HEP GROUP

1
HEP GROUP MEETING 18.12.07 work on the ATLAS
UPGRADE T.J.Fraser
2
ATLAS Inner Detector Upgrade summary
Replace the existing TRT, existing SCT Barrel and
end caps and pixel detector with possibly 5
barrels SCT and discs at ends, and new
pixel detector inside SCT barrels. For the SCT
this will mean lots more modules. Modules are at
a conceptual stage maybe with hybrid with 40
ASICs connecting to a power bus on the edge.
These would operate at high speed using 160MHz
clock with serial readout of ABCN but there are
various optionseg parallel option clocked at
slower 40 MHz. SERIAL POWERING is favoured as in
theory it would save a lot of material but this
will need serious risk mitigation strategies to
protect against failures eg OPEN in serial chain
loss of whole stave!, NOISE, SHORTS etc so must
be able to isolate a module. One stave will have
10 modules minimum. Irradiation testing of
Readout materials have started, more studies in
2008. Module sensors silicon strip detector
specification made, irradiation started, testing
from March 2008. Date for Final Design Review
beginning of 2010 for Barrel followed by
Production Readiness Review. Finish production
for Barrel modules April 2012. LHC SHUTDOWN
scheduled for 2015 2016 with SLHC running
from spring/summer 2016. Plan to add early
separation dipoles maybe IN detectors at around
6m from IP. Maybe add CRAB CAVITIES at small
angle.
3
ATLAS Inner Detector Upgrade summary
35
To reach 10 , increase beam current or change
bunch cross section and crossing angle. RF
cavities located around the IP or in 2 locations
of the LHC. 25ns spacing versus 50 ns spacing
being debated. Radiation NEUTRON MODERATOR for
INNER TRACKER likely will impact on space
envelope inside cryostat. The baseline is 5cm
lining the calorimeters. 5cm is a LOT of space
but this seems to be the optimum to reduce
fluences but cant be hermetic places where
services go through prevent this. IP magnets and
the ID reach radiation damage limit at 700fb-1
which could be at around 2014 so cant
delay! Various dates on the schedule (rough
estimates) are General TDR (Technical Design
Review) 2010 CONCEPTS FIXED ASSEMBLE PARTS
2011 2013 ASSEMBLE STRUCTURES on surface 2012
2014 INSTALL IN PIT 2015 STARTUP 2016 SERVICES
REVIEW not before March 2008 lots of work
what services can be reused reliability
available space. We are setting up a working file
for services materials which contains info
mass/space occupation etc.
4
ATLAS UPGRADE ID BARREL END REGION TJFraser WP7
meeting 08.10.07

• Looking at impact of services routing off the
  barrel end
• with current proposed layout of ID..
• Looking at Thermal barrier feedthroughs as part
  of
• supermodule assemblies
• Constraints on services Hot vs Cold gap options

5
LS LAYERS 1 and 2 R 950, R750, L 3800
SS LAYERS 1, 2 and 3
R 600, R 490, R 380, L 2000
6
services from inner barrels must route 900mm to
outer barrel end
7
services will line end of short barrels and
inner surface of outer layers
8
services from inner barrel must share space with
wheel sections inside outer barrel
9
Thermal barrier sandwich
Feedthroughs (as part of supermodule assembly,
one per supermodule) fit into gaps in the
thermal barrier. Layout of gaps would depend on
the siting of and shape of the thermal
barrier. Shape of feedthroughs would depend on
type of services if LMT type then a rotational
aspect could be included. Multi-service
feedthroughs like this could only be used if
sited close to detector end.
10
Feedthrough could be designed to come apart
or have holes large enough for connectors, with
separate seals
11
services separated into groups for channeling
outer barrel
wheels
inner barrel
services from barrel
separate thermal barrier
thermal barrier
in warm gap
insulated multi-feedthroughs one per supermodule
WARM
12
thermal barrier
outer barrel
wheels
inner barrel
services from barrel
thermal barrier
in cold gap. No feedthroughs here but would
need patch panels for readout/TTC PCBs somewhere
in cold gap.
insulated single- service feedthroughs
COLD
13
GENERAL LAYOUT Services from barrel on current
strawman layout follow a tortuous route
presents mass where not wanted and difficulties
of access during and after installation. 5
barrels all the same length would present fewer
problems for services and less mass as services
gap would be shorter. WARM vs COLD GAP Warm
gap more space needed for thermal barriers,
active cooling pipe insulation and feedthroughs
also means more material where not wanted. Cold
gap services and thermal barriers take up less
space and therefore less mass but would still
need space for patch panels for some services
inside due to limits on lengths and transitions.
No insulation needed on cooling (?) Size of
single entity installation difficult if not
split up. NUMEROUS POSSIBILITIES (or should this
be impossibilities?) for barrel end layouts until
basic decisions are made about layout and thermal
management.
CONCLUSIONS
14
Barrel services minimum space allocation in Z
Strawman layout and Spider layout
WARM and COLD GAP versions
Services envelope in Z In order to reserve space
allocation - need to use places in the layout
where the maximum accumulation of services in
Z occur cannot use an average as it will not be
possible to squash services into gaps to equalise
the occupation. Need to know which existing
services are to be kept and if the
existing channels/ducts have to be re-used -
before making useful layout in R/PHI for barrel
ends.
example x-section in Z for services
accumulation, barrels 1, 2 and 3
cooling connectors
bus tapes/cables with twist factor
thermal barrier feedthroughs
uninsulated cooling pipes manifolds

