All queries about the ATLAS SCT Barrel Detector Modelling, Prototyping, Assembly, Integration and Installation Web Page - please contact Janet Fraser on:
UCL Phone [00-44]-(0)20 7679 3441
CERN Phone 00 41 22 7672020
UCL FAX [00-44]-(0)20 7679 7145
e-mail tjf@hep.ucl.ac.uk
The first three episodes of a series of films following just one of the teams of physicists involved in the research at the LHC.
The latest episode (4) in the series including an update on the 'incident' that is delaying work at the LHC.
The ATLAS Detector showing the SCT tracker surrounding the pixel detector
The latest enews about the ATLAS Detector at CERN
Current ATLAS MEETINGS at CERN
What's on at CERN
ATLAS Upgrade at the SuperLHC
During 2007 work started on the UPGRADE of the ATLAS Inner Detector for the Super-LHC! The Luminosity Upgrade aims for a x10 increase in luminosity using a new PS
(proton synchrotron) and a new Linac, Linac 4, on the CERN site. The ATLAS Detector will also require upgrading to accept higher interaction rates (about 400 events per bunch crossing). Trigger and Data-AQuisition systems will need upgrading, and the ATLAS Inner Detector will be completely replaced.
See preliminary work by TJF on ATLAS Upgrade ID off-barrel services
as part of WP7 UK group
Map showing how to get to UCL from Gatwick and
Heathrow
(NB: ADD one hour onto your journey time until further notice!)
The CURRENT ATLAS SCT barrel detector at the LHC, a long history:
2007 saw work on connecting up and testing of ATLAS BARREL SCT detector services. This work is recorded here.
In 2006 work at UCL and CERN concentrated on finalising the INSTALLATION SEQUENCE of the SCT Barrel part of the INNER
DETECTOR in the ATLAS CAVERN, together with feasibility studies concerning
last-minute changes in services and routing at CERN. The installation and connection of SCT BARREL READOUT fibres at the ROD racks situated in USA15 of the ATLAS cavern is also inlcuded.
Dog-leg design used for model was Univ. Geneva 252149P2 G.Barbier (1998)
25/01/07
The .ppt file 'an Illustration of R/O Fibre Connection Sequence' 25/01/2007
can be found at
http://www.hep.ucl.ac.uk/~tjf/fibreinstz4.ppt
This is an updated version of the original dated 20/12/06.
Here are some comments on the R/O fibre connection (in progress):
The R/O fibre installation in the cryostat was very well done, with fibre bunches in the correct
order (see entry 03/04/06 etc) for connecting up to the R/O fibres from the SCT barrel.
Some TTC Y sections were the wrong length but this was known about (see entry dated 08/03/06).
A photo of the neat bunching of R/O fibres installed in the cryostat.
A close-up of the R/O fibre bunches in the cryostat. Note that these fibres
were pre-installed and are up to 125 meters long extending from the cryostat region near the centre of
ATLAS out to electronics racks in cavern area USA15. The connection of these R/O fibres to the ROD (DAQ)
crates is recorded - see entry dated 18/08/06.
Unforseen problems in connecting R/O fibres were:
Some of the R/O fibres from the barrel had labels stuck on with kapton tape which were too
close to the MT connectors, where the ribbons have to make a 90 degree bend out to the
connector housing. It was considered too dangerous to try to move the labels.
Here is a photo of the first octant of fibres to be connected showing
some Kapton covered labels very close to the MT connectors in the housing in Quadrant C2.
A general view of both sets (octants) of R/O fibres either side of the HEX
near the housing in Quadrant C2.
A photo of the HEX (5 pipes) in Quadrant C2 with the two sets of
R/O fibres either side of the HEX (one set for each octant)
The bulkhead region of the HEX C2 showing capillary connections and fibres
at the TE (Thermal Enclosure) bulkhead. The fibres in furcation tubing are to be fixed in position
over the LMTs on the Inner Heat Spreader Plate, before Outer Heat Spreader Plate is fitted.
Some of the R/O fibre ribbonised lengths from the Barrel were non-standard making the positioning
of the splice heatshrink sections very difficult.
The available space envelope in Z allows for a ZERO packing factor (part of the SCT End Cap detector
extends into what could have been contingency space).
LMT cooling pipes occupy space in Z underneath the LMTs and extend into the spaces between the
LMTs on the HSP surface.
A general view of LMTs over cooling pipes on the Heat spreader plate
Radial Stiffener at the far left and the Heat ExChanger on the right of Quadrant C2.
The splice Heatshrink sections have to fit in
the space together with the LMTs in places where there are no gaps between the LMTs.
The HSP (Heat Spreader Plate) radial stiffeners extend in Z and fill the available envelope where
occur so that services cannot be placed over them.
The Outer HSP has to go on over all the services forming the limit of the envelope in Z, so
fragile services such as R/O and FSI fibres must be routed so that they are not crushed when
the HSP goes on.
Quadrant C3 with R/O fibres and FSI fibres still in the ISSS before
unfolding.
Quadrant C4 with R/O fibres and FSI fibres still in the ISSS
24/08/06
Installation of ID BARREL - 2 pics
The ID barrel being moved by crane from SR1 to the ATLAS building
The ID BARREL being installed in the ATLAS cavern - still in the
cradle.
24/08/06
The completed SCT BARREL optocable connections to the BOCs
A BOC with connections completed
TOP crate of Y22
BOTTOM crate of Y22
TOP crate of Y23
BOTTOM crate of Y23
View down through the base of the rack with heavy cables
held behind a bracket at the left side.
View of both racks completed.
See also the (Document) step-by-step guide to SCT-BOC cabling (contact Martin Postranecky)
18/08/06
Preparing the optocables for connection to the crates
Optocables laid out on the floor of USA15
The optocables are as installed in the underfloor trays with the
stripped-back lengths protected inside heavy tubes. The excess length has
to be accommodated by stowing under the floor beneath the rack.
a closer view of the cables separated into bunches
colour coded upper or lower crate (green or yellow), types Rx or Tx (red/orange or blue).
The space under the rack floor showing optocables and cooling pipes
etc. The excess heavy optocable lengths had to be stowed away under here.
Removing the protective tube. The thick layers of black tape that
join the tube to the optocable have to be cut away and the polystyrene
bung removed to enable the separation of the tube and cable with stripped-back fibre ribbons.
A view of the long workbench which supports the whole cable end.
Once the protective tube, fibre wrapping and cable stiffeners are
removed from the ribbons, they
are carefully formed into a single uniform stack (of either 8 or 6 ribbons) for sleeving. The stack
is maintained in the same order as it emerges from the heavy cable. Note that in a few cases the ORDER
of the ribbons, according to the labels on each, were not sequencial - a note was made of this.
The tool used to fit the protective Round-It braid over the fibre ribbon
stack.
The application of the Round-It braid.
Two people are needed, one to apply the braid using the tool and another to hold
the ribbon stack in place to make sure the ribbon does not twist.
The remaining lengths of stiffener are removed
once the ribbons are protected by the braid.
The connector ends are threaded through a 60mm length of coour-coded heatshrink
sleeving which is used to secure the join between the heavy cable
and Round-It braid . The sleeving is shrunk over the join .
Connector housings are fitted whilst the above is being done
The ends of the ribbons are examined for signs of damage before the
housings are fitted.
An example of a ribbon with a small 'nick' in the blue covered single fibre
on the outer edge of the ribbon.
Closing the Infineon connector housing and final
step .
Connecting the SCT BARREL optocables to the BOCs in the ROD racks
The optocable is lifted into the rack with the delicate ribbons
protected in a bag and then the heavy cable is secured at the
side of the rack whilst the ribbons are supported.
The fibre ribbon stack, enclosed in Round-It braid is threaded inside
the top plastic conduit alongside other bunches.The labelled ribbons
are matched with the destination SMC sockets on the fibre mapping chart
The dust-covers on the SMC sockets are removed and the open sockets
and optoribbon connectors cleaned before each is connected.
14/07/06
The TTC fibre ribbons were found to have different length 'Y' splits
on one sector
This meant the lengths had to be adjusted during cable laying and the new layout
was modelled using these new lengths.
A view of the new layout from above
A view from an angle - note fibre ribbons crossing HEX region
View showing modelled connectors in the housing fitted accoring to
final version of the mapping.
close-up of set closest to fibre channel in cryostat
close-up of set on the 'far' side of the HEX.
04/07/06
29/06/06
29/06/06
27/06/06
23/06/06
26/04/06
03/04/06
28/03/06
20/03/06
08/03/06
07/03/2006
06/03/2006
03/03/2006
14/02/2006
10/02/2006
08/02/2006
12/01/2006
09/01/2006
22/09/2005
15/08/2005
12/07/2005
16/06/2005
09/06/2005
09/06/2005
05/05/2005
30/03/2005
Also in SR1 - work on cabling Barrel 3 for testing!
10/03/2005
03/02/2005
09/12/2004
08/11/2004
03/11/2004
15/10/2004
28/09/2004
23/09/2004
06/09/2004
03/09/2004
05/08/2004
29/07/2004
26/07/2004
12/07/2004
25/06/2004
23/06/2004
17/05/2004
19/04/2004
26/03/2004
19/03/2004
18/03/2004
16/03/2004
04/03/2004
03/03/2004
02/03/2004
Further description of 'Camembert' for R/O fibres
28/02/2004
28/02/2004
24/02/2004
18/02/2004
13/02/2004
30/01/2004
A new layout of Barrel 3 services cage boxes has been made.
28/11/2003
27/11/2003
24/11/2003
17/11/2003
Here are the photos of the trial assembly on the model cryostat at UCL
11/11/2003
01/10/2003
Real barrel assembly!
11.09.2003
11.09.2003
01.09.2003
13.08.2003
11.08.2003
08.08.2003
15.07.2003
17.06.2003 and updates
21.05.2003
An attempt was made to solder the litz wire (20 wires) to the gold surface of a sample of the REAL foil (Kapton/Al/Ni/Au) threaded through a 6mm diameter hole, with wires fanned out on the surface.
14.04.2003
09.04.2003
13.03.2003
27.02.2003
09.02.2003
27.01.2003
24.01.2003
23.01.2003
17.01.2003
05.12.2002
04.12.2002
04.12.2002
20.11.2002
05.11.2002
11.10.2002
27.09.2002
27.09.2002
25.09.2002
30.08.2002
09.08.2002
22.07.2002
10.07.2002
27.06.2002
18.06.2002
09.06.2002
07.06.2002
07.06.2001
10.05.2002
10.05.2002
04.04.2002
26.03.2002
14.03.2002
15.02.2002
14.02.2002
07.01.2002
22.10.2001
The provisional measured lengths of readout fibres will be confirmed when the exact specification of the fibre splice protector is finalised and all lengths of fibres and low mass tapes tested using a full set of model harneses on the model barrels.
See also lists of numbers of fibres and their positions
on the 4 barrels and ducts to PPB1 (go to item below dated 07.06.2001 - still valid at 22.10.2001)
04.10.2001
A model 1/4 section of the Thermal Enclosure Bulkhead (designed and made at RAL) has been installed. This is in 2 sections, allowing low mass tapes, optical fibre ducts, caplillary tubes and DCS to pass through; specially made grommets used with sealant will provide a gas-tight seal. The cooling exhaust pipe bulkhead is seated in the thermal enclosure (one per quadrant).
The layout of low mass tapes at PPB1 has been revised.
A complete set of cooling output pipes has been installed.
New photos of the LMT PPB1 cover and the fibre housing
The full length of the single cover is shown with the fibre
housing at the far end. NOTE that the housing is shown on the outside of the
set of 3 adjacent LMT stacks - but the housing is really located on a U bracket
over the MIDDLE stack of LMTs in the row of 3.
In each quadrant the housing is located either side of the HEX, fitted over the
2nd stack of LMTs out from the HEX.
Here is a close-up of the Housing showing how it is fitted
on the 'U' bracket over the tapes and their cover. (but see NOTE above)
A photo-guide to the SCT barrel INSTALLATION order for the LMTs
The new installation order for the tapes was tried out on the model -
The .ppt file can be found on
http://www.hep.ucl.ac.uk/~tjf/instorder0706.ppt
Please note that a slightly modified version can be found in Richard Apsimons transparencies
shown at the ID week on 6th July 2006.
ROUND-IT braid with ID=5mm is used to protect bare ribbon fibre in the ROD
racks
The ROUND-IT braid is fitted over the bunches of 6 or 8 fibres
Medium diameter Heatshrink is used over the interface between the heavy fibre cable and the
bare fibre ribbon covered by the braid. The Heatshrink is not shrunk yet.
Here is a photo with the Heatshrink shrunk. This prevents the
braid from slipping.
The thin covers fitted over the PPB1 LMT boards are designed to
protect and enclose the delicate crimped cables and connectors from tools etc
but are NOT designed to withstand being trodden on!
Here is a photo of the fitted cover showing the limit of the
envelope in R which is defined by the height of the 'rear' section of the cover.
Connecting SCT Opto-cables to BOCs in USA15 at CERN
First - some pics of the crates and racks installed in USA15 at CERN:
TTC crate in TTC rack
Front of our 2 racks with RODs (ReadOut Drivers) and TIMs (TTC Interface Modules)
BOCs (Back Of Crate)
Rear of rack with BOC and conduits for optofibre ribbons
Rear of rack with both optofibre conduits
LMT PPB1 connectors are identified for easy connection
view of the two PPB1 boards with FCI (blue)cable connectors
on one side of the board only marked with a white
'blob' and JST (white)cable connectors going to the same side of the board marked the a black 'blob'.
This should eliminate confusion
between the two FCI or two JST connectors on cables of the same length connecting to the same board.
close-up view of the connectors
Note how the FCI (blue) connector on
the right has become partially disconnected!
This is a very common occurrence.....the 'Oxford clip'
must be fitted immediately after each set of 2 FCI and 2 JST connectors are connected, before going on to the
next set.
