larp lhc phase ii secondary collimator cd1 review. - e. doyle 15 dec. 2005 1/25 larp phase ii...

LARP LHC PHASE II SECONDARY COLLIMATOR CD1 REVIEW. - E. Doyle 15 Dec. 2005 1/25

LARP Phase II Secondary Collimator RC1 Review SLAC 12/15/05

Prototype Engineering Studies

503 618

522

35495

0 1406


LHC Phase II collimator concept development

• LHC & NLC collimators compared• FLUKA/ANSYS jaw thermal response simulations

– A major % of our effort so far– model evolution

• Baseline Jaw concept– Material selection– Cooling channel arrangement– Jaw support– Jaw geometry

• Baseline collimator concept– aperture control– jaw support/actuation– RF parts– cooling water supply

• Unresolved issues


Phase I Supporting System

Fixed base Fixed base

Adjustable baseAdjustable base

Pivoting cradle Pivoting cradle (manual (manual

adjustment +/-10º)adjustment +/-10º)

Jaw actuation Jaw actuation systemsystem

Vacuum TankVacuum Tank

SLAC will utilize as much of CERN support and actuation systems as possible – must work in horizontal (shown), vertical (90o) and skew (+/- 45o)

1.48 m


Phase I Mechanical design

Steppers & roller screws – jaw ends independent

Return Springs

C/C, copper, SST jaw, 1.2 m long

Vacuum tank

Bellows not shown (4x)

Coolant in/out

SLAC concept substitutes - rotate-able jaws for 1.2m long Phase I jaws

- Larger vacuum tank to allow max jaw diameter


NLC Consumable Collimator

6.0

Note short length of collimation material.

L/D = .02


Phase II Baseline layout

beambeam

•136mm diameter x 950 mm long jaws (750 mm effective length due to taper).

•Vacuum tank, jaw support mechanism and support base derived from CERN Phase I.


NLC & original LHC specs – major differences

Specification NLC LHC Comments

beam pipe ID 1cm 8.4cm LHC: two opposing beam pipes

gap range (full aperture)

0.2 – 2.0mm

0.5 – 45mm

Jaw diameter 318 mm 136 mm LHC: Function of beam pipe diameter/spacing and gap range

jaw length 10 cm total 6mm active L/D=.02

95 cm total, ~75 cm active L/D=5.5

LHC: length controlled by 1.48m flange-flange space and need for flexible transitions; thermal bending problem results from length

jaw deformation – toward beam

5 um 25 um NLC: short jaw => no bending; close coupled support controls effects of swelling

SS power, per jaw

~1 W ~ 12 kW Cooling NLC: radiation - 4C temp rise. LHC: water cooling, possible power densities in boiling regime

Bottom Line: LHC & NLC collimators are different animals

*

* This spec infeasible, has been relaxed


Water cooled

2-d & (3-d rectangular) model 3-d “Hollow cylinder” model

- Uniform or limited arc cooling

“Solid” model

Tubular cooling channels Uniform ID Cooling

– simulates helical or axial channels

H2O simulation

– helical flow shown

Progression of ANSYS models – increasingly realistic

beam


Material thermal performance (Hollow Cylinder Model)- O.D = 150 mm, I.D. = 100 mm, L = 1.2 m- NLC-type edge supports

material cool

ing

arc

(deg

)

pow

er (k

W) p

er ja

w

Tmax

( C

)

defl

(um

)

Tmax

wat

er s

ide(

C)

pow

er (k

W)

Tmax

( C)

defl

(um

)

Tmax

wat

er s

ide(

C)

BeCu (94:6) 360 0.9 24 20 4.3 41 95Cu 360 10.4 61 221 43 52.0 195 829 117Cu - 5mm 360 4.5 42 117 39 22.4 129 586 117Cu/Be (5mm/20mm) 360 5.3 53 161Super Invar 360 10.8 866 152 60Inconel 718 360 10.8 790 1039 66 54.0 1520 1509 85Titanium 360 7.4 214 591 42 36.8 534 1197 77Tungsten 360 14.9 183 95 79 74.5 700 335 240Al 360 3.7 33 143 18.5 73 5272219 Al 360 4.6 34 149 26 23.0 79 559 46C R4550 graphite 360 0.6 25 5 3.0 41 20

10s, primary debris + 5% direct hits

SS @ 1 hour beam life transient 10 sec @ 12 min beam


material reasons for rejection in favor of Cu

BeCu (6% Cu-loaded Be)Be is prohibited by CERN management, except when no alternatives exist; low cleaning efficiency; fabrication difficulty

Super Invarpoor thermal conductivity => high temperature (866C); desirable properties (low thermal expansion coefficient) disappear at 200C

Inconel 718

poor thermal conductivity => high temperature (Tmp = 1400C < 1520C transient peak) & very high deflection (1039um SS, 1509um transient)

