[Halld-tracking-hw] Carbn Fiber Parts

Fri Jul 16 06:35:13 EDT 2010

Hi Everyone -

    we have carefully gone through the plastic donuts and feedthrus. We 
find that the
ODs of both the donuts and the feedthrus are slightly under their 
specifications:
donuts 0.6 to 0.7 mil too small and the feedthru about 0.3 mil too small 
(on average).
We are extremely uncomfortable with the run-out that this will lead to 
when we assemble
the chamber. We believe that this can adversely affect the the location 
of the wires at a
level larger than 5 mils. We are also concerned that the run-out may 
exceed the spacing
that we allowed for between tubes, making installation of the last 
element in a layer difficult.
Finally, the run-out makes us prone to gas leaks at the input end of the 
tubes which worries
us in terms of guaranteeing good gas flow in the chamber.

     I know that this introduces a delay, but we are very uncomfortable 
with the parts in hand.
We feel that the following changes need to be made to reduce our 
perceived risk with
the chamber assembly.

1) Increase the OD of the donut as it now stands by  1.0 to 1.5 mils.

2) Increase the OD of the feedthru as it now stands by 1.0 to 1.5 mils.

3) Increase the size of the pin-hole in the pin holder so the crimp pin fits
     into the hole.

    Curtis

-----Note on measurements-----
All measurements at CMU were performed with a brand new Mitutyo 293-344
digital micrometer that reads to five decimal places and is guaranteed 
accurate
to 0.00005 inches (1 micron).  The has a ratchet-friction thimble so 
that you cannot
squeeze the component too tightly.

The ID of the straws which we received is difficult to accurately 
measure. We have made
a part that slips in when we are cutting the straws that is 0.6106. From 
this, and the nature
of the fit, we estimate that the ID of the tubes are about 0.611.

We also note that because all the holes in the carbon fiber endplate 
were reamed using a
0.4375 tungston-carbide reamer, then checked with pin gauges, we believe 
that all holes
are 0.4375, which is the lower end of the specification.

Each of the parts that we received were measured many times in diffrent 
configurations.
Those numbers are reported here, the largest first, follwed by the 
smallest. The largest
number was measured with the micrometer on the small ridge.

Donut  OD:   Design Specifications is 0.609 to 0.611
               --   the average of all of our measurements is 0.60784

   0.60880   0.60730
   0.60880   0.60710
   0.60875   0.60720
   0.60870   0.60700
   0.60875   0.60715
   0.60860   0.60690
   0.60850   0.60680
   0.60850   0.60695
   0.60860   0.60680
   0.60850   0.60700

Donuts OD when mounted on the  0.4375" pin-go-gauge. The two measurements
were taken 90 degrees apart near the glue trough.

   0.60880, 0.60835
   0.60870, 0.60815
   0.60870, 0.60845
   0.60875, 0.60810
   0.60890, 0.60860
   0.60880, 0.60830
   0.60860, 0.60830

Donut ID:  Design Specification is 0.4375 to 0.4395

    All were checked with our 0.4375 go-gauge, 0.4396 no-go-gauge and 
passed.
We noted that a 0.4380 pin-gauge would go through the holes.

Feedthru OD  Design Specifications are 0.4355 to 0.4375
                 -- the average of all our measurements is 0.43491

    0.43595  0.42400
    0.43570  0.43410
    0.43570  0.43400
    0.43570  0.43435
    0.43570  0.43410
    0.43565  0.43395
    0.43550  0.43450
    0.43560  0.43400
    0.43555  0.43430
    0.43565  0.43410

Feedthru OD when mounted on the 0.25000 pin-go-gauge. Measurements were 
taken
90 degrees apart near the glue trough

   0.43575, 0.43525
   0.43570, 0.43525
   0.43560, 0.43515
   0.43580, 0.43520
   0.43565, 0.43525
   0.43570, 0.43520
   0.43565, 0.43525

In all of the above cases, the typical smaller measurement will lead to 
the biggest slop
in the fits. Also, if we take the measurements when mounted on the pin 
gauges to be
the reference, then the donut OD is about 0.6-0.7mil under specification 
and the feed-thru
is about 0.3 mil under specification. Essentially, with all the OD's 
undersized, and the IDs
in the middle of the specifications, we have a build-up of loose tolerances.

1) Strawtube to donut is about 2.5 to 3.0 mil. This affects our ability 
to guarantee a gas-tight
     seal at the down stream end of the chamber as the gas enters the 
straws.

2) The feedthru to the donut is about 3.0mil. This also affects our 
ability to guarantee a
     gas-tight fit at the down stream end of the chamber.

     The two comments on gas above worry me. If we have leaks near the 
input to the straws,
     will we be able to guarantee good gas flow through the straws? A 
may be paranoid, but
     our performance is very sensitive to gas (as we have seen from our 
recent work with
     the preamplifiers).

3) The feedthru to the endplate is about 2.5mil. This most directly 
impacts the centering of the
     wire in the hole. This needs to be combined with the tolerances of 
the pin-holder.

4) The OD of the pinholder to the ID of the Feedthru is snug because of 
two small ridges
     near the top of the pin holder. The fit at the base of the 
pin-holder is very loose. I am
     not certain if this is an issue or not, but mention it for 
completeness.

The above 1-3 also impact where the straw will be located relative to 
the center of the
hole in the end plate. We have a potential build up of 
(2.5-3.0)+(3.0)+(2.5). The concern here
is that this will be a systematic that pushes the straws apart, so it 
could consistently
be above 5 mils. I do not believe that this will impact the 
electrostatics, but it could build
up going around a layer of straws, and impact our ability both our 
ability to insert the last
straw in a layer as well as place the "close-packed" layers in place. I 
believe that our design
allowed for a 5 mil gap between straws.

-- 
Prof. Curtis A. Meyer		Department of Physics
Phone:	(412) 268-2745		Carnegie Mellon University
Fax:	(412) 681-0648		Pittsburgh PA 15213-3890
cmeyer at ernest.phys.cmu.edu	http://www.curtismeyer.com/