Quality Assurance

1
Quality Assurance
A status report A procedure for ensuring the
quality of the finished tracker has been designed
following our experience with the construction of
the prototype and advice from D0. The various
components of the system have been individually
validated, or the problems identified and we have
confirmed that suitable solutions exist. A plan
exists to provide a working system in time for
the delivery of the fibre doublets and the
construction of a fifth station. The
construction of the fifth station will be
undertaken at a speed that will allow the QA
procedures to be optimised.
2
Fibres arrive their properties having been
measured
Clear fibres run from several station connectors.
(internal clear fibres)
Fibres run from the doublet layer
20 or 22 bundles are bought into a station
connector
Patch Panel
are collected into bundles of seven
External clear fibres
Problem As soon as the fibres are glued into
the connectors no further corrections to
errors,omissions or faults is possible
Connector to the VLPC cassettes
3
Bundles may have too many or two few fibres
Fibres incorrectly positioned in connector
Misalignment between connectors
Clear fibres incorrectly positioned in connector
Or fibres may be crossed over
All fibres susceptible to cracking leading to
unacceptable light loss
External clear fibres
Patch Panel
Possible Faults
4
Method Inject light.and track its path
Silvering of the fibres prevents straightforward
injection into the rear of the plane
Solution illuminate the plane with a UV LED
Question of damage to the scintillating fibres
(considered later)
Objective Identify faults before it is too late
to correct them OR reject the component before
too much time has been spent on it.
5
Scanning table
Plane held on vacuum chuck
LED and collimation and focussing system
Scanning system
6
Bundle fibres up into sevens do about 10
bundles.
Hold the fibres together with a rubber sleeve
(already used to make the prototype)
7 fibres
Place the bundles in a comb to hold them
together and in place
Bundling
For these pictures 370nm was used throughout
7
Clearly observe the two layers Counting seven is
easy
Number determined by experiment to minimise the
total construction time. it takes n minutes to
make a bundle we check every G bundles which
takes a time of (Gm s) minutes. If the
probability of making a mistake is p (known
empirically to be lt1). Total time for a plane is
Gn (Gm s) p(G1)n/2T/G with T
bundles per plane.
Unbundle and repeat if required
Confirming bundles
This stage may be done by eye or automatically as
appropriate
8
Fibre bundles
Can measure lt 5 of the nominal output
Light leak
Cameras
Light source
comb
At this point we will also measure the light loss
from the silvered end of the fibres
Final scan for bundling
9
Jig for combs and station connector
10
Fibre bundles
Camera
Light source
connectors
Sideview
Scan after placing bundles in connectors
LED
Suitable bend radius
11
All fibres in their correct bundle
All bundles in their correct station connector
Silvering OK
Ideally we would also check light out put at this
stage but until the fibres are glued into the
connectors and the ends polished a reliable
measurement of the light output is not possible
Status
12
Once the station connectors have been cut and
polished. At this stage we measure light
output. We will also make this measurement on the
old stations which have been disassembled this
will allow us to validate the UV measurements
against the test beam results
Light Output
13
Circle radius of 12 pixels used, Integral
intensity in arbitrary units. Background black
is effectively 0

• Automatic Measurement
• An astronomical program confirmed
• we can identify individual fibres
• we can measure their light output

In this picture no attempt was made to control
the illumination
14
Scan film
15
Check wiring of clear fibre runs
Inject red light and measure output
fibre. Injection done in order by hand
photographs taken of the output and the two cross
correlated subsequently. The experimenter will
record the input connector/fibre number, the
programme will determine the correct input
connector/fibre number from the output fibre
number. The experimenter will not be able to
influence the comparison in order to combat the
see what you expect effect
16
Charged particle tests
The station will then be irradiated by a Co57
source of gammas to check response to charged
particles.
Orthogonal scanning table holding the source
Light tight box with two orthogonal computer
driven scanning tables
Scanning table holding the plane
These measurements will also made on station 4
and the test beam results used to validate the
measurements.
17
UV light damages the scintillation fibre need
to minimise damage and quantify the dangers.
Wavelength Currently using 370nm LEDs which give
good results. D0 have used longer
wavelengths MICE have ordered longer wavelength
LEDs and we will use the longest wavelength which
gives satisfactory results. D0 also pulse their
UV LEDs and we will implement this if it is
necessary
Test measure the damage and set a limit of 1/10
of a photo-electron for the maximum damage make
sure the total illumination for normal checking
is less than 1/5 of this dose
UV damage
18
Hitachi U4100 sample compartment
Measure effect of UV irradiation on the
fibres. Using high performance photometer
19
Time Line
20
100 correct fibre runs Light loss from all
sources bend, cracking, misalignment lt
10. Light yield measurements good to better than
10 System to be ready by end August for the
delivery of the first doublet layers. Steady
construction of a station 5 to validate the
system. By end-September the validated QA system
will be in place. Ready to start production. Once
the planes have been constructed they will be
geometrically characterised with a CMM.
Summary