Titanium poor thermal conductivity => deflection 2.7 x Cu (591um, SS)

TungstenHigh temperature on water side (240C => ~30bar to suppress boiling); high power density - can't transfer heat without boiling

Aluminumrelatively poor cleaning efficiency, water channel fabrication difficulty

Cu - 5mm walldeflection only ~50% lower than 25mm Cu, loss of safety zone between surface and water channels

Cu/Be (5mm/20mm)deflection only ~30% lower than 25mm Cu; Be prohibition; fabrication difficulty

Cu chosen as best balance between collimation efficiency, thermal distortion & manufacturablity


Central Aperture Stop

-Swelling neutralized

-Bending neutralized

NLC-type Aperture control

- Swelling neutralized

- Bending toward beam

Shaft support (Phase I)

-Swelling toward beam

-Bending toward beam

Aperture-defining support

Choice of Aperture-Defining Scheme - To Minimize Bending and Swelling Deformation toward Beam


61C

Note transverse gradient causes bending

Interesting side trip: 64% less distortion if cooling is limited to a 36o arc centered on beam path.

89C

Note axial gradient

x=221 m

Spec: 25msupport

support

x=79 m

Note more swelling than bending

64% less distortion

360o full I.D. cooling 36o arc cooling


Limited cooling arc: free wheeling distributor – orientation controlled by gravity – directs flow to beam-side axial channels.

Pro: Far side not cooled, reducing T and thermal distortion.

Con: peak temperature higher; no positive control over flow distributor (could jam); difficult fabrication.

360o cooling by means of helical (or axial) channels.

Pro: Lowers peak temperatures.

Con: by cooling back side of jaw, increases net T through the jaw, and therefore thermal distortion; axial flow wastes cooling capacity on back side of jaw.

water

beam

Helical and axial cooling channels illustrated

beam


Helical cooling passages chosen for manufacturablity, beamline vacuum safetyPer CERN’s Phase I design – no water-vacuum weld or braze

• 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:

1. Pitch may vary with length to concentrate cooling

2. Two parallel helixes to double flow

3. Spacer between coils adds thermal mass, strength

4. Electroform jaw body onto coil


Collimator TCSM.A6L7 Cooling scheme Helical Axial (36o)

# channels 1 2 Diam (m) .008 .006 Velocity (m/s) 3 3

Cooling

Total flow (l/min) 9 10 SS Power (kW) 11.7 Beam heat Trans Power (kW) 58.5

Jaw peak 86.5 91.5 Cooling chan. peak 68.3 69.7

SS

Water out 36.0 36.1 Jaw peak 231 223 Cooling chan. peak 154 130

Temp (C )

Trans

Water out 43.6 47 SS 394 107 Deflection (um) 4 Trans 1216 778 SS 43 75 Eff. length (cm) 5 Trans 24 31

Max Cu temp 200

Possible boiling

Max water return temp

Deflection 325 & 750 (SS & trans)

All temperature simulations based on 20C supply. For CERN 27C supply add 7 to all temperature results. CERN max water return temp 42C

Exceeds spec, or other possible problem as noted

Baseline Jaw Performance


Estimates for downbeam collimator heat loading and deflection – scaled from worst case baseline (green)

Steady State Transient No. name alias Power

(kW) Defl (um)

Eff. Lngth (cm)

Power (kW)

Defl (um)

Eff. Lngth (cm)

note

1 TCSM.A6L7 TCSM1 11.7 394 43 58.5 1216 24 simulated 2 TCSM.B5L7 TCSM3 2.7 137 75 13.5 422 46 scaled 3 TCSM.A5L7 TCSM5 .69 35 75 3.44 108 75 scaled 4 TCSM.D4L7 TCSM7 .18 9 75 .92 29 75 scaled 5 TCSM.B4L7 TCSM11 .20 10 75 1.01 31 75 scaled 6 TCSM.A4L7 TCSM13 .19 10 75 .96 30 75 scaled 7 TCSM.A4R7 TCSM15 .16 8 75 .81 25 75 scaled 8 TCSM.B5R7 TCSM21 .22 11 75 1.10 34 75 scaled 9 TCSM.D5R7 TCSM25 .19 9 75 .93 29 75 scaled 10 TCSM.E5R7 TCSM27 .13 6 75 .62 19 75 scaled 11 TCSM.6R7 TCSM29 .25 12 75 1.23 26 75 scaled


Jaw diameter – limited by vacuum tank size and required range of motion. Vacuum tank size limited by proximity of opposing beam pipe in all collimator orientations.

Jaw Diameter Determined


•jaw tends to bow toward beam due to heating. Stop prevents reduction of gap

•Jaw ends spring-loaded to the table ass’y (next slide) … move outward in response to bowing

•Shallow groove => smooth contact surface safe from beam accidents

•May use two stops to control tilt

Adjustable central gap-defining stop

Stop in/out position controls aperture, actuator external, works through bellows.