Barrel fibre ribbons have to follow a difficult route around the cryostat flange after
making a 90 degree twist after the route described below (see 28/03/06).
The modelled route around the cryostat flange
Another view showing the change in direction
A view showing how each bunch is separated out all the way along to avoid tangles
and crossings-over in the cryostat bore.
Note that the bunches are arranged here in the 'final' configuration for routing
to the PPB1 connector housings. (Yellow labels=12way, Blue lables=6way ribbon).
Barrel fibre routing OUTSIDE the cryostat .
A trial was made routing fibre ribbons from
the end of the heavy cable, through a restricted space envelope. The ribbons each have very large labels
on sliding sleeves which will not fit into the available space on any other part of the route. A layout
had to be devised that could accommodate all the ribbons together with the labels.
Here is an annotated photo of the layout of ribbons and labels. Note that when
doing this in situ a small flat sheet of metal or similar has to be used to form the 'base' onto
which the ribbons are laid and held in position using Kapton, until the correct channel is available -
which will protect the ribbons. If this is not laid out in a compact way there is a very real risk of
damage to the fragile fibre ribbons in this region which is completely exposed.
Barrel fibre routing at PPB1 with real housings and defined fibre lengths
General view of the two housings and fibres indicating different lengths
Upside down trial with fibres and housings on a board. The stiffness of the
fibre ribbons, their short length and their grouping into sets held with Kapton sleeves meant that
gravity was not a serious problem here - but more extensive trials will be made when PPB1 lids are available
and housings can be fitted to the model cryostat
A view of the fibre routing in the flat plane, showing that fibres do not exceed
the envelope in R. A special cover will be fitted after final connections are made.
An angled view showing fibre ribbons and ducting.The ribbons stay within the
low-profile area of the PPB1 LMT cover.
Final trials of EC PPF1 fibre routing and trial with ribbon length differences
Fibre routing and connections within the definitive ENVELOPE
View from the front showing ribbons routed down the side of PPF1
Unobscured view of fibre ribbons/connections in PPF1, assuming all ribbons
approx. the same length to connectors
A trial with fibre ribbons of different lengths, showing that the
excess length can be accommodated by allowing the ribbons to bend within the definitive envelope.
The excess length could be up to 20mm.
Preparation for SCT/TRT Barrel installation in the ATLAS cavern - LMT installation
A photo of the Low Mass Tapes modelled on the 360 degree end-barrel
model showing a sample octant with harness pair names eg B6P11, B4P8.
The same photo showing relative amounts of 'twist' in the LMTs between
the TE bulkhead region and the 90 deg bend onto the cryostat. The 'twist' occurs to a greater
degree on the tapes at outer edegs of the octant.
A view directly onto the tapes showing how they line up in the PPB1 cryostat
region.
Further trials of EC PPF1 fibre routing
An annotated photo of the fibre routing through PPF1
Another view of fibre routing and connections in the PPF1 region
An annotated photo of the fibre ribbon bunches showing lengths at PPF1
SCT EC work : A prototype PPF1 was used to do feasibility studies of fibre routing
The path of the EC ribbon fibres is a narrow one either side of the LMT PPB1and
a view from the other side where
the ribbons have make a 'floating connection' without any PPB1 type housing to support them.
The preliminary routing of the fibres including connections is made here within
the outline of the structure of the PPF1 and a view from the other side.
Here are the fibre ribbons showing the ideal arrangement of connectors IF fibre sets are of
equal length.
The PPB1 region inside the cryostat was examined in order to find out where the optofibre housing
fitted and a template with a full-scale drawing of the housings was offered up to check the fit.
Here is an annotated photo of the region
A template is held up to show where the optofibre housings will be located -
the SCT barrel and TRT barrel will fill the space on the right, so the housings are very close to the
TRT/SCT barrel end where all the services are routed out to the cryostat.
A close-up view of the template . The housing is not the final design.
Another view from approx Z=0 looking outwards to the cryostat flange Note that TRT services
will be routed along inside the enclosed channels.
Accommodating different lengths of optofibre ribbon in the area between ducting and PPB1 housings
Sets of fibre ribbons are enclosed in heavy cable from the racks to outside the cryostat. The
numbers of ribbons in each cable (which enclose either TTC or Data fibres, not both) mean some sets of TTC
ribbons from the same cable have to connect up to both housings. As the ribbons from one heavy cable are
all the same length and pre-fitted with MT connectors, some excess length has to be accommodated outside
the ducting.
Here is one option for taking up the extra ribbon length for the TTC set from one heavy
cable, where some ribbons have to connect in the connector housing on the nearside of the ducting and others have
to cross the HEX and connect in the housing on the far side.
Here is a better solution. By allowing ribbons to slide relative to one another and make
different size loops (respecting the minimum bend radius), the differece in length can be taken up in the space outside
the duct. A cover will be fitted over the fibres to protect them after connecting up.
Optofibre routing from 'outside' with model PPB1 housing - preliminary (with new 'Y' TTC lengths)
The previous exercise (see entry for 09/01/2006) was repeated in more detail. The main difference was that the LENGTHS
of the semi-rigid TTC 'Y' split ribbons was discovered to be 160mms when the real cables were unpacked at CERN.
The shorter length means that there is fortunately now enough space for the ribbons to bend around 90 degrees
to reach the fibre ducting inside the 'low profile' (in R) part of the Low Mass Tape PPB1 board covers
An annotated photo of the layout Note that there are 11 TTC 12 way ribbons which split
into half, the separated 6-way halves protected by a semi rigid covers terminating in MT connectors. These correspond
to the 22 Data 12 way ribbons each terminating in MT connectors
Each connector housing takes 44 connectors - 22 TTC from 11 ribbons and 22 Data from 22 ribbons from the outside.
NB: On the barrel side there are 22 TTC ribbons and 22 Data ribbons each terminating in MT connectors ie no split
ribbons.
There are two housings per QUADRANT, one each side of the HEAT EXCHANGER cooling pipe run, but only one fibre duct
serving both, so fibres to the 'far' housing have to route OVER the HEX run.
Note that the exact dimensions of
the HEX are not yet known.
Fibre ribbons emerging from the ducting are held in stacks using Kapton made into a 'sliding
sleeve' which keeps ribbons togther but allows ribbons to slide relative to one another for length adjustment.
TTC and Data ribbon stacks
The specific region where the R/O optofibres bend from the vertical to the cryostat was modelled.
The R/O fibres are in single fibre form, each protected by furcation tube sleeving when on the barrel and
outside it. The fibres are ribbonised some way between the TE bulkhead and the PPB1 and this transition
is protected by a rigid heatshrink section which covers the splice region as well. Beyond this, the
fibres are in 12-way ribbons with avarage length of 240mms, the ends terminating with MT 12 connectors.
The fibre ribbons have to make a 90 degree bend just before reaching the connector housings on the cryostat wall.
Here are some pictures of this region:
the ribbons making the bend to the model housing
another view showing the position of the fibre housing in Z and the space available
for the fibres to bend. The 'wall' on the left represents the available envelope for the fibres which have to
route over the already installed Low Mass Tapes.
Line-up of heat-shrink sections so that RIBBON LENGTH is just right for connecting the
MT connectors and ribbons do not have to cross over each other to reach the allocated connector slot inthe housing.
Another view
Note that:
The lengths of the ribbons from the heatshrink section to the MT connector may vary due to splice
repairs etc
The available envelope in Z in the vertical plane allows for only a single row of heatshrink sections so it is necessary
to lay these out flat, side-by-side, in the correct order for connection, shifted forward or back in the set depending on
the ribbon length to reach the connector housing.
All slack is taken up and necessary crossings-over are made in the region where fibres are sleeved in furcation tubing.
Here is a view of the ribbons where they bend.
A close-up view of the 'twist' on the bend.
They have to twist 90 degrees from 'flat'
in order for the MT connectors to fit in the housing slots. Experts agreed this was acceptable.
A first attempt at R/O fibre routing at PPB1 after it was known that the TTC fibre ribbons would
be split and protected by semi-rigid covering originally 195mm long.
See later changes to this!
The preliminary layout with connector housings at the top, left and right of the
space which will be occupied by the Heat Exchanger pipes.The fibre ducting on the cryostat is at the bottom
right. This layout shows that there is insufficient space to allow the TTC fibre ribbons to bend from
connector housing to duct as the fibres cannot route over the raised section of the PPB1 covers.
A close-up of the TTC 'Y' section at the housing. The connectors are in a double formation.
The prototype TE bulkhead section was used for services sealing trials.
An octant (with harnesses of real Low Mass Tapes, real furcation tubing (optofibre covering), TPs resembling
DCS wires, fibre ribbons for FSI and wire of the same diameter as capillary tubing) was sealed using all 3 sections of the bulkhead (rear, middle and front).
The bulkhead rear section was bolted onto a plywood surface representing the 4 barrel flanges, with services
emerging as they would off the barrel ends. The trial was purposely done upside down as this
was going to be the most difficult to access and seal on the real bulkhead.
The sealant used was a mixture of Techsil and talc as thickener. See Document by Brian Anderson.
DCS and LMT harness sealed in bulkhead, fibres in grommet unsealed.
All LMTs sealed, fibres being sealed with unfilled sealant.
Fibres with pressure plates visible, held with temporary screws - middle bulkhead
Services sealed in bulkhead - general view - awaiting front section of bulkhead.
Services sealed - close-up view - temporary screws will be removed then the front
outer section of bulkhead fitted.
All services and all 3 setcions of the bulkhead sealed
All three sections one octant of the prototype Thermal Enclosure Bulkhead were made at UCL and tried out
with grommets also made at UCL
view showing FRONT part of bulkhead in the foreground - this is the final part
to go on. The middle part houses shaped fibre grommets, capillary and DCS grommets.
another view as above.
full view looking down on the outside radius.
full view looking down on the inside radius.
another view.
another view.
The first part of the section of TE bulkhead made at UCL was checked
The 'rear' part of the bulkhead is the first part to be installed, and houses the Low Mass Tapes
in their grommets. An anomaly was discovered in this new -final version of the design when compared
to the earlier basic design.
Here is a photo illustrating the shift in the position of the grommet
and another showing the full section.
Here is a photo showing the problem area with Low Mass Tapes.
The problem was resolved by making a small modification to the production bulkhead before fitting.
Trials routing a full set of capillaries on the barrel end
The capillaries (all the same length) were routed from the barrel input end out through the
Thermal Enclosure Bulkhead (where there will be a grommet to hold each tube in place) - to the
approximate position of the Heat Exchanger.
a view of the capillaries bent to avoid too many overlapping.
another view showing full length out to HEX area.
a simple way to label fibre ribbons
the non-sticky label is secured on the ribbon using sticky Kapton tape If the label
needs to slide then the Kapton tape and label must be the same width.
trials with capillary bending on the barrel end
photo shows how a single capillary can be bent with 35mm bend radius to lie
as flat as possible on the LMTs on the barrel end, taking up excess length. The bending was done using
a 35mm former in-situ on the barrel end.
photo of the 4 barrel and cryostat model showing position of Barrel Heat Exchanger
Approximate position of Heat Exchanger marked on a photo in space between TE bulkhead and cryostat.
Installation of TRT Inner Bore Heaters in SR1 at CERN
Preparation of heaters in the Lab - soldering cables to the connecting tabs.
Thermal imaging tests on the heaters in the Lab. at CERN.
Glue layer application to TRT inner bore - for first two heaters
Glue application and glue layer spreading.
Application of the first heater to the carbon fibre surface.
Smoothing down first heater.
First two heaters glued in place.
Thermal imaging of one heater glued in place on TRT inner bore.
Spreading glue for next heater.
Four heaters glued in place.
Taping the guide for the next heater.
Six heaters glued in place- the remaining (upper) two will be installed later after repairs to thermistors.
Barrel 3 in cable nest
R/O fibre and TPP storage in the ISSS
R/O fibre storage showing length of rigid section
TPP box with PPB1
TPP box with PPB1 another view
A trial set of FSI fibres was made at OXFORD and tried out at UCL
Note there is here a 130mm rigid section where the splice from single fibres to fibre ribbon is made. the red ends are pre-attached to the T POST.
Tpost area B3 and B4 flanges
Tpost area B4 and B5 flanges
Tpost area B5/6 flanges
close-up of lightpath tube and insert obstacle on flange
A prototype fibre grommet was made in the UCL workshop
Here is a photo of the grommet made of PEEK
It will be possible for UCL to make the production set of these.
Trial glueing heater to inner bore of TRT
Here are photos of the various stages of the trial glueing of one of the uppermost heaters at UCL using a full-scale model of the inner bore of the TRT:
stage 1
stage 2
stage 3
stage 4
stage 5
stage 6
stage 7
Final Barrel 6 services Cage Layout
Here is the B6 services cage layout after final changes to the DCS localtions
Note that lists of the services cage fibre boxes exist as .xls files as well.
A prototype Input Capillary Grommet made of PEEK designed by RAL was made
Here is a photo of the 7-way tube grommet.
A metal sleeve is fitted in one location. Each capillary tube would be fitted in advance with such a sleeve which would be glued in position when the grommet is fitted. The sleeve acts as strain relief to prevent the capillary from slipping through the grommet after sealing.
Here is another view of the grommet
Here is a photo of the grommet in two sections
Here is another view of the sections
Note that the sleeve must be fitted on a straight section of capillary tube where there is adequate length to allow
for adjustment to bring it into the correct position as the sleeve is designed to be a close fit and cannot be moved overbends in the pipe.
A document was made describing the history of the TE bulkhead modelling work at UCL
Here is the document with pictures of the bulkhead modelling to date .ppt file)
A template is needed for the Temporary Services Support - here is the design
In order to maintain the services (LMTs, fibres and DCS wires) in the correct configuration for
(a) sealing in the Thermal Enclosure Bulkhead and (b) connecting up to PPB1, a Template as part of the Temporary Services Support is required. This Template will be used to guide and fix the services
to the support as close as possible to the TE bulkhead during and after sealing.
