LHCC referee meeting L.Rossi - INFN / Genova CERN - March 19 - 2001 1 Status of insertable pixel Layout in 2000 (Dubna layout) Reasons for the change

LHCC referee meetingL.Rossi - INFN / Genova CERN - March 19 - 20011

Status of insertable pixelStatus of insertable pixel Layout in 2000 (Dubna layout) Reasons for the change (and boundary

conditions) Design of insertable pixel

layout insertion method service routing and patch panels

Simulation results coverage and resolution material

Road to ECR


Dubna layout Dubna layout 3 barrel layers

(Rb= 5cm,R1=9.3cm ,R2=12.7cm ) and 10 disks (6 with 11 sectors and 4 with 9 sectors).

Services at high R, except for b_layer running at R~8cm to allow for independent insertion of b_layer.


Reasons for changeReasons for change Rad-hard submissions to DMILL (tested in detail in 2000) revealed

unexpected problems for our FE design. After some trials with DMILL, migration to DSM technology was judged necessary. This has generated a schedule delay of at least 15 months and made impossible (or very risky) to install the pixel inside the barrel ID on the SR building (on 4/2004).

Installing pixel separately from ID gives also more freedom in case of production problems (more likely for a novel detector) and allows easier upgrade & repair (all vertex detectors undergo upgrade sooner or later).


Layout updateLayout update Why we did not propose it before?

mechanics more difficult (and a lot of redesign if half-shelled) more material (services at lower radius, more massive mechanics) conflicts with SCT (inner bore + alignment)

A solution was found to keep the present design approach (shell and disks) and install pixel in absence of the beam pipe section just outside the barrel, i.e. break vacuum (as in long shutdown scenario), remove section of b_pipe,

and install pixel during ~2 months (<< long shutdown timescale)

The layout redesign implies some shrinkage after envelope agreement with SCT [smaller radii by ~5mm (R1=8.7cm ,R2=12.2cm) and less and smaller disks (3+3, 8 sectors each)] less modules to build (2094 1744; i.e. – 17%) less robust PR at large . some (~2%) acceptance losses for 3-hit coverage.

Further optimization difficult without loosing one hitshrunk from outside (SCT bore) and inside (beam pipe)

and in-the-middle (requirement of b_layer independently extractable rails)


Insertable layoutInsertable layout All services go at low radius up to the barrel end b_layer services all out on one side

keep independent b_layer insertion/extraction

1400m

m=2=2

=2.5=2.5


Barrel design also updated to slightly increase clearance between staves safer mount/dismount and pigtail routing (clearances~1.5mm (1 for

b_layer)) Flex -hybrid module (as opposed to MCMD) adopted also for b_layer


Design of insertable pixelDesign of insertable pixel A permanent CF tube is installed with ID

this allows to slide in pixel+services sliding in through the beam pipe flange

SCT

TRTTRT

34

50m

m

3400mm (end-plug face)

PP0Overlap regionZ=[750,1070]

PP1In End-Plug

Z=3400

Length of service panelPixel + Pigtails

795mm (Barrel Tube half length)

Pixel Frame is 1400mm longWithout pigtails

Side C Side A


The insertion process is done when the Forward Calorimeter is retracted and the relative beam pipe section is out.

First the far services and the pixel frame are inserted

Far Services

Far ServicesPixel Frame

Beam pipe support

Inner Detector Bore

Beam pipe

Far Servicesin Quadrants 2.65m

Services need to extend beyond end of Beam pipe support frame to allow termination of overlapping pigtails

Z=3450 (this bears some investigation)500mm

(comfortableHuman-width)

2270mmTo end of Pigtails


Far services terminated, pixel detector inserted

Far Services

Far Services

Terminate Far Services

2270mmTo end of Pigtails

300-500 for Cantileveredsupport

2770mmOver mechanical

supportZ=3450

Far side barrel pigtails need to fold back over frame to give access to PP0

Z=

6220

mm

Z=

5720

Cantilever structure necessary to support beampipe during installation—same as for ID forward installation


Then the near services are connected and slided in

Far ServicesNear Services

3400 (end of support tube)

Support wire at 3450 re-attachedFrame end just inside tube(Frame end to PP0 gap is 50mm)

(2.65m + 50mm)

2.7m6.1

m

Far Services

Termination to B-Layer services occurs inbound of Support wire


Assembly order require connections at different locations (PP0 and PP1) but simplifies PP2

Pigtail

PP1

“Type I” 2.7m

“Type II” 3.2m to PPB2

PPB1

Pixel Volume

Patch PPF1

To PP3

PP0 Location

PPB2

No Longer Used

Voltage drop on low mass cables becomes more challenging.Voltage drop on low mass cables becomes more challenging.Timed well with DSM change (everything changes)Timed well with DSM change (everything changes)


The b_layer must be clam-shelled as it does not pass through the beam pipe flange without opening. the b_layer can (in principle) be inserted without breaking the vacuum.

