of thermal strain.
Parameters 150mm O.D., 25mm wall, 120cm
long Grooves 10mm deep, 50mm spacing 10kW heat,
evenly distributed 45 deg cooling arc
Case Tmax C Deflection (um) Deflection (um)
Jaw edge ref axis ref
Straight 59.5 33 100
grooved 59.5 15 74
19
Mechanical Model 2005-06-01 Used NLC Concept of
Central Strongback with mid-collimator jaw gap
adjuster
20
June 15-17 CERN/SLAC Collaboration Meeting

• Attendees
• CERN Ralph Assmann (Project Leader, Tracking),
  Allesandro Bertarelli (Mechanical Eng.), Markus
  Brugger (Radiation Issues), Mario Santana (FLUKA)
  
• SLAC Tom Markiewicz, Eric Doyle (ME), Lew Keller
  (FLUKA), Yunhai Cai (Tracking), Tor Raubenheimer
• Radiation Physics Group Alberto Fasso, Heinz
  Vincke
• Results
• Agreement on basic design of RC1 (1st rotatable
  prototype)
• Transfer of many of CERN mechanical CAD files
• Lists of
• Further studies required
• Outstanding Engineering Issues requiring more
  design work
• Project Milestone List Action Items List
• Test Installation of New FLUKA

21
Conceptual Design of RC1 (1 of 2)

• Mechanics must fit within CERN Phase I C-C
  envelope
• 224mm center-to-center with 88mm OD beampipes
• 1480mm longitudinal flange-to-flange
• 25mm adjustment/jaw (22.5mm relative to beam
  w/5mm allowed beam center motion
• and use Phase I alignment and adjustment scheme
• Two 75cm Cu cylindrical jaws with 10cm tapered
  ends, 95cm overall length with axes connected to
  vertical mover shafts
• 136mm OD with 9mm taper
• Each jaw end independently moved in 10um steps
• Vacuum vessel sized to provide 8mm clearance to
  adjacent beam and allow gross/fine 0, 45, 90
  positions
• Relaxed mechanical deformation specifications
• lt25 um INTO beam guaranteed by adjustable
  mechanical stop(s)
• Ride on groove deep enough to not be damaged in
  accident case
• Adjustable between 5 and 15 sigma (2-6mm)
  centered on beam
• lt325 um (750um) AWAY FROM beam _at_ 0.8E1p/s loss
  (4E11p/s)
• Flexible support on adjustment

22
Conceptual Design of RC1 (2 of 2)

• Assumed worst case heat load (FLUKA)
• 11.3 kW/jaw steady state, 56.5kW/jaw transient
  (10 sec)
• Cooling boundary conditions
• 200 C maximum temperature of Cu
• 27 C input H2O temperature
• 42 C maximum allowed return H2O temperature
• Two Cooling Schemes under consideration
• Helical tube more secure H2O-vacuum interface
• Axial channels w/ diverter superior thermal
  mechanical performance
• Sufficient pressure (3 atm.) to prevent local
  boiling in transient
• Flexible supply lines to provide 360 rotation
• Other
• Vacuum lt1E-7 Pa (1.3E-5 torr)
• RF shielding scheme has been proposed

23
Proposed layout 136mm diameter x 950mm long
jaws, vacuum tank, jaw support mechanism and
support base derived from CERN Phase I
24
Vacuum tank enlarged to accommodate jaw motion.
Relative location of opposing beam pipe near
interference in skew orientations 10 deviation
25
Helical cooling passages fabrication concept
Based on CERNs design no weld or braze between
water vacuum

• Tube formed as helix, slightly smaller O.D. than
  jaw I.D.
• O.D. of helix wrapped with braze metal shim
• Helix inserted into bore, two ends twisted wrt
  each other to expand, ensure contact
• Fixture (not shown) holds twist during heat cycle

• Variations
• Pitch may vary with length to concentrate cooling
• Two parallel helixes to double flow
• Spacer between coils adds thermal mass, strength
• Electroform jaw body onto coil

26
Adjustable central gap-defining stop

• Stop prevents gap closing as jaw bows inward due
  to heat
• Jaw ends spring-loaded to the table assemby
  move outward in response to bowing
• May use two stops to control tilt
• Slot deep enough to avoid damage in accident
• Stop far enough from beam to never be damaged
  is out of way at injection