Here is a photo of an octant section of the Template Note that each position of each services bunch will be labelled with the names of the harnesses.
Here is a .ppt file with drawing and photos of the prototype Template
Here is a sketch of the Template with suggested dimensions (.ppt file) - these can change to fit the design (Geneva) of the main Temporary Services Support.
Barrel 4 Services Cage Layout
Here is the layout - .ppt file
A first-prototype PEEK grommet was made and R/O fibre furcation tubes and FSI (alignment) fibre ribbons were sealed in it
Here is a photo Note that the model section of bulkhead into which the grommet is fitted is not realistic - an angled tape grommet can be seen in the same section but in reality fibre grommets and tape grommets occupy different bulkhead sections ( fibres - middle section, tapes - rear section).
Here is a photo showing the X-section through the bulkhead A bunch of fibres are shown coming out from the surface of barrel 6. The rest of the fibres are shown as coming from other barrels
Here is another view of the grommet and fibres Note that the barrel 6 fibres are sealed in first followed by those from barrel 5, barrel 4 and barrel 3. Fibre grommets will take varying combinations of barrels depending on the grommet size and its location on the bulkhead.
The Readout Fibres from the barrels are connected at PPB1 (.xls file)
Here are layouts showing how the fibres are connected at PPB1
Each harness has two bunches of fibres each terminating in MT12 connectors. The connector sets are arranged in blocks per octant so there is one sheet for each octant showing how they are connected.
The cooling input capillaries from the barrels are routed through the TE bulkhead and out to the cooling area
Here is a .ppt file showing a diagram of the layout and routing of the input cooling capillary pipes sets passing through the Thermal Enclosure bulkhead and out to the cooling system areas.
Barrel 5 Services Cage Layout Update
Here is the updated layout - .ppt file
Trials have been done soldering TRT heater wires to heater pads
The gap between the TRT barrel envelope and the SCT barrel envelope is very restricted so some trials have been done in order to measure how much space a soldered wire would occupy.
Here is a close-up photo of a Kapton covered wire soldered to a 'MINKO'heater pad
and another pic.Note that the Kapton covered wire is the one with the RED sleeve.
Barrel 5 Services Cage Layout +Z and -Z end - provisional
Here is a .ppt file for the layouts
Note that for barrel 5 as for barrel 6 the layout here is dictated by the layout of the Thermal Enclosure Bulkhead feethroughs as the TE bulkhead is relatively close to the barrel 5 flange so there is very little possibility of routing fibres at an angle across from exit points on the barrel to feedthroughs at a different phi position.
Final trials of the B6, B5 and B4 cooling exhaust manifolds were carried out on the model
The barrel cooling exhaust connectors have to penetrate through gaps in the low mass tapes when the services are fitted in their feedthroughs in the TE bulkhead.
In order to make sure that the connectors from
each manifold are in the correct position, trials had to be done with the LMTs fitted in feedthroughs
on a prototype 360 LMT section ring specially made for the purpose according to the latest RAL TE bulkhead design.
Here the 360 model is shown stripped back to enble the LMT bulkhead ring to be fitted
Here is the complete 360 ring with LMTs in position
A Barrel 6 model manifold in position
Plan model of a Barrel 5 manifold and a plan model of a Barrel 4 manifold. (Circle indicates position of connector stalk).
PPB1 Type II cable connector clip (Oxford design developed from UCL prototype) trial
Here is a photo of the clip
Here is another photo showing the cable side of the clip
The clip is pushed over the two JST connectors and the two FCI connectors and a
small rod threaded through holes in the clip, under the cables from one end.
Here is another photo showing the rod (same length as the clip)
NB Care must be taken to fit the clips in the right order so that there is always space for rod insertion from one end of the clip.
Barrel 6 Services Cage Layout +Z and -Z end - provisional
Here is the .ppt file for the layouts
Note that for barrel 6 the layout is dictated by the layout of the Thermal Enclosure Bulkhead feedthroughs as the TE bulkhead is only a few mms away from the barrel end. Fibres bunches leaving the barrel 6 end must pass directly through the bulkhead grommets.
The CERN ROD rack mock-up now has the full number of thick cables and cooling pipes
Here is a pic of the 36 thick cables distributed equally between top and bottom crates
The blue cables are modelled and there is a single 'real' black cable which is heavy and stiff but can be bent over at the top. Cable ties are used here as strain relief and there is adequate space for two layers of cables to one side of the rack.
Cooling pipes are on the other side of the rack. Some racks may have the reverse configuration ie
cooling pipes on the left and cables on the right.
Some photos taken of the real CERN racks at CERN can be found here
Thermal Enclosure Bulkhead region - fibre bunching and LMT stacking options
It is necessary to avoid having large stacks of LMTs and large bunches of R/O fibres passing through the TE bulkhead as both fibres and tapes have to make a 90 deg. bend outside the bulkhead when stored for insertion of the barrel SCT into the TRT. There is very little clearance in this region.
Various options were tried out on one octant of the barrel end model and the best selected as the baseline for use in the design of the TE bulkhead:
(a) Fibre bunches of 54, 126, 126, 90 fibres; one stack of 24 LMTs adjacent to cooling exhaust bulkheadrejected as 24 LMT stack too large
(b) Fibre bunches of 54, 126, 126, 90 fibres; two stacks of 12 LMTs each This was chosen as the BASELINE but in octants where there is a cooling exhaust manifold close to a bunch of 126 fibres (eg manifold G) this bunch of 126 fibres will be split into two bunches, one of 54 fibres and one of 72 fibres, routed one each side of the manifold which will remain in the centre of the gap between tapes rather than being moved to one side.
(d) Fibre bunches of 54, 126, 216, leaving big gap for exhaust manifold in position marked by red dot; 24 LMTs rejected as 216 fibre bunch and 24 LMT stack too large
(e) Fibre bunches of 54, 126, 216; 12 and 12 LMT stacks again leaving big gap for exhaust manifold indicated by red dot - rejected as fibre bunch of 216 too large
View from the side showing how fibre bunch of 216 is too high (envelope in Z)
LMT Grommet trials on short TE bulkhead block section
24 LMT stack. A layer of Sealant is spread between each LMT. The grommet here is oversize.
12 LMT stack
6 LMT stack
Barrel 6 manifold - new design space trials.
Model exhaust manifolds were specially made so that the outer dimensions were accurate. For the +Z end, types B, C, D, G, H and K were made (each is different). Trials were done on the end barrel model to see if this new design would be feasible
Manifold G in central position between LMTs- old Alu connector is visible
Manifold G from above
Manifold B in a similar position Note old Alu connector position<
Manifold B again - note proximity of INPUT connector under LMTs (indicated using a spacer)
Another manifold - fits in gap between LMTs.
Services Assembly for insertion of the 4 barrel SCT into the TRT
view of one octant of TPPs stacked, with a block of 12 model FSI spools. The number of FSI spools is now 17 but there should still be space for these.
PPB1 (LMT-Type II cable support board trials in situ
Full view of the LMT PPB1 boards assembled on the support board Note that this is a single width support board but in the final design the support board takes three PPB1 pairs ie is over three times the width. The length at the 'front' end is also shorter than shown here.
Fastening the drain wires in the T clamp - in situ Good visibility is needed ie good lighting and a second person to check from a different angle.
Fastening T clamp screws - a special tool will be needed for these operations to minimise the riskof dropping screws and/or damaging thin cables.
Checking the 'FRONT' clamp before 'REAR' LMTs are positioned over the 'FRONT' PPB1 connectors. A special bar between posts will be needed in order to keep the thin cables neatly bunched (to be designed).
'REAR' LMTs shown passing over the 'FRONT' PPB1. Note need for restraining bar across bunch of thin type II cables.
View of whole board showing need for a clip to hold LMTs in position at 'FRONT' end of board
Services during Assembly - TPPs stacked for first assembly stage where NO access to connectors is required for testing.
LMT Temporary Patch Panel boxes stacked with NO access to connectors Note that the boxes are stacked so that the connector surfaces face each other in the two layers.
Another view of the TPPs stacked. The LMTs must maintain the correct angle exiting the TE bulkhead ie TPP boxes must be arranged so that the LMTs are not forced to assume a different angle that could place stress on the sealant/grommet in the TE bulkhead.
Open top 'Camembert' Note foam disc maintains bend radius and protects fragile area where fibres enter sleeving (breakages have occurred here during harness handling).
The dimensions of this 'Camembert' are 140mm outside diameter and 27mm outer thickness - but these can be varied according tothe number and length of fibre harnesses housed. Slots for the fibres are angled so that there are no sharp edges
A thin layer of foam goes over fibres before lid is fastened
Closed Camembert.
Note that the fibre splice region is outside. This length is protected by rigid heatshrink sleeving and a rigid splice protector inside the heatshrink.The (approx.) 240mm length ribbonised fibre terminates in the MT12 connector, accessible for testing purposes.
Camembert with bunch of fibres passing through TE bulkhead. Note that these housings are very lightweight (plastic) so that no stress is placed on the fibres or LMTs amongst which they are stacked during assembly/testing
The fibres will pass through the TE bulkhead at four or five separate feedthroughs per OCTANT.
The different bunch sizes are 126 fibres, 90 fibres, 72 fibres and 54 fibres.
Each separate bunch will be housed in a separate Camembert of appropriate dimensions.
Cooling input transition pipe - non-demountable on Barrel 3 - in-situ bending sequence
Note that the transition pipe would be permanently soldered to the barrel end connection. The pipe would be pre-bent close to the solder joint and be OVERLENGTH, with a temporary connector fitted to the end BEFORE the cooling unit is mounted on the barrel.
The temporary connector on the overlength pipe will be used during tests until the 4 barrels are assembled together when input transition pipes from all barrels are connected to capillary pipes at the TE bulkhead.
Stage 1 - pipe in the position it would be on the services cage - overlength with temporary connector.
Stage 2 - pipe cut to final length and permanent connector fitted
Stage 3 - the large radius 'services cage bend' is straightened out and a new smaller radius bend made to bring the pipe out in R, so that it lies between the LMTs of other barrels
Stage 3 also - Same view - note Barrel 3 LMTs in this octant are still folded back - ie not yet moved to their final position
Stage 4 - the LMT harness underneath the cooling input transition pipe soldered joint is now folded over the pipe in it's final position
Stage 5 - remaining Barrel 3 LMTs are folded over. Bends are made in the pipe to take up excess length and the connection is made at the Thermal Enclosure Bulkhead.
Final PPB1 support boards trials done on the bench (see report 03/03/2004 for in-situ trials)
Clamp block spacers reduced in width - now 6 cables Type II cables fit perfectly
Intermediate clamp removed - no braid - now clamp is a good fit
View of T clamp holding cables securely, with drain wires and grub screws visible
PPB1 support board assembly trials on the bench - results of the near-final design trials:
'Grounding of the shielding braid' layer proved to be too thick
this prevents the profiled clamp from fitting over the cables
the top section sits too high for fixing in place with screws
'Grounding of the shielding braid' removed - clamp profile would be a better fit if spacers did not pull cables at each end inwards
without the braid, the top section (T section) can be screwed on
fastening grub screws on the 'T section'
Drain wires threaded through 'T section' successfully
View showing length of support board - 6 cables won't fit between spacers - but easy to fix by removing some material from the inside of each.
Services during Assembly - a possible arrangement of LMT Temporary Patch Panel boxes at the testing stage when access to the TPPs is required for connecting up test cables.
Arrangement of LMT TPP boxes on 'temporary patch panel structure'.
Another view of the same layout.
The LMT TPP boxes would be held in place in slotted frames - still to be designed.
Note that all model boxes are labelled only on the surface where the connectors would be accessible - so here all connectrors are accessible.
All the TPP boxes are aligned so that the LMTs are as close as possible to their final angular position when connected to the final support boards on the cryostat. (One of these boards can be seen at the top right on the model cryostat in the first pic.).
'Camembert' - possible R/O fibre housing arrangement for all stages of assembly after removal of the services cages.
There will be four 'Camembert' and one DCS wire spool per OCTANT per barrel end. At present it is expected that there will only be one DCS wire spool per quadrant ie some octants will have no DCS spools
A full-scale model of the CERN ROD rack has been made at UCL in order to try out possible layouts of R/O fibre cables and bare ribbons connecting to the two crates.
General view of whole CERN rack model
Close-up of possible fibre routing from cables to top crate. Note that the thick cables contain either 6 or 8 ribbons and all the ribbons are 12-way.
Side view of fibre ribbons.
Close-up of fibre ribbons routed through a slotted duct keeping ribbons well outside the minimum bend radius
Note that each thick cable has been stripped back so that the bare ribbons are approx. one meter in length. These are routed to the top crate via the slotted duct allowing take-up of excess lengths. Loops of hanging ribbons are thus kept to a minimum. Some protection of the ribbons may still be necessary if there is a possibility that access to rack cooling etc will be needed.
A new layout of fibre routing and DCS routing to services cages has been made for Barrel 3
Here is a .ppt file containing 4 diagrams of fibre and DCS routing
1) +Z end of B3 showing the new readout fibre routing with fibre bunches exiting at two positions in phi per octant.
2) -Z end of B3 as above. Note this new layout allows 2 fibre harnesses to go into one box (see below for diagram showing layout of boxes).
3) +Z end of B3 with new DCS cooling temperature sensor wiring route. The association of cooling pipes with harness (module) rows is shown with positions in phi
4) -Z end of B3 with new DCS cooling temperature sensor wiring route - NOTE - on this end there are
two extra temperature sensors attached to the cooling pipes where the cooling pipe and associated module row fall on either side of the vertical separation routing to USA15 and US15.
Here is the .ppt file with 2 diagrams showing the layout on
1) +Z end. The positions of cooling input pipes and barrel interlinks are also shown.
