1
Low mass carbon based support structures for the HL-LHC ATLAS pixel forward disks R. Bates* a , C. Buttar a , I. Bonad a , F. McEwan a , L. Cunningham a , S. Canfer b , C. Lockett b , J. Pater c , R. Thompson c , T. Jones d a: SUPA school of Physics and Astronomy, University of Glasgow, Glasgow, UK b: Advanced Materials Group, RAL, STFC, UK c: School of Physics and Astronomy, The University of Manchester, UK d: Department of Physics, The University of Liverpool, UK * Contact e-mail: [email protected] Introduction Pixel mechanical test structure Pixel 2012: International Workshop on Semiconductor Pixel Detectors for Particles and Imaging , September 2012, Japan 0 5 10 15 20 25 30 35 -34 -33 -32 -31 -30 -29 -28 Position (mm) Temperature (deg C) Thermal testing and Simulations Proposed design Test object simulated and measured in dry air Thermal run-away on the test structure Thermal measurement test apparatus Thermal conductivity results Foam & CFRP Conductivity of BN filled Hysol 9396 epoxy Phi overlap on modules all disks on in ring to be placed on the same side gives simpler tape design - Lower contact area to high modules => Detailed FEA to take place Disks with have an inner radius of 130 mm and outer radius of 315 mm 3 rings of quad(hex) modules required. Services for middle ring will be a challenge Zoom showing space around modules Disk 4 Mixed module sizes (Hex, quad, quad) l measurements are steady state stems are verified against copper and silicon mperatures measured with 4 wire PT100s at from 4 wire electrical heaters Through-thickness measurements for glues and thin films Sample between Cu rods Pressure and DC340 at interfaces Measured at RT In-plane measurements for thin sheets Sample surrounded by radiation shield Under vacuum (10 -5 mbar) Measurements from -30C to +20C Plastic insulation supports Sample Aluminium thermal shield 1 Shiel d heate r 2 3 4 Pt100_5 H 2 O Wate r cool ed Heate r wires Pt100_ 6 Sampl e heate r Ancho r Heat sink Pt10 0 wire s Sample • ASIC power set to 1.56 W/chip (0.41 Wcm -2 ) • Nominal detector power ~ 3 mWmm -2 at 0C • Exact power depends on sensor thickness and desired collected charge • Factor of 4 in safety Cross-section of loaded cooling structure Silicon readout IC (150 μm thick) Solder bumps (20 μm thick) Silicon detector (300 μm thick) DC SE4445 gel (200 μm thick) CFRP K13C2U 50 gsm (200 μm thick) Allcomp Foam (2.3 mm thick) • Solder bumps at 5.3% fill factor • DC SE4445 100% coverage • No glue between foam and pipe • Input data from measurements • No glue between foam and pipe • HTC CO2 = 10k Wm -2 K -1 • HTC still air = 5 Wm -2 K -1 • Good agreement with data 0 2 4 6 8 10 12 14 16 18 -30 -28 -26 -24 -22 -20 -18 -16 -14 -12 -10 Qref T=0C (mW/mm2) Temperature (C) Pixel dummy heaters • 500um thick Silicon • 1um Tungsten with Cu solderable strip - Thermal simulation verification - Disk loading Allcomp foam 0 100 200 300 400 500 600 0 1 2 3 4 5 Glue thickness (um) Resistance (K/W) nterface resistance foam to metal ~ 0 K/W 0 50 100 150 200 250 0 5 10 15 20 25 30 35 Distanc e Temperature (C) Heat sink Q 500µm of hysol + 35% BN Senso rs Coppe r Coppe r Allcomp foam Glue penetrated into the foam Glue 1 Glue 2 • 35% BN by weight: k=1.4 W/mK • Degassing under vacuum makes little difference • Little change after irradiation IR camera plot of temp through sample k = t i k i t i = 2 k f FAW ρt for 90 direction K13D2U kf = 800 W/mK: K13C2U kf = 620 W/mK. Density = 2.19 gcm -3 CFRP K13D2U 0-90-0 100 gsm K13C2U 0-90-0 100 gsm Thickness (um) 230 254 0 - Predicted 318 229 ± 7 0 - measured 294 ± 20 285 ± 12 90 - predicted 159 114 ± 6 90 - measured 144 ± 20 Through thickness 1.20 & 0.96 1.3 & 1.1 • Large Hadron Collider (LHC) at CERN luminosity will increase by an order of magnitude around 2020 to the high luminosity HL-LHC • The ATLAS experiment will require a new tracker radiation damage and event density & rate • This tracker will consist of pixel and strip devices both as barrel and endcap disks. The new pixel detector Smaller pixels -> to minimize occupancy & increase spatial resolution Extremely radiation hard silicon detectors. New front-end electronics to cope with the higher data rates. These devices will be supported by low mass thermal-mechanical structures. • Enclosure at 20C flushed with N2 • CO2 from gas bottle - via chiller - & needle valve • Evaporation T = -33C • Measure with IR camera Test box / Cooling system

Low mass carbon based support structures for the HL-LHC ATLAS pixel forward disks

  • Upload
    pakuna

  • View
    37

  • Download
    2

Embed Size (px)

DESCRIPTION

Q. Low mass carbon based support structures for the HL-LHC ATLAS pixel forward disks. Copper. Glue 1. 500µm of hysol + 35% BN. Silicon readout IC (150 μm thick). Allcomp foam. Sensors. Solder bumps (20 μm thick ). Glue penetrated into the foam. Anchor. Aluminium thermal shield. - PowerPoint PPT Presentation

