From: Arthur W. Brooks
Sent: Thursday, March 08, 2007 4:20 PM
To: Wayne T. Reiersen
Cc: Thomas G. Brown
Subject: Shim Thickness Variation

Wayne,

In response to your questions affecting the shim thicknesses:

1) What is the range of shim thicknesses we should provide (4"x4")?

2) How out of parallel might the flanges be positioned? ie How much variation between the max/min thickness on a shim?

3) Due to local effects (flatness) how much additional variation is expected?


The shim thickness will be governed by the mating flanges surface variations from ideal or more accurately, the variation of one flange relative to its mate. This variation will be due to MTM machining tolerance and also to any realignment of flange surface we do to try and correct coil current center. 

We have data from MTM on the flange face variation which ranges from min of -0.026" to a max of +0.033" with a max range of 0.046" for any given flange face. Plots of the data for A1 and A2 show fairly smooth poloidal variation in the data. However, the poloidal gradient is still significant. Calculations done on all the data for the coils in the first period show variations as large as 0.010" over a 4" shim size, though many are less than that. Below is a table of the largest gradient over 4" in each flange:

A1D	0.0091
A1E	0.0043
A2D	0.0056
A2E	0.0027
B1D	0.0025
B1E	0.0048
B2D	0.0043
B2E	0.0073
C1D	0.0092
C1E	0.0089
C2D	0.0100
C2E	0.0066

This data is measured relative to MTM's Tooling Ball Locations. The recent discovery of a mismatch in PPL measurements of the Castings relative to MTM's and the subsequent decision to forge ahead based on PPL's measurements throws some uncertainty into the flange data. Our plan is to remeasure the flanges based on our Conical Seat Locations for coils where they are still attached. 

The realignment of the coils based on aligning 6 coils together (ie period 1 which has all coils wound) has a fairly small impact on shim thickness contributing a max variation of +/- 0.006". (Initial attempts to align individual coils or even A1 thru A3 together required larger total coil motions and was deemed not desirable.) For the first period it actually serves to slightly reduce the flange variation (the -0.026" to +0.033" range drops to -0.022" to +0.028"). I cannot say yet if the next two periods will be so lucky.

Art