2) -Z end as above
A trial using the new design of cover for the LMT PPB1 boards on the cryostat has been made.
Here is a photo with the cover screwed onto the the pair of LMT PPB1 trial assembly boards shown in the entry dated 17/11/03.
Clearance and LMT installation trials have been made using a model of the new design of Cooling Exhaust Connector.
Here are photos of the model exhaust connector on the barrel end model
Note that this exhaust connector is integral with the on-barrel cooling unit therefore the precise angle of the connector has to be set during cooling loop manufacture. This angle will be different for each cooling loop.
Assembly sequence - when the 4 barrels are assembled together the LMTs and their connector boxes have to be stacked within the clearance envelope during installation - a trial has been done.
Here is a photo of the 4 barrel model with ONE OCTANT of LMTs and model connector boxes stacked . This works as there is sufficient space but has to be done with boxes and tapes in the right order. The method of supporting the structure has not yet been decided.
A trial assembly of the LMT PPB1 boards on the support board in situ on the cryostat has been made.
The assembly on the cryostat is of course more difficult than on the bench. It took one and a quarter hours to fit the FRONT and REAR PPB1 boards, fit clamps and connect up the Type II cables. In the real situation the temporary PPB1 boards would have to be removed first. We have found that connecting and disonnecting the Type II cables takes up the most time and has to be done very carefully. Here are some comments as a result of the trial:
1) Marking the connectors so that it is obvious how they should be connected and where will be essential.
2) The various loose components and tools will need to be kept in an 'unspillable' container
3) Very good directional lighting will be needed.
4) The square spacers can probably be modified to allow 6 cables to fit between them and the corners should berounded
5) There should be an additional clamp for the REAR LMTs at the FRONT end of the support board
6) The multiple cables should be bunched so that there are no stray cables to get in the way when the adjacent boards are fitted and the final cover fitted.
A new trial assembly of the LMT PPB1 boards on the support board has been made.
The assembly design and drawing has been made at RAL (drawing number TB-0049-401 Brian J. Smith).
Here are some photos of details of the PPB1 LMT assembly using production low mass tapes and real type II cables.
Here are some pics taken on 01.10.2003 at RAL of services cage trials using a TEST harness
on the REAL barrel 3
New length cooling transition pipes have been modelled
If cooling transition pipes on one barrel all have to have the same length (different for each barrel) then all but one of the pipes will have to be overlength. This presents problems as the pipes will have to make more bends in order to take up the surplus length in an already congested area.
Here is a new set of photos illustrating the layout
NOTE: It will have to be proved that the pipe can be bent to the 'pattern' in situ -
we will use real transition pipe and special bending formers for the next attempt.
Barrel 6 has been left as is (see record for 11th August) as there is a special problem here during the 4 barrel installation sequence. It will not be possible to have all the pipes the same length.
latest update of the TE bulkhead services leaktest/materials trials document
Here is the document revision number 1.5
The proposed Barrel Numbering Scheme is illustrated
Here is a sketch with examples
The order of harness installation is illustrated, together with fibre
routing and DCS wire routing orders
Here is a .ppt file with sketches showing harness installation orders, harness types and locations and DCS and fibre routing
on the barrel.
Cooling transition pipes and capillaries have been modelled.
This is however only a preliminary layout as capillary lengths may be standardised. A new set of transition pipes with realistic lengths will then replace
existing pipes on the model.
This replaces the old layout where capillaries extended from the barrel end through the Thermal Enclosure Bulkhead to terminate in connectors at PPB1.
Here is a series of photos of the model including all four quadrants in detail. Note that capillaries (see first quadrant) must join the exhaust pipes as soon as possible and that each pair of capillaries, although they may be from connectors in set one or set two must cross over the exhaust pipes to join alongside the exhaust pipe of the SAME cooling unit.
The full set will be included on the model later.
A trial layout of fibres (furcation tubing) has been made to represent fibres from ONE OCTANT ie 396 fibres in total. The new layout is necessary due
to the difficulty of fitting all 396 fibres in furcation tubing through a single 'slot' in the Thermal Enclosure Bulkhead. The fibres will most likely have to pass through the TE in four separate bunches at different places in phi - see sketch of 15.07.03.
Note that the ends of bunches have been left without splices or ribbon. These will be added later, as the real harnesses arrive at RAL - as the positions of the splices can vary. We can then build a representative picture of how the real layout will look.
Here is a photo of the fibres coming through gaps in the LMTs, crossing the TE bulkhead area (there is no TE bulkhead on the model at present) and terminating at the stage where there would be a splice.
The end of each set of 6way and 12way fibres is colour coded:
B3 = red, B4 = green, B5 = blue, B6 = yellow.
The ends are in the position where they would be in order for the ribbon fibre
ends (not attached) to connect to the fibre PPB1.
Here is another view of the same octant
and here is a close view of fibre bunches
The fibres in bunches now have to cross LMTs to reach the PPB1 position in phi. Here is a view showing how they keep well within the allowed envelope in Z, going over LMTs and under cooling pipes.
There are 396 single fibres in furcation tubes per octant which have to pass through the Thermal Enclosure Bulkhead. The limited space available means that the fibres cannot now pass through in one single bunch but have to be split up into several bunches. This means that fibres have to deviate from the original route on the barrel ends.
A possible solution is to have four separate bunches spread out around the octant. (3 bunches of 90 fibres and 1 bunch of 126 fibres in the original 'duct'
where there is more space.)
Here is a .ppt file sketch of the possible layout
showing how bunches from various barrels are combined into 4 bunches going through the TE bulkhead.
This is a replacement of Version2 dated 15.07.03
Now also see photos - AUGUST 8th 2003 above where the bunching is as on the sketch.
Leak testing and materials testing of sealant and grommets which allow services to pass through the thermal enclosure bulkhead are being done at UCL. A Report on the testing of the SCT barrel thermal enclosure bulkhead services sealant (B. E. Andersen, D. Attree, J. Fraser) can the found here. Tests are ongoing - updates of this .doc file will replace old versions.
The barrel end cooling tube referencing disk as described in Ned Spencer's document of 21.04.2003 is being modelled at UCL. A first attempt is described here.
The document describes a foil annulus placed over the power tapes and cooling tubes. We think a foil in one piece may be unmanageable, so trials are done with the foil in sections for each quadrant, however it is possible that two halves would work. The foil used for modelling was 100um kapton/Al but the real foil to be used is 100um total with Kapton/Al/Ni/Au layers (exact composition not known).
This shows 3 tab links (RH section of quadrant)from the foil to the heat-spreader plate tabs which would be pre-soldered to the heat-spreader plate under the tapes. The tabs are placed as evenly as possible but have to avoid high tape stacks either side of the cooling pipes and wide fibre ducts (not shown) next to the tape stacks.
This is just a possible solution.
Both the heat-spreader plate tabs and the foil tabs are 70mm long x 10mm wide. The tabs from the foil have to pass through the thermal enclosure bulkhead with the power tapes so have to be pre-aligned at the same angle as the tapes.
Their positions on the outer edge of the foil (B6 outer diameter) are symmetric about the cooling pipes and are the same for each quadrant.
The heat-spreader plate tabs are positioned so that they emerge through the tapes at right-angles to the foil tabs. This facilitates the join between the two tabs and allows for some play in the looped tabs. A solder joint here would be difficult - it may be best to use special clips 10 x 10mm surface area.
The 3 tab links on the LH section of the quadrant are shown here
Here is a view of the whole quadrant showing all 6 tab links
Another view showing the foil section for one quadrant marked with colours where litz wire connections would have to be made between cooling pipes under the foil and the outer surface of the foil.through pinched holes
The RED dots represent litz wire connections from the cooling exhaust manifolds
The BLUE dots represent litz wire connections from the input capillaries
The GREEN dots represent litz wire connections from the cooling return loops
NOTE that the positions of these are different for each Quadrant. Here there are 6 exhaust, 12 capillary and 10 return loop connections.
Here is a view of the underside of the solder join showing the Kapton surface of the foil. The other end of the litz wire is soldered to an eyelet tab. The tab with wire would be pre-attached to the cooling pipe.
Here is a view of the solder area from the gold side of the foil. Note that there is solder/flux splatter. More trials need to be done using wire with chemically cleaned ends and some means of flattening the splayed wire ends. It may help to pre-tin the surfaces.
here is a view of all four quadrants. A large foil area may make it difficult to thread litz wire connections through from underneath.
We have to prove that soldering to this foil will not cause damage to the power tapes underneath.
All 4 barrel services cage 'envelopes' were tried out on the 4 barrel model
Trials were done in order to illustrate the amount of space available if services cages were all left on each barrel at the time of assembly.
here is a view of the services cage 'envelope' - without services - on barrel 3 only
Here is a view of all 4 services cage envelopes - without services attached to the 4 barrel tooling rings
The first final harness clips have been made
The clips are made by moulding VECTRA A 130. The moulding has proved accurate
and within spec. with only a very small amount of finishing work necessary.
Here is a photo of the RH harness clip
Here is a photo of the LH harness clip
See also clip type, order and location on barrel
NOTE all types of clip are the same for each barrel
It was necessary for people at CERN to know the mapping of the LMTs from the 4 barrel assembly in the SR building so that cabling for testing could be installed
Here is the full 360 deg LMT mapping diagram for one barrel end showing the layout of each harness for each barrel and quadrant.
NOTE that the naming of Quadrants and harnesses ('staves') may be different from the 'physics' terms for the same regions!
The temporary PPB1 connected to the permanent PPB1 forms a long box shape approximately 318mm long x 42mm x 26mm.
Each LMT harness terminates in such a box, therefore there are 22 per octant. Accommodating 22 of these during testing and installation (allowing access for disconnecting and reconnecting) may be a problem as they would have to occupy no more than 1/8th of the space around the barrel. Trials with 22 model connector boxes were done.
Here the LMTs are hung on a temporary support with the connector boxes - these would be fixed to the support rather than left looseduring the testing phase.
It can be seen that the 22 boxes occupy a large amount of space but the arrangement could allow for some overlapping
Here is a diagram of the possible arrangement of the connector boxes during testing (.ppt file).
See also note of 17.01.2003 for layout of LMTs from TE
During the installation phase, the boxes will have to be stacked together within a very limited space inside the radii of the Thermal Enclosure, at the barrel ends.Here the 22 connector boxes for one octant are stacked as closely as possible. As LMTs have to be folded within the same space there does not appear to be sufficient space for a total of 176 within the constraints at the barrel end. Removing the Temporary PPB1s may be the only solution, leaving the permanent PPB1s and LMTs in the stack.
Diagrams showing the permanent harness clips locations in Z along the barrel have been produced relating harness clip type with it's location
Here is a diagram of the clip orders and types (.ppt file) with recess dimensions, numbers of fibres and LMTs served by the clips.
Here are diagrams of the LH and RH clip locations - distances from Z=0 (.ppt file).
Barrel 3 end clamps fitting
The manufactured version of the Barrel 3 end clamps which serve as LMT clamps, cooling manifold brackets and fibre/DCS wire routing were found to require packing of various thicknesses to produce an efficient clamping effect during trials on B3 when at Geneva. Here is a diagram of the packing required, with alternative arrangements (.ppt file) as a result of trials done on the barrel 3 cylinder at Geneva.
NOTE that this should not be necessary for barrels 4, 5 and 6. Clearance between barrel surface and inner radius of clamp can be minimised and the LMT recess
should have a radius on each inner corner rather than a sharp right angle to effect better clamping.
Barrel end clamps - fibre and DCS cable routing
DCS cable from temperature and other monitors on the barrel has to be routed around the barrel end circumference sharing the same routing space on the barrel end clamps as the readout fibre but exiting at different points from the fibres.
Here is a photo of the trial done to see how many DCS cables (twisted pairs, each single cable 0.54mm thick) will fit together with 3 sets of 18 fibres from 3 harnesses in the same space. 12 twisted pairs of DCS cables could be fitted under the clamp on the right which holds the maximum number of sleeved fibres (54).
here is a diagram showing the optimum routing of DCS cables (.pptfile) from the barrels, through the TE bulkhead and out to the DCS connector PCBs at PPB1. The best position for the DCS connector PCBs would be at 90 degs from the horizontal, with cables routed straight outwards from the TE bulkhead.
Any surplus cable can then be wound up and stored flat in a suitable location between the TE bulkhead and PPB1.
Latest Harness layout reverts back from layout of 30.08.2002 to the basic layout of 09.06.2002, with the 3 fibres from each dogleg running together then separating into 2 readout and 1 TTC fibre sets before the fibre PPB1.
View of whole LH model harness
Close-up vierw of 2LH doglegs and fibres held in position with Temporary clips (moulded plastic).
Another close-up but with fibres held by the Permanent Clip made of PEEK. Note that the fibres runclose together along sticky Kapton tape which keeps them in ribbon form.
View of whole RH model harness
View of part of RH harness. Note that the fibres are routed as close to the upper edge of the harness as possible.
Close-up of 3 doglegs showing fibre loop with temporary clips.
View of RH harness with Permanent PEEK clips
Low Mass Tape Configuration after 4 Barrel Assembly
Photos of the sequence in which the LMTs should be connected at PPB1 One octant at the +Z end is shown. Note that the PPB1 boards are represented by material with the correct length and width and the harness numbers and 'stave' (position in phi) numbers are clearly shown on each.
Here is a diagram showing the schematic layout of sets of LMTs at the TE and PPB1 for one octant (.ppt file).
Further work on trials with PPB1 boards and connectors
This shows the LMTs going over the connectors and type II cables - Note that one of the FCI connectors has become disconnected.
We have found that the power (FCI) connectors (blue) are easily disconnected - any slight movement of a stiff power cable can cause this. To solve the problem, a special connector clip was designed. This fits over one row of four mated connectors (2 FCI and 2 JST) and should, provided there is also adequate strain relief on the cables, prevent any movement or disconnection of both FCI and JSTs.