Citation preview

Page 1: Low mass carbon based support structures for the HL-LHC ATLAS pixel forward disks

Low mass carbon based support structuresfor the HL-LHC ATLAS pixel forward disks

R. Bates*a, C. Buttara, I. Bonada, F. McEwana, L. Cunninghama, S. Canferb, C. Lockettb, J. Paterc, R. Thompsonc, T. Jonesd

a: SUPA school of Physics and Astronomy, University of Glasgow, Glasgow, UK

b: Advanced Materials Group, RAL, STFC, UK

c: School of Physics and Astronomy, The University of Manchester, UK

d: Department of Physics, The University of Liverpool, UK

* Contact e-mail: [email protected]

Introduction Pixel mechanical test structure

Pixel 2012: International Workshop on Semiconductor Pixel Detectors for Particles and Imaging , September 2012, Japan

0 5 10 15 20 25 30 35-34

-33

-32

-31

-30

-29

-28

Position (mm)

Tem

pera

ture

(deg

C)

Thermal testing and Simulations

Proposed design

Test object simulated and measured in dry air

Thermal run-away on the test structure

Thermal measurement test apparatus

Thermal conductivity results Foam & CFRP

Conductivity of BN filled Hysol 9396 epoxy

Phi overlap on modules all disks on in ring to be placed on the same sidegives simpler tape design- Lower contact area to high modules => Detailed FEA to take place

Disks with have an inner radius of 130 mm and outer radius of 315 mm3 rings of quad(hex) modules required.Services for middle ring will be a challenge

Zoom showing space around modules

Disk 4 Mixed module sizes (Hex, quad, quad)

All measurements are steady stateSystems are verified against copper and siliconTemperatures measured with 4 wire PT100sHeat from 4 wire electrical heatersThrough-thickness measurements for glues and thin filmsSample between Cu rodsPressure and DC340 at interfacesMeasured at RT

In-plane measurements for thin sheetsSample surrounded by radiation shieldUnder vacuum (10-5 mbar)Measurements from -30C to +20C

Plastic insulation supports

Sample

Aluminium thermal shield

1

Shieldheater

2 3 4

Pt100_5

H2OWater cooled

Heater wires

Pt100_6

Sample heater

Anchor

Heat sink

Pt100wires

Sample

• ASIC power set to 1.56 W/chip (0.41 Wcm-2)• Nominal detector power ~ 3 mWmm-2 at 0C• Exact power depends on sensor thickness and desired collected charge• Factor of 4 in safety

Cross-section of loaded cooling structureSilicon readout IC (150 μm thick)

Solder bumps (20 μm thick)

Silicon detector (300 μm thick)

DC SE4445 gel (200 μm thick)

CFRP K13C2U 50 gsm (200 μm thick) Allcomp Foam (2.3 mm

thick)

• Solder bumps at 5.3% fill factor• DC SE4445 100% coverage• No glue between foam and pipe

• Input data from measurements• No glue between foam and pipe• HTC CO2 = 10k Wm-2K-1

• HTC still air = 5 Wm-2K-1

• Good agreement with data

0 2 4 6 8 10 12 14 16 18-30

-28

-26

-24

-22

-20

-18

-16

-14

-12

-10

Qref T=0C (mW/mm2)

Tem

pera

ture

(C)

Pixel dummy heaters• 500um thick Silicon• 1um Tungsten with Cu solderable strip - Thermal simulation verification - Disk loading

Allcomp foam

0 100 200 300 400 500 6000

0.51

1.52

2.53

3.54

4.5

Glue thickness (um)

Resis

tanc

e (K

/W)

Interface resistance foam to metal ~ 0 K/W

0 50 100 150 200 2500

5

10

15

20

25

30

35

Distance

Tem

pera

ture

(C) Heat sink

Q

500µm of hysol + 35% BN

Sensors

Copper

Copper

Allcomp foam

Glue penetrated into the foam

Glue 1

Glue 2

• 35% BN by weight: k=1.4 W/mK • Degassing under vacuum makes little difference• Little change after irradiation

IR camera plot of temp through sample

k =tiki∑ti∑

=2k fFAW

ρtfor 90 direction

K13D2U kf = 800 W/mK: K13C2U kf = 620 W/mK. Density = 2.19 gcm-3

CFRP K13D2U0-90-0100 gsm

K13C2U0-90-0100 gsm

Thickness (um) 230 2540 - Predicted 318 229 ± 7 0 - measured 294 ± 20 285 ± 12 90 - predicted 159 114 ± 6

90 - measured 144 ± 20Through thickness 1.20 & 0.96 1.3 & 1.1

• Large Hadron Collider (LHC) at CERN luminosity will increase

by an order of magnitude

around 2020 to the high luminosity HL-LHC• The ATLAS experiment will require a new tracker

radiation damage and event density & rate• This tracker will consist of

pixel and strip devices

both as barrel and endcap disks.

The new pixel detector

Smaller pixels -> to minimize occupancy & increase spatial resolution

Extremely radiation hard silicon detectors.

New front-end electronics to cope with the higher data rates.

These devices will be supported by low mass thermal-mechanical structures.

• Enclosure at 20C flushed with N2• CO2 from gas bottle - via chiller - & needle valve• Evaporation T = -33C• Measure with IR camera

Test box / Cooling system