Self aligning bearing

Leaf springs allow jaw end motion up to 1mm away from beam

Adjustable central jaw stops (previous slide) define gap

- Flexible bearing supports allow jaw thermal distortion away from beam

CERN’s jaw support/positioning mechanism. Vacuum tank, bellows, steppers not shown.

Flexible end supports used in conjunction with central gap-defining mechanism


Rigid round-square transition

Spring loaded fingers ground two jaws through range of motion

RF contact overview

Sheet metal parts flex to follow jaw motion

Clearance problems to be resolved

Concept satisfies CERN RF requirements

- Need sufficient contact pressure

Cooling issues not addressed


CERN design:

Jaw supported on individually moveable shaft at each end, controlled by steppers external to tank.

Bellows allows full range of jaw motion

Continuous one-piece cooling tube brazed to jaw, exits tank at each end through shaft.

Unresolved interferences between RF parts and cooling and support parts

Jaw rotation mechanism not devised

Substantial forces to rotate jaw

Mandrel to support coil not shown

Flex Cooling Supply Tube Concept

Space required for flexible water connections results in 95 cm maximum jaw length.


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.

Stub-shaft (bearing not shown)

Support shaft

relaxed (as shown) # coils 4

O.D. 111mm (4.4in)

full 360o rotation # coils 5

O.D. 91mm (3.6in)

torque 9.1N-m (81in-lb)

Detail of Flex Cooling Supply Tube


ANSYS simulation: Axial stress for un-grooved and grooved jaw with axially uniform heat input.

Case Tmax °C Deflection (um)Jaw edge ref axis ref

Straight 59.5 33 ~100

grooved 59.5 15 ~74

Grooves reduce bending deflection

Note: RF taper requirements may make this concept un-feasible


Specifications for baseline Phase II collimator

spec value Beam sigma 200um location Centered in pipe +/- 5mm Beam pipe Spacing 224mm c-c opposing beampipes diameter 88mm OD clearance 8mm vacuum tank to opposing beampipe Jaw Length 95cm including 10cm tapers on ends Diameter 136mm Material Copper cooling Embedded helical channel cooling No water-vacuum joints if possible Special features Circumferential slots to reduce thermal-induced bending, if no RF problems deformation <25um toward beam; <325mm away, steady state; <750um away, 10 sec transient Peak temp. 200C operating, 250C bakeout Range of motion 25mm per jaw, including +/- 5mm beam location drift Damage extent 15mm Aperture stop Range of motion Positively controls aperture from 5-15 sigma (2-6mm full aperture), must float +/-

5mm as jaws are moved to follow beam drift Heat load Steady state 11.7 kW Transient 58.5 kW Vacuum pressure <1e-7 Pa (7.5e-10 Torr) Vac. tank length 1.48m flange-flange flanges CERN quick disconnect Clearance Clears opposing beampipe with +/- 10o adjustment in all orientations Cooling Supply 27C return 42C max RF contacts configuration Sheet metal parts per Figures 7-9 subject to CERN approval

*

* Relaxed from original spec

baseline design deviates


Unresolved Issues

Jaw actuation mechanismHow to handle mass of rotary jaws (fail open springs)Availability of CERN actuation mechanism for SLAC use is being discussed

Jaw rotary indexing mechanismforce to rotate jaws acceptable?concept not developeddo we know angular position of jaw at all times?

RF parts – taper requirement details not clearcentral groove in jaws (smooth track for central aperture stop)strain-relieving grooves in jawswhat is the acceptable range of taper angles for the jaw ends

Heat generation in thin RF partsNeed details of CERN support stands, etcEffects due to accident

does accident cause unacceptable gross distortion of the jaw?do RF fingers work in contact with damaged surface?How much material melts and where does it go? – depends on jaw orientationis central aperture stop safe from contamination by melted material?Beam tests may be required


LHC Phase II Collimation

BONUS SLIDES


Phase I Jaw - Design Principles1. Jaws in C/C or graphite2. Cooling Cu-pipes (2 x 3 turns) and plate pressed against the jaw, brazed to the bar.3. Main GlidCop® support bar

Glidcop support bar

Cu cooling plate c/c


Material Properties


Heat Transfer by Boiling Water


Construction possibility – jaw with axial cooling channels


NLC Aperture-Defining GeometryOne independent variable (stop roller spacing, s) defines

aperture.

s = stop roller spacing


NLC Jaw Indexing Mechanism


beam

beam

NLC suspension and aperture-defining mechanism: limitations in LHC environment. 150mm jaw full 30 mm retraction showing interference with opposing beam pipe. Vac envelope with pass-through for beam 2 as possible solution.

larp lhc phase ii secondary collimator cd1 review. - e. doyle 15 dec. 2005 1/25 larp phase ii...

Documents