Here is a photo of the clip in place over the connectors
and another view from the side
Fibre Routing at PPB1
There was found to be a severe limit on the clearance volume in the region where fibres leave the PPB1 connector area and enter the ducting on the cryostat wall.
The fibre MT connector block is modelled here
Optical fibre is in ribbon form after each splice (seen here on the left) and each ribbon (part of the barrel harness, 250mm long) is connected to long fibre ribbon which exits through ducting by means of MT12 connectors at PPB1. These connectors fit together to form a single block called Fibre PPB1. There is one for each octant.
Here is another view of the model connector block
Here is a view of the trial routing the fibre ribbons from the connector block, keeping them within the limit of the clearance volume, to the channel which will form the ducting between other services.
Another view of the same trial
The first design of the Barrel 6 Support Bearing (Geneva, G.Barbier) was machined at UCL out of Al tooling plate. The final bearing will be carbon fibre.
Here are some views of the machined bearing
Here is the .ppt file with notes on trials made with the bearing on the models
and photos of the bearing trials:
Here we can see the bearing on the barrel 6 flange. Note that, with the tapes moved it can be seen that the bearing clashes with the rectangular white tube which represents the end-barrel alignment light path protection tube.
A view of the bearing on the B6 flange showing the gap between the flange and the Thermal Enclosure Bulkhead Note that the four recesses for the low mass tape harnesses need to be enlarged so that the tapes can exit straight off the barrel surface without 'pinching'. The recess edges should not be sharp. Adequate clamping of the tapes is needed here.
A view of the bearing on the surface of barrel 6 showing the gap between the surface and the Thermal Enclosure Bulkhead from the inside.
Fibre routing trials on the bearing:
A view from the barrel side of the bearing showing
fibres from four harnesses routed along th bearing. Note that small clips similar to those used for the fibres on the barrel end clamps will be needed to keep fibres in place on the bearing.
Another view of the fibre routing
A view from outside the barrel showing the fibre routing through the Thermal Enclosure Bulkhead
PPB1s with type II cables on cryostat wall
Two PPB1 boards with cables connected up are positioned on the cryostat wall. We found that the long support board would need a profiled bonded surface in order to adhere permanently to the inner radius of the cryostat wall as the whole structure is heavy. The low mass tapes will need strain relief before the right-angle bend onto the cryostat.
Trials with real PPB1 connectors and real TYPE II cabling
Type II cable ends were stripped and ferrules fitted.An attenpt was made to fit the cables with ferrules onto a PPB1 board. It can be seen that there is not enough space in the clamp (strain-relief) region for a row of 6 cables fitted with strong ferrules.
The 6 cables without ferrules fit in the space
Here is a sketch of 'ferrule clamp' - a possible solution to making a secure clamp for the Type II cables?
This shows stripped Type II cables clamped - the straight bar provides inadequate clamping force
A view showing the length of stripped cable needed to enable connecting We noticed that the individual cables make a VERY stiff bunch, the power cables being extremely stiff.
Another view showing stiffness of cableswith one JST (white) and one FCI power connector (blue) connected up on a real PPB1 board.
Fibre routing through the Thermal Enclosure Bulkhead - assessment of size of apertures required for fibres from one octant only
The .ppt file with diagrams of the fibre numbers and routing for one octant of one barrel end can be found here
Shown here is a fibre bundle in a 65 x 5mm slot placed above the approximate region in the TE bulkhead where it would be located.
There are 342 sleeved fibres in this bunch, from one octant of all four barrels.
The remaining 54 fibres (part of the set from barrel 6 go through an Auxilliary slot) further along the bulkhead - and curve round to join the main fibre duct.
An alternative slot shape 32 x 10mm for the main bunch of 342 fibres was made.
65 x 5 bunch on bench
10 x 5 Auxilliary bunch on bench
32 x 10 bunch on bench
NOTE that these fibre bunches, like all other services, will have to be sealed in special grommets at the TE to prevent gas /thermal leakage - this means that the slot sizes will have to be larger than shown here.
Cooling capillaries - new routing brings the capillaries alonside exhaust pipes (heat exchange proposal)
The capillary routing remains unchanged from the previous version in the region from the barrel ends and through 4 exists per end through the thermal enclosure bulkhead.
Outside the thermal enclosure, the capillary route changes, with the capillaries bending round to join the exhaust pipes
which serve the same barrel cooling units. Note that the exhaust pipes have connections at the barrel end and the thermal enclosure bulkhead and again beyond the right-angle bend in the PPB1 region - BUT the cooling capillaries are continuous without connections until the PPB1 region.
A close-up view of the region where capillaries meet
exhaust pipes showing a possible layout.
Trials with separate PPB1 covers on the model cryostat
A view of a set of 6 PPB1 boards (3 pairs) with separate covers on support boards fixed to the model cryostat. the covers screw onto the PPB1 boards and the PPB1 boards are stuck to the cryostat wall.
Another view of the set of 6 FRONT and REAR PPB1s with covers on the cryostat
First trials using Type II cables on model PPB1 boards fixed to the support board.
6 Real Type II cables were fixed to a model PPB1 board screwed onto a model support board. This shows the FRONT board which goes onto the support board first
Here is a close-up of the modelled connector area on the front PPB1 board.
The REAR board with LMTs passing OVER the FRONT board.
This shows trials with the first design for covers for the PPB1 boards. Each board has a separate cover.
A view showing the construction of the covers
New model harnesses have been made.
An annotated photo of a LH harness showing the new version of the fibre routing requested by Taiwan in order to separate the TTC fibres into a separate set along the harness length.
A close-up of the LH dogleg showing the TTC fibre on the right of the 3 - routed up to the separate set of fibres. A permanent harness clip is shown here.
Close-up of harness with TEMPORARY clip and with fibres on kapton adhesive tape.
The first design of the 'tandem' PPB1 full-length support plate allowsfor a few mm sliding adjustment.
Here is a full length view of the back of the support plate.
A close-up view of the sliding section between fron and rear boards
A view of the model octant with low mass tapes arranged on the cryostat with front and rear sets spaced as for sliding plate
All parts required for the PPB1 sliding support plate on the bench
Dimensions in Z of dogleg to interface PCB to LMT end for LH and RH doglegs have been measured
Here is a drawing of LH and RH doglegs with interface PCB, LMT end and tall parts of the dogleg to LMT connector shown.
Note that the main dimensions are with reference to the BRACKET CENTRE line. (Dimensions in grey are wrt dogleg centre line as on the original dogleg drawing by Mark Jones).
The highest parts of the connector are shown.
These measurements were made in order to get an accurate distance of the LMT end from the MODULE PHYSICS CENTRE as the these positions are used to define the
tape lengths on the barrel.
An updated list of LMT lengths is now being made.
Routing of cables from temperature sensors and other sensors (DCS)
Sensors on the barrel will have cables running from barrels over LMTs at the barrel ends, through the thermal enclosure bulkhead and out to PCB connectors. Here is a very preliminary layout for the DCS cable routing from cooling pipe temperature sensors.
Taiwan - latest dimensions of the readout fibre splice protector and furcation tubing
Here is a drawing (not to scale) of the Taiwan splice protector sample with the transition from fibre sleeved in furcation tubing to bare fibre to splice inside the splice protector to ribbonised fibre . The whole length is protected by heatshrink tubing 'SUMITUBE 125deg C OFT F diameter 5.0 VW 1 SUMI-PAC CSA HS X'
Note that the yellow and other coloured furcation tubing used was found to allow light leaks so the decision was made to use black furcation tubing - however Taiwan notice that the inside and outside diameters are greater than the coloured tube diameters. (O.D. for yellow tubing was 0.9mm).
The multicoloured sleeving on the ribbonised fibre must now be covered where it comes close to the module detectors. This length will be included in the revised table of fibre lengths requested by Taiwan to allow for a 2nd repair between splice and dogleg (table of lengths previously allowed for one repair only).
Valencia SCT Week report on status of model work and end of meetings update
Status report page 1
Status report page 2
Complete barrel 3 model harnesses
Here is a photo of two complete barrel harnesses (LH and RH) to the latest dimensions with latest fibre layout and length, with spliced sections and latest dimension model LMT boards.
Here is a close up of a section of the harness showing temporary harness clips made of moulded plastic (Oxford).
Services cage - trials with 230mm long LMT box
An additional outer support ring is necessary at each end of the barrel, at the
edge of the services support extensions. This means that the length of the boxes housing the services has to be reduced from 250mm to 230mms. Trials were done to determine if it was possible to accommodate a barrel 3 harness (the longest) in a box 230mm long.
Here is a view of a box 230mm long with real, correct thickness tapes (B3 harness length) folded, terminating on the connector board immediately outside the box on the right.
NOTE that the minimum radius of the tapes is now 13mm.
Here is another view of the box with the connector board lifted at an angle for type 2 cable connections to be made on the underside.
End-Barrel 3 model sector with harness interlink
The model -Z end sector with 2 doglegs, fibres for 2 harnesses, the harness interlink which links 2 harnesses and lies under the fibres, together with the folded dogleg redundancy arm. This arm at the harness end is unused as the interlink is connected instead.
PPB1 region showing the new length model LMT connnector boards with real type 2 cables
Here is a view of the model cryostat wall with the two connector boards. The 6 real type 2 cables are stripped back to allow access for the 6 connectors. The sleeved cables are held togther using strain relief brackets (not shown). Note LMTs going to the rear connector (eg barrel 3) pass over the LMTs, stripped cables and connectors of the front connector board. Type 2 cables from the front connector board (eg barrel 5) go between the plate attached to the cryostast and the rear connector board.
The LMT connector board 268mm long x 38mm wide
New design of fibre duct (Ian Wilmut RAL)and cooling capillary tube layout
Here is a section on the 360 deg end barrel model showing capillary tube routing up to and beyond the Thermal Enclosure and the layout of the fibre ducting (for readout and alignment fibres) up to the thermal enclosure. Not that this is as it would appear BEFORE the addition of the cooling exhaust pipes which would lie over the other services.
Lists of items with dimensions to be stored on the B3 services cage
Services which run from the barrel to the patch panels have to be stored on services support extensions at both ends of each barrel, within a limited space envelope. Low mass tapes are stored in enclosures on these extensions (cages)
with optical fibres, DCS wires and cooling capillaries stored in enclosures
between each set of tapes.
Here is a provisional list of items for the services cage at the barrel 3 +Z end
Here is a provisional list of items for the services cage at the barrel3 -Z end
note that the lengths given on these lists are ONLY the lengths in need of storage on the services cages.
Here is a photo of a possible layout of low mass tapes in the enclosure
The new long LMT connector board is now 268mm long
Here is an updated dimensioned drawing of the PPB1 area along Z which replaces that of the previous entry to allow for the extra length.
The new long LMT connector boards for PPB1 have been modelled and assembled on the vertical quadrant sector
The sector - see previous entry for February 2002 had to be modified by lengthening the section of the cryostat in Z to accommodate the long boards.
Here is a dimensioned drawing of the area
Here is a photo of the front and rear connectors, each 260 (+5mm extra for strain relief extensions) long with a 40mm gap between the front board and the rear board - with the 'old version' 90mm long connector boards either side. The new connector boards start at Z = 880 whereas the 'old' boards started at Z = 800.
Note that starting the boards at Z = 880 creates more space for Optical Fibre connectors and their support brackets in the space in front of the tapes where they curve through 90 degs. It also allows for the horizontal tape cover plates
to nearly meet the vertical heat spreader plates so that they can be connected using Al Kapton strip (grounding and shielding) This will be modelled later.
Here is a view of the Rear board and the Front board
Note that the cables used on the model are not the real cables but have a similar diameter. They will be replaced with real cables and more realsitic model components. The outer sleeve of the cables is stripped back 'n' mms from the connector.
Latest news: the outer isolation of the conventional cable will be removed up to the cable clip to provide electrical contact between cable shield and PCB shield layer. The cable strain relief will be between the front and rear boards and beyond the rear board.
Here is a view of the boards from under the cryostat
and photo of the whole octant with tapes from all four barrels going through the thermal enclosure bulkhead out to PPB1 old and new.
The boards will be adjustable by approx 10mm in Z in order to take up excess length where this occurs (see entry below - 'heat spreader plates' for reason)
Low mass tapes and model connectors were assembled on the new vertical quadrant model and two perspex sheets were used to model the heat spreader plates which 'sandwich' the tapes from the thermal enclosure bulkhead region out to near PPB1
Here is a photo of the model showing the space occupied by 90mm long boards (MOLEX connectors), the front and rear boards separated by a 20mm gap. The connector boards start at Z = 800.
Here is a photo of the tapes sandwiched between two Perspex sheets as they would be sandwiched between two Al heat spreader plates.
The difference in length between tapes of the same barrel at different locations in phi means that allowance has to be made for some excess length (if all tapes for the same barrel are to be made the same length). An undesirable wave effect is created between the plates where for good grounding they should be in contact with the plates along their whole length. The excess will have to be taken up elsewhere.
A set of Castellation Clamps have been tried out on the small barrel 3 model
The castellation clamps (also known as barrel end tape clamps and cooling brackets) serve as support brackets for the on-barrel cooling units.These clamp/brackets are screwed to the barrel flange castellations.
Each cooling exhaust manifold is supported in two places by these brackets and each input capillary tube is supported where it joins the barrel.
In the region of the U return sections of the cooling unit no support at the barrel end is supplied as these are intended to be 'floating'.
Here is another view of the model section.
Here is a view of the castellation clamps during fibre routing trials.The fibres are held in place with small clips screwed to the clamps. Note that the on-barrel harness clamps used here are a variety of prototypes, not the final version.