24
29mm
10
21
84mm
TJF
15
Barrel services minimum space allocation in Z
(Strawman layout)
COLD GAP version
B5
B4
B3
B2
B1
Z0
0
How will services from Bs 1,2 and 3 be supported
on the inside of B4? Will need separate support
cylinder or rings for this, adding to
space occupancy and material. Services at barrel
ENDS also need supports for connectors/strain
relief.
84mm in Z
gt160mm in Z

gt150mm in R (insufficient clearance)
services in Z bus tapes, connectors, cooling
pipes, connectors and manifolds
TJF
16
Barrel services minimum space allocation in Z
(Strawman layout)
WARM GAP version
B5
B4
B3
B2
B1
Z0
0
Thermal barrier will need to serve as services
support as well if not, then add this to
thickness. Cooling exhaust pipes will need
insulation 6mm thick ie 12mm added here.
96mm in Z
gt172mm in Z

gt162mm in R (insufficient clearance)
TJF
services in Z bus tapes, connectors, insulated
cooling pipes, connectors and manifolds
17
Barrel services minimum space allocation in Z
(Spider layout)
COLD GAP version
B5
B4
B3
B2
B1
Z0
0

130mm in Z
Advantages of simplified barrel layout avoids
two 90 bends in services route just one large
services spider so can organise services into
channels only one type of support needed
for connectors/strain relief at barrel ends
0
84mm in Z

services in Z bus tapes, connectors, cooling
pipes, connectors and manifolds
TJF
18
Barrel services minimum space allocation in Z
(Spider layout)
WARM GAP version
B5
B4
B3
B2
B1
Z0
0
Thermal barrier will need to serve as services
support as well if not, then add this to
thickness. Cooling exhaust pipes will need
insulation 6mm thick ie 12mm added here. Not as
good as the Cold gap version but better than both
Strawman layouts
142mm in Z
96mm in Z

TJF
services in Z bus tapes, connectors, insulated
cooling pipes, connectors and manifolds
19
ATLAS Tracker Upgrade - Services at Barrel Ends
Scenario where services are routed through
existing services channels on cryostat shown on
the next 3 slides ie Old TRT channels used for
fibres, power and sensor cables and input cooling
pipes Old SCT cooling exhaust channels used for
same purpose in the Upgrade Could possibly work
for 108 supermodules (with lots of manifolding
for cooling pipes) Wouldnt work if outer barrels
were included these would need to use existing
power cable channels but there would be no space
in the existing cooling exhaust channels, so
new channels eg one per quadrant would have to be
created in order to keep each set together for
maximum cooling efficiency. Design of layout on
the cryostat is crucial to the design of the
layout on the barrel ends!
20
7
7
rows per Quadrant
6 exhaust pipes to old cooling channel
in cryostat
exhaust manifold
11
6
9
input pipes in TRT channels (need manifolds)
5
7
4
7
Evap. cooling routing off barrel end.
5
6
4
input exhaust exhaust manifold
3
45.0
22.5
3
11.25
half length cooling loops
TJF 28/06/07
21
rows per Quadrant
exhaust cooling only
11
9
power and sensor cables go in old TRT channels
7
power and sensor cables in 4 dedicated channels
per quadrant shown as one bunch per supermodule
45.0
22.5
11.25
TJF 28/06/07
22
exhaust cooling only
rows per Quadrant
11
9
power, sensor cables, fibres and input pipes go
in old TRT channels
7
single fibres
ribbon
optofibre routing one fibre from each PCB
joins one 12 way ribbon 9 ribbons per quadrant
45.0
22.5
11.25
four PCBs per supermodule
TJF 18/07/07
23
CURRENT ATLAS SCT BARREL view of end barrel
services for the 4 barrels these form a dense
thicket on the barrel ends and beyond. The
upgrade will try to avoid this and minimise mass
and complexity.however this wont be easy!
24
CURRENT SCT BARREL installation of one barrel
inside the others.
Large services support structure necessary to
store long lengths of services and connectors
within the profile of the barrel. The
services radiating from the barrel on the
radial support have to be folded into
this structure for integration with the TRT
- Transition Radiation Tracker
Silicon modules on cylindrical support
cylinder made of carbon fibre.
25
CURRENT SCT BARREL and TRT being installed in the
cryostat in the ATLAS pit. It slides in on
side rails.
The orange painted cradle is removed once the
SCT is installed.
This is the SCT services support structure.
The SCT barrel is hidden inside the TRT barrel
26
Conclusion It has taken over 10 years to arrive
at the stage where the Inner Detector is ready to
run in the present form with TRT and SCT barrel
and end caps all worked on in parallel by
different Groups - but there will only be 6 years
to produce the Upgrade version, so radical
departures in philosophy and design are unlikely
to be chosen unless they present a
simplification of the current design!