27
Swelling vs. bending deflectionEffect depends on
jaw support scheme
28
Flexible end supports used in conjunction with
central gap-defining mechanism
Adjustable central jaw stops (not shown) define
gap Flexible bearing supports allow jaw thermal
distortion away from beam
Self aligning bearing
Leaf springs allow jaw end motion up to 1mm away
from beam
CERNs jaw support/positioning mechanism. Vacuum
tank, bellows, steppers not shown.
29
RF Contact Overview
30
RF contacts cutaway view
31
RF contact variation removes step
32
Jaw Support Concept unresolved issues
interferences with RF parts
33
Flex cooling supply tube concept
34
Detail of flex cooling supply tube
Stub-shaft (bearing not shown)
Contiguous with helical tube inside jaw. Formed
after assembly-brazing of jaw and installation of
bearing on stub-shaft Exits through support shaft
per CERN design Material CuNi10Fe1, 10mm O.D.,
8mm I.D.
Relaxed (as shown) coils 4
Relaxed (as shown) O.D. 111mm (4.4in)
full 360 rotation coils 5
full 360 rotation O.D. 91mm (3.6in)
full 360 rotation torque 9.1N-m (81in-lb)
Support shaft
35
Power Deposition on First 120cm Secondary
Collimator in 12 Min. Lifetime from 20cm C
primary hit files (kW per jaw)
Sensitivity to aperture and to source of halo
H, V, or S
56.5kW/jaw (11.3kW/jaw for 1 hr beam lifetime)
assuming 80-5 interaction ratio split from 20cm
primary hit maps, but adjusting for hadron
absorption in back 40cm of C primaries and
adjusting for shorter 95cm length
Primary Collimator (source) TCSM.B6.L7 Jaws at 7 s TCSM.B6.L7 Jaws at 7 s TCSM.B6.L7 Jaws at 10 s TCSM.B6.L7 Jaws at 10 s
Primary Collimator (source) Copper Al_2219 Copper Al_2219
TCP.D6.L7 (TCPV) 73 26 51 19
TCP.C6.L7 (TCPH) 61 22 49 19
TCP.B6.L7 (TCPS) 92 28 56 20
Notes 1. Collimator data, ray files, and loss
maps from LHC Collimator web page, Feb. 2005. 2.
Must add contribution from direct hits on
secondary jaws.
36
Continuing evolution of ANSYS model to include
realistic cooling channels and cooling water
response to heat generated by beam
Solid jaw has less DT across diameter and short
cut for heat flow
H2O simulation helical flow shown Fluid pipe
elements Water temperature responds to heat
absorbed from jaw More realistic simulation Axial
pipes can simulate axial flow Friction can be
simulated
37
Compact (136x950) jaw variations - performance
comparison
38
Final Expected Performance of RC1 Design
39
Outstanding RC1 Unresolved Issues

• Jaw positioning
• Acceptance of estimated deflection by
  CERN281/869um
• Design concept for central stop gap adjust, 5
  central position float,..
• Bearings springs attaching jaws to vertical
  movers
• Load capacity of steppers
• Jaw alignment perpendicular to collimation
  direction
• Jaw rotation
• Specification of mechanism on crowded jaw
• Force required to rotate jaws against cooling
  coil
• Misc
• Spring arrangement for H, V, S orientations
• Springs to ensure that device fails open
• Motors, cables, temperature sensors, position
  probes,
• Cooling
• Possible local boiling in transient condition
  need for P3 atm. H2O system
• More flexible yet vacuum safe water supply for
  helical cooling
• Vacuum safe water connection/diverter for axial
  cooling scheme

40
Other Studies Planned

• Tracking Studies Hit maps for each of 11 IR7
  collimators/beam and efficiency with 60cm C-C
  primaries and Cylindrical 75cm jaws which
  includes effect of tertiary collimators and
  absorbers
• FLUKA energy deposition with 60cm primaries
  cylindrical 75 cm jaws Complete self consistent
  package of tracking, FLUKA ANSYS results to
  support design choices unambiguously
• Better definition of RC1 thermal tests
• Remnant Prompt Dose Rate calculations
• Engineer damage assessment mechanism into design
• Thermal shock studies
• Studies/experiments to verify
• assumed extent of damage in accident
• Where metal slag will wind up
• acceptable peak temperature of jaw

41
What is the damage area in a missteering
accident?
Missteered beam (9E11 protons) on secondary Jaw
Copper Jaw
Cross section at shower max.
Copper
2.5 cm
Fracture temp. of copper is about 200 deg C
Assumed Damage threshold seems inconsistent with
FNAL experience
42
FY2005 Deliverables
RC1 CDR Draft 9/30/05 32 pages figures
Collimator Assembly Test Area (SLAC-ESB)
43
FY2006 Work PlanNon-beam Mechanical Thermal
Performance of RC1

• Single jaw thermal test one jaw with internal
  helical cooling channels to be thermally loaded
  for testing the cooling effectiveness and
  measuring thermal deformations.
• Heating by commercial electric resistance heaters
  coupled to the jaw with thermal grease
• Operated in a tank purged with inert gas to
  protect the copper jaw from oxidation.
• Flexible cooling supply that wont be intended to
  allow rotation of the jaw.
• A FE model of the test jaw will be used as a
  benchmark to evaluate the response of the test
  jaw to the test conditions.
• Full RC1 prototype a working prototype for bench
  top testing of the jaw positioning mechanism,
  supported to simulate operation in all necessary
  orientations, but not intended for mounting on
  actual beamline supports with actual beamline,
  cooling, control and instrumentation connections.
  