A New Quardrant Sector of the cryostat, PPB1 and the TRT services envelope has been constructed and fixed to the four model barrels now assembled together and is being used to test harness lengths
Model harnesses for one octant of the four barrels (Right-hand and Left-hand) are being made and tested on the four assembled barrels with the patch panel sector.
Here are two model harnesses on the bench, complete with optical fibres and low-mass tapes of the correct length.
Here is a section of the optical fibre part of the harness showing the fibre splice protector .
Here are two harnesses (LH and RH) on the full-scale barrel 3 model before barrel assembly.
The four barrels assembled with the model sector of the cryostat and TRT services envelope (green) and with the model harnesses being fixed in position.
Provisional Optical Readout Fibre lengths from opto package to splice and from splice to PPB1 connector have been measured
Here is a drawing showing the positions of the fibre bunches and splices on the barrel end
A provisional list of measured readout fibre lengths
List of Left-Hand harness fibre lengths
List of Right-Hand harness fibre lengths
Note correction to RH Barrel 5, harness 2, dogleg 6: length is 288 not 188.
Refer to the above drawing for positions of fibres, splices and ducts. Note that the Barrel Harness Overall Layout drawings on the RAL system (A.Nichols)
(0-TB-0049-272 to 0-TB-0049-279) will be on CDD.
The 360 deg Barrel end model has been updated with the following featu
res:
The new 'expanded' layout in Z has allowed for greater clearances between low mass tapes/fibre ducts/alignment features/cooling pipes.
Here is a diagram of a cross-section in Z of the barrel end services showing the new spacing.
Here is a view showning a section of the model with Thermal Enclosure Bulkhead.
Here is a view of a section of the model.
Two sets of model capillary tubes have been installed serving one quadrant.
Here is a photo of the 2 capillary bunches
The low mass tapes have to be wrapped in shielding material between the thermal enclosure bulkhead and the connectors at PPB1. 2 harnesses have been wrapped with the 50 micron AL/Kapton sheet.
Here is a photo of the wrapping sequence
and a photo of the two harnesses wrapped - resulting in 2 bunches (together) being 4mm thick.
01.10.2001
The design of the permanent harness clamp has been modified to fit with the new design of the bracket pads.
Here is the plan view for the clamp. The side sections have been reduced in width to minimise material and there is now a fibre hook on both sides of the clamps - to serve both LH and RH harnesses.
Here is a section through the clamp. The thickness of the sections that go under the bracket pad extensions has been made thicker as the pad extensions are higher off the barrel surface.
Here is a photo of the new prototype machined out of PEEK and a photo of the clamp on the model barrel with bracket pads
31.08.2001
New drawings of the permanent harness clamps have been made.These secure the harness on the barrel cylinder surface and are held in position by the bracket pad extensions.
Here is a plan view of the LH clamp For the RH harness the 3-fibre hook is on the side marked D
Here is a section through the LH clamp
Note that there are 6 different types of clamps for each side (LH and RH harnes) to secure the different numbers of readout fibres joining the harness:
Clamp 1 nearest Z = 0 for LH harness - 3 fibres
Clamp 2 - 3 fibres through clamp, 3 under hook
Clamp 3 - 6 fibres through clamp, 3 under hook
Clamp 4 - 9 fibres through clamp, 3 under hook
Clamp 5 - 12 fibres through clamp, 3 under hook
Clamp 6 - 15 fibres through clamp, 3 under hook
The new drawings of the pads by Gerard Barbier for the Test Sector are now on CDD: ATLISBB_0012 and ATLISBB_0013.
07.08.2001 and 28.08.2001 (see NEW drawing)
Prototype temporary harness clips have been made. These secure fibres and tapes in position before and during harness installation on the barrel and are then replaced by permanent clamps made of PEEK or VECTRA A130.
The temporary clips are positioned with the centre at the same Z as each redundancy arm connector of the doglegs and can remain in place whilst the permanent clamps are installed after the harness is positioned on the barrel as these are centred on the centre line of the dogleg in Z. The 5 (white) temporary clips are then removed.
Other temporary clips can be removed at later stages :
the red clip is removed when 2 harnesses are in place to allow fibres to be routed away from the tapes at the barrel end
The blue clips which hold the 6 tapes together off the barrel end can remain in place on the services support cage until tapes are folded up during the barrel installation.
one harness with clips
4 clips: 3, 6, 9, 12 fibres
one clip shown open
one clip detail
Detail showning fibre/tape position
NEW drawing of clip with dimensions - note each of the six clips is slightly different - the width of the fibre guides varies with the number of fibre and the thickness of the low mass tape guides varies according to the number of tapes held by the clip. The outside dimensions of the clips do not vary.
When opened flat, the clip is 70mm x 25mm x 2.0mm. When closed on the harness the clip is 33.5mm x 25mm x 4.0mm
the base for the clip is made of plastic sheet which can be scored and folded without breaking at the fold. The extra heights for tape and fibre guides are made of plastic sheet of various thicknesses glued to the base. Ideally these
should be moulded plastic - 6 types should be possible as a lot will be needed.
Note that the use of VELCRO bonded to the plastic clip as an easy method of fastening may not pass cleanroom standards - so an alternative to this may be needed unless a cleanroom standard VELCRO is obtainable.
05.07.2001
A new version of the multi-structure has been drawn
Here is the drawing of the structure which serves as harness end-clamp, fibre routing from harness to duct and cooling support (both Exhaust manifold support and cpillary input pipe support.
Here is a detail drawing of the clip that holds the fibres on the structure en-route to the fibre ducting off the barrel end.
14.06.2001
A model of the new version of dogleg 0 was tried out on the small sector - the new version has the new outline shape. The new prototype clip is also shown - this is designed to SLIDE sideways from the right into position under the bracket pad extensions.
Here is a photo.
07.06.2001
Here is a list of readout fibres and splices for each barrel, including dimensions of fibre tubing etc.
LIST of readout fibres and positions
A Schematic of fibreand splice layout from barrel to PPB1
05.2001
Readout fibre ducting at the barrel end - The RAL setup for fibre splicing requires 8 cms of bare fibre in front of each splice. The actual splice area is protected by a length of heat-shrink tubing 40mm long.The distribution of these therefore restricted to areas where there is enough space. A model of the fibre duct, serving 1/8th of one barrel end is shown with areas coloured to indicate fibres:
yellow = single fibres in furcation tubing
red = fibre splices
purple = fibre ribbon
A second fibre duct is shown in green - serving the other half of the quadrant. Note that the fibre duct is now positioned OUTSIDE the tapes ie between the tapes and the cooling pipes in Z. Please refer to a new drawing of the barrel end services region for the new layout in Z (see August 2001)
Here is a full view of the 360 deg. model of the barrel end with fibre ducts at the far quadrant
Here is a closer view and a detail of the model fibre duct. Note that the model duct is 5mm thick and 45mm wide max.
Prototypes of on-barrel harness clamps - here is a photo These have been machined from PEEK but production clamps could be moulded from VECTRA A130. See drawings below.
New drawings of the on-barrel harness clamp - version 27.04.01 have been made to allow for the greater number of sleeved fibres on the harness and also for the early first bend in the dogleg from the barrel surface towards the fixing points on the bracket.
This drawing shows a plan view of the clamp
This shows a section through three clamps which allow for the varying number of low-mass tapes and the varying number of sleeved fibres. Note that the earlier version allowed for 6 different heights and this may still be possible - depends on the method and cost of manufacture.
This shows a detailed section through the clamp for the dogleg nearest the barrel end where tape and fibre numbers are greatest. A prototype is being machined out of PEEK at UCL but the final manufacture will probably be moulded.
Note that the clamp does not 'pin' tapes, dogleg or fibres to the barrel. It's purpose is to hold the harness in place whilst allowing some movement in Z
due to thermal changes. Strain relief is applied at the barrel end by means of the 'multi-structure'.
The prototype 'multistructure' at the barrel end is shown on photos of the short 1/8th sector model in items below. Here is a drawing of the structure . More trials will be done on this and a new low-mass version will be machined.
11.04.2001
note - this reversed layer harness proved not to be necessary
Trials with reversed harness at barrel end
If the assembly of the harness has to be done with the dogleg closest to the end of the barrel first, (ie closest to the barrel surface) then tapes from the other 5 doglegs have to go OVER the dogleg-to-tape connector rather than flat against the barrel surface as shown in the previous item of 28.03.2001.
This reversal means that tapes are compatible with the current connector layout at PPB1.
Here is a photo of the short 1/8th sector model with 5 tapes going OVER the connector. NOTE that blocks of the correct height are used to represent the connector but the other dimensions are not known. It is important that the connector or board does not extend beyond the edge of the tapes on the side which would be under the lowest corner of the module.
The height of the 'bridge' of tapes over the connector is just less than 6mm, BUT the 6 harness tape clamps are designed so that tapes can move under them - to allow for thermal changes on the barrel, so in theory the tapes could move to form a bridge over the connector which was higher than 6mm at the barrel end. The PEEK 'multistructure' at the barrel end of course secures the tapes in place.
.
Here is a view from above
Here is the same layout with sleeved fibres added
.
28.03.2001
Optical fibre on the barrel harness
Each optical fibre (data) from the optopackage has now to be sleeved separately using 900 micron furcation tubing. This means that, with 3 fibres per dogleg, there will be 18 sleeved fibres after the last dogleg at the barrel end. These have to be ribbonised (12 to a 12-way ribbon and 6 to a 6-way ribbon, then the ribbonised fibre has to be joined (spliced) to the fibre ribbon which travels out to PP1 where they connect by means of Infineon MT12 and MT8 connectors to long
fibre ribbons which go out to the electronics area outside the detector.
Here is a photo of an attempt to route fibres past the last dogleg connector, with all 18 fibres following the same route.
Here is a drawing with x-section through the sleeved fibre bunch.
Here is another attempt to route sleeved fibres past the last dogleg connector,
this time with fibres in 2 bunches, routed round each end of the dogleg connector. Note that the fibre joins are modelled using
red heatshrink tubing and the joins are positioned on the 'multi structure' which is fixed to the castellations on the barrel end. This could have channels
within the block structure which would support the joins (ie they would NOT
lie on top as in the photo)
NB please ignore the dog-eared model dogleg - these photos are only to show
possible ways of routing fibres. Note also that the black harness clamp is an old version which is not profiled to accept the 5 tapes + dogleg or any sleeved fibre - see item dated 15.03.2001 for a newer version of the clamp
new photos will be taken when the new design of dogleg can be used!
15.03.2001
Here are drawings of an updated version of the on-barrel harness clamp.
A plan view of the clamp
sections through the 3 types of clamp Low, Medium and High - to accommodate increasing numbers of tapes.
Note that the 'bridge' sections d, e and f which go over the fibre bunch will now also have to be modified to
take increasing numbers of sleeved fibres.
another version of the clamp will be designed when the dogleg 0 design is finalised
12.03.2001
End barrel services - suggestion to create more space in Z
A possible solution to provide more space at the barrel end area would be to occupy the space in Z out to Z = 792 by moving the cooling pipes and manifold to centre on Z = 783.5 (present layout - centred on Z = 780)
Drawing of the present layout of services in Z beyond the barrel flange (cooling, tapes and alignment structures).
Drawing of a possible layout with cooling pipes moved out in Z by 3.5mm. This prevents tapes from being pushed down by the cooling pipes and fittings onto the alignment structures. Note that the optical fibre ducting would go between tapes ie would occupy the same envelope in Z.
01.03.2001
Alignment jewel prototypes
Here is a photo of some prototypes machined out of PEEK at UCL
27.02.2001
Cooling exhaust pipe system
The new cooling exhaust pipe system prototype made at RAL is tried out on the 360 deg end barrel model. The 'spider' pods have to fit exactly onto the cooling exhaust manifolds at the ends of the barrels where there is space between low mass tapes. Each quadrant has to have a different configuration.
Here is a photo showing the 'spider' with barrel 3 in the front
'spider' from the side and
'spider' from the outside where the manifold would join to pipes running to PP1 outside the thermal enclosure.
09.02.2001
Short 1/8th -Z end model
First trials with dogleg/bracket/low mass tape on the short end model.
Here is a photo and
another view. This will be updated with new version dogleg etc.
27.11.00
A new version of the Barrel Harness Installation Cover
has been designed.
Here is the drawing of the cover. Note that this version is designed for harness installation before installation of cooling whereas the old first version was designed for harness installation with cooling pipes already on the barrel (see 6.11.98 and 11.8.98). This new drawing is based on the LH harness drawing TD-1006-767 by Andy Coyle (RAL).
This cover is made of a reinforced (ie does not bend) strip of clear plastic with sections formed to
the same (bent) shape of the doglegs.
This cover is placed over the fully assembled
harness on the bench (jig) and the harness attached to it using small straps
underneath the harness using captive screws. Each strap (thin plastic) has a
tab to facilitate removal using flat tweezers. The straps are postioned
on dogleg and tape areas away from connectors and optopackage. A larger strap
fastens the ends of the tape bunch securely.
The covered harness is then lifted onto the barrel (handing points on cover) and placed in position, locating the open slots around the module fixation part of the bracket (or any other suitable location point). The ends of the tapes are supported by a second person whilst this is done.
Once the covered harness is in position, each dogleg is fastened to the bracket at the dogleg fixing point, through holes in the cover, then the straps are carefully removed and the cover lifted off.
Harness retainer clips are fixed in position, held by the bracket 'feet'.The tape ends are held in a temporary position until the 2nd of the pair of harnesses is installed when the pair is joined and final adjustments made. The tapes of both harnesses are then clamped down at the barrel end.
Note that this drawing (not to scale) is preliminary as the final design of the dogleg is not yet known. The exact dimensions of the cover itself and the positions of straps will be added when the dogleg design with position of optopackage and optical fibres is known.
22.11.00
Here are photos of the 1/8th section of Barrel 3
The outer surface section with the end barrel alignment T post and its socket, the support structure for optical fibre, cooling manifold and harness low mass tape clamp. Also visible at the right is the outer part of the alignment jewel housing N214-009 (insert plug) and its locking tab.
