From batesr@physics.gla.ac.uk Wed Sep 22 13:29:01 2004
Date: Thu, 2 Sep 2004 14:53:39 +0100 (BST)
From: Richard Bates <batesr@physics.gla.ac.uk>
To: Anatoli Holodenko <holodenko@mail.ihep.ru>
Cc: Inna Grigorieva <Inna.Grigorieva@g23.relcom.ru>
Subject: TPG


hello Anatoli and Inna,

I need your help to understand the number of TPG plates that we have
ordered, the yeild of turning these into spines and therefore the number
of spines that we will be able to make.

What I need to know is if we will be able to make enough spines to make
all the modules that we require.

I have looked at the Contract we have with Atomgraph and it says that they
will supply 1900 long plates and 600 short [total = 2500].

this compares with the requirement of 1892 long modules (sum of middle and
outers) and 480 short (inner) modules [total = 2372]. The number of good
spines has to be equal to this number. Note that we loose some spines and
some are trashed at module production sites. At present we have 92 spines
that have been trashed. So the number of spines produced must be more than
this number. At present is would appear that we should make about 10% more
spines than modules required to make sure we have enough spines.

So lets assume we need 480 inner spines and 1892 long spines. We can
therefore have a yeild of 80% making inner spines from the raw TPG and
still get enough inner spines. BUT we heve to turn all the long TPG plates
into outer/middle spines. This seems to be a very silly thing as we know
that the yeild at NIITap is not 100%. I got some numbers from Larisa in
August from which I understand their yeild to be 66% for long spines and
83% for short spines. this is just the yeild inside NIITap. Please see
numbers at the bottom of the page.

Also we have ordered a total of 2250 spines (not 2372) which requires an
overall yeild going from raw TPG to spines of 90%. BUT the ratio of long
to short spines is wrong. We have ordered 1901 long spines (middle +
outer) and only 349 short (inner) spines. Therefore we need to make one
more long spine that we have order long TPG plates and we can have a yeild
on short TPG plates to inner spines of 58%. This is for the whole process
until the spine is received at the module production site.

So there is a disagreement between the amount of TPG we have ordered and
the number of spines that we require. AND there is a disagreement between
the number of spines that we have ordered and the number of modules that
we wish to build. I can understand a little bit the problem with the
number of modules as the required module number might have been increased
after the orders to take into account bad modules during production.


Have I made a mistake regarding this. Did the number in the order refer to
TPG plates after processing at NIITap? I dont think this is the case.

Please advise.

If we are only ordering 1900 long TPG plates from NIIGraphite then we need
to order quite alot more TPG to be able to finish the production of
spines. Even to produce the number of spines that IHEP said that they
woudl produce let alone the number of modules that we wish to make.

I am on holiday for 1 week and 1 day starting on friday. So this gives you
a week to understand the numbers before I am back. I can then advise to
the SCT management if we really will have such a short fall in TPG and
place another order for extra TPG plates. We would place an order for both
long and short plates but less short than long as we do not have such a
short fall in short TPG plates.


The numbers that I received from NIITap are:

Translation:
---------------------------------------------
1. Received from Inna:
   - long samples 1744 items;
   - short samples 538.
2. #good TPG prepared:
   - long 642;
   - short 243.
3. # not passed:
   - long 139;
   - short  25.
4. Faults during the production:
   - in polishing:
      long - 17 + 43 (used for production of Iner)=60;
      short - 12.
   - in laser cutting: long - 13; short - 4.
   - in pressing - long - 13, short - 4.
   - in process of making them thinner (?) *): long-13, shirt - 5
   - in covering by Parilen - long - 56, short -2
5. Total number of faults:
   - long - 324;
   - short - 48.
6. Samples in production line:
   - long - 298;
   - short - 32.
7. Remain unused:
   - long - 479;
   - short - 258.
-------------------------------------

therefore we have a yeild for long plates of 642 / (642 + 324) = 66%
and for inners = 243 / (243 + 48) = 83%

-- 
-------------------------------------------------------------------------------
Dr. Richard Bates
Physics and Astronomy Department  Tele : +44 141 330 5882
Kelvin Building                   Fax  : +44 141 330 5881
The University of Glasgow         Email: r.bates@physics.gla.ac.uk
Glasgow                           WWW: http://ppewww.ph.gla.ac.uk/~batesr
G12 8QQ
-------------------------------------------------------------------------------