Far Services

1.2m

Far Services

TOP View

VerticalSupport

Halves are inserted around wire--requiredtooling is not shown

Halves are closed and fixed together

B-layer is inserted into Pixel Frame using samerails in pixels as for baseline


Cartoon of b_layer insertion

Preliminary design of tooling structures is consistent with space available now. Length of 1.2m required

Halves are held together by longitudinal actuation


Finish of b_layer installation

Far Services

Far Services

To terminate the services to the B-layer, and the rest of the detector, the detector must be withdrawn to gain access

B-layer services are terminated first as they will beobscured by the rest of the pixel services

The B-Layer is then pushed into the frame into its final position. Depending on the length of the Pigtails on theB-layer, this step may be avoided

B-layer Services


Patch Panel Octant Name

Barrel Layers 1&2 Disk

Tube Total

6-Module Bundle

7-Module Bundles

Bundle Total

Staves Serviced

Sectors Serviced

1A 3 2 5 (3+6)=9 6 15 12 3

2A 3 1 4 (3+6)=9 6 15 12 3

3A 3 2 5 (3+5)=8 5 13 10 3

4A 3 1 4 (3+6)=9 6 15 12 3

5A 3 2 5 (3+6)=9 6 15 12 3

6A 2 1 3 (3+5)=8 5 13 10 3

7A 3 2 5 (3+6)=9 6 15 12 3

8A 3 1 4 (3+6)=9 6 15 12 3

1C 3 1 4 (3+5)=8 6 15 10 3

2C 3 2 5 (3+6)=9 6 15 12 3

3C 2 1 3 (3+6)=9 5 13 12 3

4C 3 2 5 (3+6)=9 6 15 12 3

5C 3 1 4 (3+6)=9 6 15 12 3

6C 3 2 5 (3+5)=8 5 13 10 3

7C 3 1 4 (3+6)=9 6 15 12 3

8C 3 2 5 (3+6)=9 6 15 12 3

1CBL 2 4 4 8 4

2CBL 3 6 6 12 6

3CBL 3 6 6 12 6

4CBL 3 6 6 12 6

Sid

e A

Sid

e C

Sid

e C

B

-La

ye

r

7A6A

8A

2A

4A 1

A

5A

3A6C7C

5C

3C

1C 4

C

8C

2C

4CBL

2CBL

3CBL

1

1CBL

Needs update for new stave count (90)

Services routing (octants)


PP0PP0

Radius of PP0 is approximately 180mm Starts at Z=750 and goes to Z=1070 (50mm gap between PP0 and

Frame)

Z_400.7Z_440

65

0

49

5

58

0

R_190

R_90

10

0m

m

205mm

260mm

50mm120mm

Z_7

00

130mm

50mm

30mm

160mmBi-Stave Assy

6X 7X

370mm

En

d o

f Pix

els+

Pig

tails

Z_1

070

End of Pixel Frame


Rail system on support tubeRail system on support tube

Vee and Flat rails were chosen to provide pseudo-kinematic support for the detector during delivery to the support points

Rails are used only for delivery, not support


Support tube mock-upSupport tube mock-up

Sufficient space acquired in old Bevatron Generator Room (@LBNL) Enough space to simulate entire assembly sequence

Mockup in three pieces to simulate independent parts of tube Goal is for full length of entire tube to simulate all installation scenarios Detector rails are removable, should modification be necessary


Scope of tube mock-upScope of tube mock-up Mass and Envelope Geometry of final detector Frame

Attempting for similar friction as well At least two full quadrants of “dummy” service panels

Initially 1-quadrant, both sides, eventually half of all services At least two octants (both sides) of mechanically accurate

connectionNeeded for installation simulation

At least one octant of electrically active service connectionsThis is to provide verification that terminations stay terminated through

procedureMight prove useful for thermal mockup

Provision for Dummy B-Layer installationRequires also mechanical connections for installation simulation

Expected to be ready (tube + services) by end of March


Simulation resultsSimulation results Coverage

first disk cannot come too close to barrel end (service routing)

Barrel-to-disk two hit hole

Assume all tracks Assume all tracks coming from (0.0.0)coming from (0.0.0)


Geometric hit losses (not including electronics , dead channels)

Vertex smearing includedVertex smearing included


200 GeV 200 GeV Resolution does not change appreciably

e.g. in b_layer


200 GeV 200 GeV Transverse impact parameter slightly improves

MCMD --> flex (51.5 to 50m pixel size), smaller radii of layers 1&2 and larger stave tilt angle

Longitudinal i.p. worsens at high less disks and algorithm not yet tuned for long clusters in b_layer


1 GeV 1 GeV No change of i.p. resolution vs Dubna layout, sensible change vs

physics TDR layout mostly due to increase of the b_layer radius (4 -> 5cm) beam pipe X0 (1mm Be -->1.6mm Be) and radius (24 -->38mm) increase

not yet included in simulation (new beam pipe not yet released).


MaterialMaterial increase mostly due to support tube + services at low radius

mean increase up to ~2 is ~0.6%X0


Material with less layersMaterial with less layers We will likely start-up with 2 layers


Low-energy tails in EM calorimeter spectraLow-energy tails in EM calorimeter spectra Study of effects on EMC and on b_tagging just started

Tails event fraction with Erec < 0.92 Etrue

Layout Tails =0.3 Tails =2.4

Physics TDR 9.5 0.6% 3.6 0.6% Insertable 11.3 0.7% 9.6 0.9%

Vertex spread includedEvents over full and =0.025 (1 cell)

Electrons ET =10 GeV = 2.4

Normalised to same number of events


Steps to ECRSteps to ECR On design/mock-up

April: put together mock-up support tube and dummy services June 18: Comprehensive Design Review of global supports + insertion

tube June-Aug: electrically active tests and thermal simulation of mock-up October: mechanical testing of assembly procedure

On simulation compare reconstruction efficiency and b_tagging with TDR study effect of material increase on tracking and EM calorimetry (and

reduce material whenever possible). Time-scale for ECR: Summer 01.

Documents

LHCC referee meeting L.Rossi - INFN / Genova CERN - March 19 - 2001 1 Status of insertable pixel Layout in 2000 (Dubna layout) Reasons for the change