Here is another view of the section showing the inner surface of the barrel with the alignmnt jewel insert. NOTE that the castellated flange does not have the inner cross-section ie it is represented by a flat plate. The model flange is fixed to the barrel surface plate using screws - these are of course not present in the real design
Here is a view showing a section through the barrel cylinder so that both outer and inner parts of the alignemnt jewel insert and
plug are visible, together with a view of the end barrel alignment T post showing adjusting gap in the socket.
27.10.2000
A 1/8th short section of Barrel 3 with castellations has been made at UCL in order to try out prototypes of various items listed below:
on-barrel alignment jewel inserts
end-barrel alignment unit
structure to support optical fibres/cooling manifold/tape clamp, tape retainers, brackets etc
A harness installation procedure is proposed which is possible after brackets are mounted and before cooling pipes are installed on the barrel. Here is the proposed sequence.
Here is a diagram illustrating the use of the structurein the harness installation sequence.
26.10.2000
New End-barrel schematic layout version 3
This new version of the cooling input capillaries and output exhaust layout shows changes to the routing of the cooling capillaries from the end-barrel to the thermal enclosure where there is now space to use conventional connectors. Note that no capillaries, pipes or fibres are routed acreoss the horizontal interlinks.
This will be modelled on the 360 deg model when the first prototype of the exhaust pipe 'spider' is ready for trial installation on the model.
04.09.2000
Low-mass tape connectors at PPB1
A first version of the low-mass tape connector designed by Andy Nichols at RAL
has been modelled and tried out on the 360 deg model. This version has the 6 tapes from one harness connected to a PCB (would be soldered) and this is in turn connected to a similar but larger PCB with 6 tapes from PPB2. Two sets of connectors are joined to form one block which is on a board fixed to the cryostat. Here is a photo of the model set of connectors positioned on the cryostat wall and a close-up.
01.09.2000
The positions of the in-barrel alignment units were modelled on the inside of the model barrel 4 section. Alignment units were modelled using simple blocks of perspex and the optical fibre routes were indicated by black wire (to make them visible!) and the light paths were indicated by red wire strung between the units. The first version is shown here.
Following discussions, some of the routing was altered; here is a close-up of the next version and also another view of the next version showing 'fibre' routing from alignment units to the end flange exit channels.
Note that this is not the final version of the routing as special fibre clips designed by Oxford will be made and tested here. These will secure the optical fibres/ribbons on the inner surface of the barrel so that there is no interference with the light paths and fibres are held in a safe bend radius.
21.08.2000
A prototype on-barrel alignment jewel insert, plug and locking tab designed by Oxford and made at UCL. Here are photos of the PEEK items -
scale indicated by a pound coin.
The three separate components and the jewel insert and plug together with locking tab separate.
The three components together . Three components in locked position.
15.08.2000
Alignment jewels, refectors and housings on the barrel end have to occupy the same space as the low-mass tapes in Z (see diagram (01.08.2000). Here is a photo of the prototype T post and socket and one reflector housing (machined out of PEEK - UCL workshop)
A side view showing adjustment screws in the socket
The T post with model light-path tubes and reflector housings was mounted on the 360 deg barrel end model with the socket holding the T post mounted on the end flange of barrel 5. The reflectors at the ends of the light-path tubes (model square section white plasitc tubes either side of the T section) are mounted on the ned flange of barrel 6. Here is a close-up.
The tapes have to fold over the alignment post after installation of the barrel.Here is a photo of the low-mass tapes and the T post.
02.08.2000
A structure which serves as a support for optical readout fibres/tape clamp and cooling manifold support has been made out of PEEK. The structure is designed to be fixed to the castellated part of the barrel end-flange and serves as a clamp which holds the low-mass tapes in position at the barrel end. Optical fibres are routed via the structure to the fibre ducts and the cooling manifold is fixed to the outer radius of the structure at one point.
Here is the structure on the end flange (without castellation) of model barrel 3 (with the tooling ring in the foreground) and another view showing the fibre route (precise dimensions of this fibre channel are as yet undefined). Here is the structure showing where cooling manifold may be supported .(On barrel cooling pipes and manifolds design not finalised)
01.08.2000
A diagram of the end-barrel services was made to INCLUDE the newly designed end-barrel alignment units after preliminary trials with designs for the alignment T post and alignment light-path tubes and reflector housings. The end-barrel alignment has to share the space envelope assigned to the low-mass tapes. The end-barrel services diagram shows the relative positions in Z.
23.06.2000
Here are the numbers of alignment fibres for each barrel
as given by Armin Reichold.
Trial with fibre channel through barrel end flange
A model section through a barrel end flange was made with a special channel
having dimensions exactly as shown on Gerard Barbier's drawing: SCT CYLINDER 3 (Part 1) 252632P0(1) detail E.
The model section was made of aluminium
with the channel part inset into the flange section. Optical fibre ribbon
and mock DCS wires were passed through the channel into the route they will
take round the barrel surface. The fibres on the 'inside' of the barrel were
stuck down, allowing for a curve through the channel of >25mm radius.
The fibre ribbons consisted of 2*12 and 1*2 way. This is two fibres more
that the maximum quoted on Armin's list which was 1*12, 1*10 and 1*2.
10 mock DCS wires were routed through together with the fibres.
Here is a photo of the INSIDE of the barrel with fibres and DCS wires going through the INSIDE of the flange.
Here is a photo of the OUTSIDE of the barrel with fibres coming out of the barrel surface and routing round the barrel.
Conclusion: The design of the channel through the barrel end flange and
surface of the barrel works well. The fibres can curve through a safe radius
and as all channel surfaces are smooth the ribbon is very easy to thread
through. The channel size is adequate for the maximum amount of fibre ribbon
and present estimate of the size of a set of 10 sleeved DCS wires (0.5mm diameter each).
If possible, fibre ribbon should not be as
small as 2 fibres only as the very thin ribbon easily separates from the bunch
and could get caught up in things. A ribbon of 8 to 12 fibres is best.
The fibre ribbons and DCS wires were stuck to the inner surface using tape but
it would be safer to use perhaps tape made of carbon fibre, stuck down with
a good adhesive. This is especially important close to the where the fibres/wires curve to enter the flange channel so that they do not curve out beyond the radius of the flange (installation hazard); this won't be a problem if they are
firmly anchored.
09.06.2000
Additions to 360 deg end-barrel model
A full 360 deg model of a section of the cryostat inner wall has been fitted to the 360 deg end-barrel model.
Here is a closer view showing the layout of low-mass tapes . Low mass tape connectors will be supported here at PPB1. The design (Andy Nichols - RAL) for these is under development.
23.05.2000
New cooling pipe layout
As a result of cooling tests in May, the layout of cooling pipes
had to be changed. Previously it was advantageous to have the input
capillary tubes running close to the output exhaust tubes. Here is a close up
of the model thermal enclosure bulkhead area with capillaries routed close to exhaust pipes. This arrangement had to be changed so that input and output pipes are kept separate coming off each barrel, through the thermal enclosure and up to PPB1. Each set will require separate insulation. Separating the capillaries from the exhaust pipes at the barrel ends is possible as can be seen from this photo of barrel end area
with capillary tubes separated from exhaust tubes. This will occupy more space at the thermal enclosure bulkhead.
Here is the updated drawing of what the 360deg new layout
would look like. Note that it has not yet been decided if the capillaries will go over or under the tapes. This will be tested on the new model.
Here is the updated schematic of the exhaust/capillary
layout for +Z and -Z barrel ends .
19.05.2000
New 360 deg end barrel model
It was decided to build a full 360 deg structure because of the differences
in layout in each quadrant; cooling pipes and manifolds, power tapes, readout fibres, alignment units and fibres, DCS wires etc each having different requirements. It is possible that this model may be used in the full-scale ID services
model under construction at CERN. It's initial use is to provide a means of
testing layout and mechanical designs for low mass tape connection, cooling
layouts, thermal enclosure bulkhead, readout fibre, alignment units, PPB1
etc. Here is a view of the basic structure and
the model with a full set of 1056 tapes.
Here is a dimensioned sketch of the model structure.
05.05.2000
possible solution for tape clamp
Here is a drawing showing how the various thickness of low mass tape could be allowed for in the construction of the tape clamp. The clamp is designed to hold the tapes against the barrel surface at the junction with the dogleg without stressing the bracket inserts in area where the tape
layers are thick.
05.2000
Thermal enclosure sealant tests
(further tests are planned using a pressurised box to test the leak rate)
A preliminary test of the sealant that could be used to seal tapes in the thermal enclosure has been done. A small amount of sealant (silicone rubber with silica filler) was put in between each tape, the set of tapes were enclosed in a specially made grommet (silicone rubber) designed by Andy Nichols (RAL), then more sealant was used to seal the grommet and tapes into an enclosure. The sealant was left to set, see photo, then later the grommet and tapes were released from the enclosure and sealant removed, without damage to tapes, grommet or enclosure, see photo thus proving that this is a reversible process.
25.02.2000
Changes have been made to the Quadrant model
The model now has all components in the correct Z layers,
indicating how low mass tapes and cooling pipes occupy the 20mm space
available.
Alignment light-path protection tubes have been added (white square section
tubes in photo). The low-mass tapes and the optical fibre ducts have to share the 10mm space in Z
with these tubes. It was found that the optical fibre ducting was in the way of the planned light path, so the ducting has to be bridged over this and the depth of the duct decreased. The optical fibre ribbons, if neatly packed, will still fit in this shallower duct.
Here is a view of the whole end quadrant showing new
exhaust piping bent to fit and extending up to the PPB1 area.The space available in Z for cooling pipes is restricted to 10mm. The thermal enclosure region has yet to be modelled in detail. Capillary tubing is now
routed from the barrel end UNDER low mass tapes until it is supported by the exhaust pipe of the same cooling unit on the way out.
A sketch has been made of the whole 360 deg view of the assembled barrel ends out to the thermal enclosure. This shows how the cooling pipe routing is arranged together with low-mass tape routing and optical fibre ribbon routing. Note that the top section of this drawing is the section
modelled - hence horizontal interlinks are at an angle!
The colour-coding for the exhaust and distribution manifolds (often mostly
hidden behind tapes) are the same as indicated on the model.
21.01.2000
Evaporative cooling pipes now on Quadrant model
The orientation of the horizontal interlinks has been established and
cooling pipe input capillaries, exhaust and distribution manifolds
and exhaust pipes have been added -
seen here in a general view also in close up view of the
octant adjacent to the long 'horizontal' interlink
and in the octant on the right
The cooling pipe manifolds can just be see between the tapes -
exhaust (orange) and distribution (green)
The optimum layout for the cooling units has been
established and can be seen in this drawing
showing layout for both +Z and -Z ends of the 4 barrels
A complete unit of 4 cooling pipes has now been modelled on
barrel 3
The cooling pipes can be seen at the +Z end
of the barrel and here is another view
Here is a view of cooling pipes at the -Z end
with another view. Note the small capillaries
at the ends of the two pipes either side of the exhaust manifold.
26.11.99
Here are some preliminary views of the evaporative cooling pipe unit
distribution ends, optical fibre route, low mass
tapes off barrel end and another view of same
area. The other end of the barrel has
the exhaust manifold
Here is an updated picture of the quadrant model
as it was used by A. Nichols of RAL to make a CAD drawing of the
low-mass tape configuration.
14.10.99
Armour for the optopackage
A trial cover for the optopackage was made out of
PTFE (could be molded plastic) and fitted on a dogleg with a model
optopackage. The final dimensions of the optopackage are not known, so
this is a very approximate model. An old discarded dogleg was used with
'old' 18mm tapes. Here is a drawing of the trial
cover , top and side view. The fibres emerging from the cover could
be sleeved with a cover made of KEVLAR which follows the curve of the fibre
from optopackage to where it joins the optoribbon. Here is
another picture with the fibres covered, as seen from the back.
The cover curves outward and passes from back to front of dogleg.
17.09.99
First assembly of barrels 3, 4 and 5
Barrel 3 was equipped with Oxford designed service cages at either end
and a row of brackets and some modules on the end row were installed.
A section of barrel 4 was installed with some brackets and modules, then
a section of barrel 5 was installed over this. Interlinks between the
barrels were attached, but this could not be done in the section where
there were low mass tapes as there was insufficient space to fit the
interlink without damaging the 20mm wide low mass tapes. The interlinks
will have to be less than 35mm wide in order to fit.
Here is a view of the assembly from the +Z end
and a view showing cages at both ends
Here is close-up view of the end with modules mounted on brackets and with tapes - one set folded onto the cage with a model
connector on the end.
Note that barrels 4, 5 dimensions and 6 and interlink
dimensions are under review
and the dimensions of these model sections will be changed accordingly.
The length of barrel 3 is currently being changed (increase of 10mm each end)
27.07.99
Optoribbon bunching with relation to
Off Detector Electronics requirements -
Tower mapping or Layer mapping?
Optoribbon bunches will be made up of ribbons from octant 'segments'
ie ribbons from the same octant (in phi) of all four barrels and
from the barrel centre to the barrel end (in z) which is a
'half stave' of 6 modules. That is 22 x 6 modules.
This bunching, followed through to the electronics crates,
may favour Tower mapping.
Diagram of barrel octant optoribbon bunching
Diagram of optoribbon bunch route to PPB2
12.07.99
Cable Clamp - new rectangular version
The new version of the cable clamp
shown in this drawing is a modified version of the original
Geneva CF clamp, but reversed, with a gap to allow optofibres through.
It also has an extra bar to hold the optoribbon in place. The clamp has
been lengthened so that it can be used for the system test setup, as this
uses connectors at the join between dogleg and low mass tape.