• Lidded vacuum tank for easy access.
• Cooling water feed-throughs and flexible
  connections as realistic as possible.
• A reasonable effort will be made to test RC1
  under heat loading but this will probably prove
  to be impractical.

44
FY 2006 Deliverables

• Final version of RC1 CDR
• Nov. 1, 2005 ?
• External review of RC1 CDR
• Nov. 14-18, 2005 ?
• Dec 12-17, 2005 ?
• Performance report on RC1
• Sept.30, 2006

45
FY2006 Staffing

• Expect continued involvement of existing
  collaborators at current level
• Devoted engineer
• negotiations with SLAC Klystron shop underway
• negotiations with SLAC Research Division ME
  shop underway
• employment requirements prepared will be sent
  to
• Fermilab, CERN, etc.
• Mercury News, etc.
• Devoted postdoc
• Advertisement prepared
• People coming through for a SLAC Postdoc
  interviewed
• In-house external shop design work as needed

46
Detailed FY2006 Work Plan
47
Inter-Lab Collaboration

• Excellent good will cooperation limited only by
  busy work loads on other systems
• Exchange of mechanical drawing files
• Installation of latest FLUKA at SLAC
• Transfer of latest mod to SixTrack
• Latest hit maps with 60cm primaries
• Invaluable 3 day visit by 4 CERN staff
• Monthly video meetings mostly killed
  April-present due to visit, other meetings,
  summer,
• Next video meeting Oct 11, 2005
• CERN review of SLAC draft CDR
• CERN participation in RC1 CDR review
• Exchange of detailed technical information will
  be crucial to delivering prototypes on time
• Drawing of support structures for H, V Skew
• Ideally, CERN would send old prototype parts
  (i.e. everything support structure with
  steppers, motors, bellows, LVDTs, except for
  the tank cylinder jaws) rather than have SLAC
  re-fab from drawings or from transcriptions of
  drawings

48
Conclusion

• To meet the Jan.1, 2008 RC2 ship date requirement
  SLAC and CERN collaborators have agreed on an
  initial set of specifications for the first
  mechanical prototype RC1
• based on extensive but incomplete studies done to
  date
• consistent with CERN beam mechanical
  constraints which uses Phase I design to
  maximum extent
• RC1 Prototype Conceptual Design, while not 100
  complete, has been written up and serves as a
  start point for construction.
• report to be finalized reviewed in FY2006
• Fabrication of RC1 in two main steps in FY2006,
  with appropriate thermal mechanical tests, should
  validate most of the design issues
• Design extension to RC2, a beam-test-capable
  prototype, will occur in parallel
• Good (but of course could always be better)
  communication and exchange of information marks
  collaboration between labs
• 100 devoted manpower required to ensure success

49
Status of Phase II Efficiency Studies

• Excellent understanding of the code
• Tracking simulations of 1m metal secondary
  collimators at 7s show inadequate efficiency
  (plot shown previously)
• CERN provided upgraded code with absorbers
  tertiary collimators added will hopefully show
  adequate performance
• Continue to understand playoff between gaps,
  lengths and materials and provide loss maps for
  use as FLUKA input for suggested modifications

50
Vertical Skew Collimators
Secondary halo in normalized phase space at the
end of collimation system
Primary collimator
Collimators are projected to The end of
collimation system
6 and 7 sigma contours
This is an independent check of the simulation
code, since the collimators are plotted according
to the lattice functions calculated using MAD.
51
Tertiary Halo Particles Escaped from the
Secondary Collimators
TCSG.E5R7.B1 last skew collimator
Number of particles beyond 10s is 73, which is
consistent with the efficiency calculation 73/144
446 5x10-4. Tertiary halo at large
amplitude is generated by the large-angle Coulomb
scattering in the last collimator. If we add a
tertiary collimator at 8s in the same phase
as the collimator TCSG.D4L7.B1 after the
secondary collimators, the efficiency should be
better than 1x10-4.
TCSG.D4L7.B1
52
Status of Phase II Energy Loss Studies

• FLUKA with simple CERN-provided input file
  modeling 40m around primary collimators used for
  all SLAC studies
• Let pencil beam halo interact in primary
  vertical carbon collimator and study energy
  deposition in rectangular 25x80mm jaws at 10s
• Assume 80 inelastic int. in primary, 2.5 in
  each jaw of secondary
• Vary jaw material provide energy deposition
  grid on jaw to ANSYS
• 2.5mm x 8mm x 5cm rectangular grid, mapped onto
  cylinder
• Understand secondary particle content, energy
  spatial distributions
• Use CERN provided loss maps for H,V,Skew halo
  with jaws at 7s and re-calculate energy
  deposition grids
• Study accident case
• Transverse extent of damaged region
• Longer term goal of upgrading to current CERN
  input structure with much richer description of
  all devices in tunnel
• For the moment, ask CERN for estimates of load on
  easier collimators