27.05.99
Optofibre ribbon routing trials
As a result of new requirements concerning safety and length of ribbon
it was decided to try routing one bunch of fibre ribbon per octant
(instead of following each low mass tape harness set). Fibre ribbons
are separated from the low mass tape of each harness at the end of the
barrel and looped round on a support close to the barrel (in this
case the 'old' model manifolds were used, but a purpose made (CF?)channel
would be needed). Fibre ribbons then follow the 'corridor' between tapes
outwards to PPB1 - seen in this photo
on the left half of the quadrant.The gap between tapes in the middel is
where INTERLINKS are situated. The gap to the right of this would be
used for the optofibre bunch from the right hand octant.
Here is the updated schematic depicting the
new arrangement of low mass tapes and opto ribbons
Low mass tapes have been moved into the 2 spare
locations at the centre to allow space either side for the opto ribbon
bunches - one octant bunch on each side.
Note that fibre ribbon can be routed
along a channel up to and beyond PPB1 - see another
photo. The channel can be enclosed - a section of trunking was used
- the ribbons from one octant fit into a channel 6mm deep x 16mm wide
in the same Z plane as the low mass tapes.A close
up view of fibre ribbon loops off barrel end. Note tapes would
normally be lower/flatter (closer to optoribbons).
Finally a photo
of the full length optoribbons
for one octant (4 barrels ie from 22 harnesses) if they were
to be routed through PPB1 without connections up to PPB2. The length
(4.5 to 5m) would probably be unmanagable at the assembly stage as
would require numerous reels.
27.05.99
New barrel 3 cylinder and sections of barrels 4, 5 and 6
These are ready to be prepared for mounting of inserts, brackets, model
harnesses and model modules which will be transferred from the 'old'
barrel when it returns from Oxford.
04.05.99
The End Barrel Quadrant has been updated
The model is now accurate according to the latest RAL drawing of the space
available for services and we intend to add cooling pipes etc when designs
are produced. Here is one view of the quadrant
- the bar across the tapes indicates the limit of the depth (10mm)
available for low mass tapes.
Another view of the quadrant - the card shows
the space allocations ie:
outer surface of flange at 755
Interlinks 755 - 765
gap for TRT cables and plate 765 - 775
SCT cables and heat spreader plate 775 - 785
SCT cooling pipe feeds 785 - 795
3mm gap and Thermal enclosure 795 - 803
Close-up view of model connector areas for
low mass tapes, with 1st set of 6 double tapes (one harness) then set of 4
and set of 2 (for another harness).
22.04.99
Discussions were held around the model about alignment unit
placement, optofibre routing, harness issues such as proposed dog leg
design change etc.
20.04.99
New dogleg designs were proposed at the harness meeting at RAL on 18.03.99
and as a result 2 different designs were modelled using Cu layers
to test if it would be practical to install a harness without causing
damage (assuming brackets and cooling pipes are in place)
29.01.99 and 22.01.99
Model of 1/4 section of ends of barrels 3, 4, 5 and 6 to PPB1
The model represents the ends of each of the 4 barrels with power tapes
coming from each barrel surface going to PPB1 using the Geneva
drawing.
The wall of the cryostat is represented by the blue surround
and the notches through which the tapes pass represent the first
set of cable connectors in PPB1 (using drawing of PPB1 by Andy Nichols
of RAL).The large plywood area represents the plane of the TRT barrel.
Here is one view of the model and
a side view.
A rough idea of how tapes separate out to the sets
of connectors. They will not take up as much space as this, as will be
in line
- this will be modelled using real connectors soon.
The services gap. The surface of TRT1
is represented by a sheet of transparent plastic.
The tapes are colour coded: red - barrel 3, blue - barrel 4, green -
barrel 5 and yellow- barrel 6.
A schematic diagram of power tape routeing
from all four barrels to PPB1 illustrating how tapes from different
barrels have to be paired to run to the connector area.
Note that this means that power tapes and opto ribbons
cannot be connected up one barrel at a time!
Tapes and ribbons
from all barrels will have to be paired according to a system before
connecting anything.
Two sets of power tapes ie two
'harnesses' of 6 double tapes go to each connector area (notches)
where there will be one set of 6 double connectors and behind these
one set of 4, then one set of 2 double connectors ie 12 in all.
The surplus tapes at each side belong to adjacent quadrants.There
are two 'spare' connector areas which we have sited in the centre
which may be cooling pipe area.
The model was used to measure tape lengths: it looks as if at present the length of the longest tape should be at least 1635mm. This is from the module closest to the centre of barrel 3 (from connection to dog-leg) to the PPB1 connectors situated towards the outside of the quadrant where the tape route is less direct. This is without the thermal enclosure, which may add to the length required. In addition, if the tapes are fitted with special temporary 'pre-installation testing connectors' which are then cropped off before 'real' connectors are fitted, the initial tape length will have to be about 40mm longer ie 1675mm. Watch this space!
First model of the 'octopod' design by A.Nichols and T.Hayler (RAL). This is the section of cooling pipes from each barrel manifold to the thermal enclosure. A view of the quadrant and a close-up of the eight pipes from the manifolds . Note that the thermal enclosure bulkhead has still to be modelled. Note also that the tape pairs are brought together approximately where the thermal enclosure will be situated.
Further work on this model: we intend to model the thermal enclosure area and the PPB1 area in more detail using new drawings from Andy Nichols (RAL) and also try out some cooling pipe configurations.
16.12.1998
New type of harness retainer
In order not to place stress on the fragile opto-fibres that come from
the optoboard on the dog-leg to join the fibre ribbon running with the
power tapes to the end of the barrel, a new version of the cable
harness retainer has been tested. Here is a view of the
6 new retainers and
another view.
These are made of plastic but should be CF. They are fixed under the
inserts (as were the rectangular retainers) but create less stress as
are fixed at 3 points only. If stress on the inserts is still a worry,
the 3 fixing points could be attached to the barrel itself ie moved
about 5mm to the side.
These retainers do not come near any connector/solder join areas and the
tapes in between the retainers remain flat against the barrel.
19.11.1998
Definition of space required
under cooling manifold
Height of gap between barrel surface and manifold - 10mm (minimum 8mm)
Length of gap (centred on mid way between each row of inserts) - 50mm
(minimum 45mm)
This means any fixtures connecting manifold to barrel need to be in line
with the rows of inserts on the barrel.
The position of the manifold relative to the barrel end (ie in z direction)
will not make any difference to this requirement.
The requirement is the same for both ends of the barrel
06.11.1998
New plastic harness installation 'envelope':
Here is a view of a newly installed harness, enclosed by the
transparent plastic installation envelope and a
close-up view.
02.09.1998
Barrel supports:
Here are two views of the steel supports,
one view
from the side and another view from the front
. Note that these supports now stand on a 250mm high plywood version
of the Oxford steel cooling platform on which the barrel will stand
during assembly of harnesses and modules.
16.10.1998
Barrel support platform
(table)- ergonomic trials:
Here is a picture of the model barrel standing
on the main section of the platform and a view showing the distance
a person 6ft 3ins (1.9 meters) tall with arms 36 ins (91cm) long
would have to reach standing and
sitting in order to touch things on the
(smallest)barrel without stepping onto the cooling platform.
11.08.1998
Harness Installation:
This would be reasonably straightforward if done before cooling pipes,
manifold and brackets were fitted, requiring only a rigid support
with separable dog-leg sections which would slide away as harness was
fastened using permanent cf cable clips (fixed under inserts).
However - model cooling pipes and manifold sections were made and fixed to
the model barrel in order to test harness installation methods assuming
barrel arrives already equipped with cooling hardware.
As the harness is made up of some very fragile components, notably the
optohybrid and fibres, it seemed best to enclose the whole section that
would go on the barrel in an 'envelope' which could be taken apart in
sections, attached to a rigid 'spine'
1) Harness (on the bench) ready to go in
envelope - this was made of smooth card (could be fairly thick TYVEK
or plastic). Note the strip of balsa wood slides in to form the rigid
'spine'
2) Harness in envelope before sealing
- note that the harness shown here has substantial model connectors
joining each dog leg to the power tapes - if flatter connectors or
a method of bonding or soldering is used this will enable the envelope
to be much flatter
3) Same - from above
4) Dog Leg envelope sections before sealing
5) Dog Leg envelopes folded
6) Harness enclosed in envelope in place on barrel -
slid under the manifold, with no brackets on
barrel. Harness is supported temporarily by
inserts
12)Three harnesses mounted on barrel with brackets
- view from -Z
Observations:
As can be seen from the above report, the brackets were installed after
harness was in place; initial attempts to install the harness AFTER the
brackets
were on the barrel were foiled by the springs and there is not enough room
between one row of brackets and the next to slide the harness envelope
through from the side.
(28.9.98 Further trials proved it was possible to place
the harness, with brackets in place, by feeding the loose ends of
the off-barrel tapes under the manifold from the barrel side, then
sliding the harness support diagonally over the brackets into position,
but this has to be done with great care.)
When cable clips were used to fix the harness tapes in position they
capture the three bare optical fibres before they
join the ribbon.
The length of fibres between ribbon and optoboard needs protection
(sleeving?) along this length as damage could easily occur at this
point and also where the bottom edge of the module comes close to the
underside of the dog leg (near the third fixing point).
Cooling pipes were modelled using Al bar and these were
very flexible, bowing out in the middle until a full row of modules had been installed. Pipes may well be more rigid,
but some kind of bracket will probably be needed at least in the middle of
each row to keep the pipe at the correct distance and angle from the barrel
surface, but which allows for lateral movement (either temporary or
permanent made of cf)
4.08.1998
Testing a method of securing a harness where it comes off the barrel (after
installation of barrel, when tapes are routed to patch panel, harness
could be damaged if not secured). Note: the tapes go UNDER the manifold.
Tests with cf brackets, springs and model modules - two Dog Legs with brackets and springs, with
three model modules in position and with a full set of six modules - view from side and view from front
Observations:
It was difficult not to catch things
on the bracket springs, which easily came loose. Although they hold the
opto board against the cooling pipe, the Al optoboard/ dog leg can slide
out of postion (would this be better or worse with grease?). It is also possible that the cooling pipe could transfer some stress to the
brackets via the springs and L shaped Al bracket.
The conclusions from this test were:
Advantages:
NOTE: the above tests were all done with the 'original' cooling pipe
dimensions and a model manifold section which may or may not
be comparable with the final design. It is obvious however
that any cooling pipe used must be flat both sides, not only for
maximum cooling effect but also for stability.
7) Harness still in
envelope whilst brackets are installed under dog leg, so optohybrids and fibres are
protected.
8) Dog Leg envelopes
removed
9) Harness envelope
is then gradually removed in stages as temporary cable clamps are replaced
by cf cable clips (retainers)which are held in place by the inserts.
The cable clamps are retained on the section beyond the ends of the
barrel, keeping the 6 tapes together but allowing freedom of movement
between tapes
10) View from end of barrel
11) View from mid barrel
A drawing of the above procedure
13) view from +Z
14) Three harnesse with model modules - view from
-Z
15) view from +Z
Modification to the Dog-Leg opto board (design for 98 System Test)
A detail of the harness with opto boards with model cooling pipes and manifold before installation
of brackets.
The opto fibres come from the opto board on the reverse of the Dog Leg ie
facing the barrel suface. Detail showing reverse of
Dog Leg. Note that the RED strip models the '12 fibre'ribbon ie 2
fibres per module and the BLUE strip models the ribbon that carries 1
fibre per module.
End of barrel with
'manifold tape clip' and view of off-barrel
tapes as they would go to the patch panel. The 'manifold tape clip'
is made from a single strip of 0.5mm Al, bent to shape, with no screws
(would need rounded edges where they touch tapes - could possibly be made of cf). Outside end of barrel showing how harness tapes are held against manifold providing cooling of tapes (!).
As a result of this, an alternative to using bracket springs was tried out -
using a modified Al opto board with arms extended by 15mm on both sides:
view of modified Al opto board in front of cooling pipe
then in position with bracket with springs
removed held against the cooling pipe by two small springs (same
width and same material as bracket springs, or could be of Al). A model module was fixed in position between the
two small springs - this is the one on the left. the module on the
right is on a bracket with springs. The extended opto board was held
securely against the cooling pipe and did not move or bend, and the module
remained in full contact with the cooling pipe (the lower type bracket was
used for this test).
1) Less mass between barrel surface and module, as there are no springs,
screws or screw fixtures. (the clips used were approx 25mm long
compared with approx 35mm long bracket springs of same material)
2) no danger of damage to opto fibres and opto hybrid on reverse of Dog Leg
3) no stress on cf bracket hence no movement of module
4) Al optoboard can't be knocked out of position
5) extra cooling surface on opto board due to extended arms
6) could adjust to thicker or thinner type of cooling pipe
Disadvantages:
1) clips on cooling pipe must be light so as not to distort pipe
2) clips must not get in the way of module installation
A possible method of taking up the 'flying' lengths of power tapes
at the ends of the barrel based on the Andy Charalambous method
with roller dimensions from the Oxford drawings. As can be seen
the roller (wheel) will need higher rims to hold tapes in place.
The following sequence would take place with harness fixed
in position on the barrel as no adjustments could be made after
this.
1: Wheel A is brought down over set of
6 tapes (one harness)
2: Arm rotates about Wheel A spindle
3: Wheel B is moved upwards
4: over wheel A
5: and down towards 'barrel end'
6: so that Wheel B keeps 2 sets of 6 tapes in position underneath
7: and Wheel A confines 2 sets of 6 tapes
above it. ** There will have to be
a clamp either over or just beyond
Wheel A to secure the ends.
Drawing of the
above sequence See last item on
this drawing - maybe the tapes and
wheels should be enclosed to prevent
damage(?)
We aim to try out new/updated designs of mechanical components on the model barrel as they are made available to us in the form of fully dimensioned drawings or samples, so please contact us (e-mail address below) as soon as you have anything new. This web page will then be updated with new photos so that everyone can comment on them before the next meeting.