Hello Dan (and Jan),
                    That's good news in my opinion, since it means that
the present mechanical thickness gives a reasonable result (0.22mm
deflection).  I expected you wouldn't join today's meeting because of
Thanksgiving.
                    As I said at wpmtg117, we should try to understand the
tolerance to put on the z-deflection of the endplate.  On playing with the
numbers, I came to the conclusion that the magnitude of the deflection is
not as important as the error on the knowledge of this magnitude. If the
TPC clock is good to one nanosecond, which corresponds to about
80 microns for the T2K gas, then this seems to be a reasonable ballpark
for the knowledge of the deflection. Therefore a deflection of 0.22mm is
also quite reasonable. Of course if the deflection were very large, say
22mm or 22cm, then this argument might break down, but we don't have that
problem with your design.  I also looked into the deflection we had for
the Aleph TPC endplate, and Werner Wiedenmann's study concluded that we
had about one mm deflection, which he corrected the data for using
Z->mumu events from the Z-peak running (the Aleph TPC had 6mb
overpressure, as you know). I am still working on figuring out what
Werner's correction corresponds to for the error on the one-mm deflection
of the endplate, and also on looking for other arguments to understand if
this ca.100 micron tolerance on the knowledge of the deflection is o.k.
                    We can never reduce the deflection to zero, as your
mail below shows, Dan, but this would be adding material with no real
benefit.  So let's stick with 100mm endplate thickness and 0.22mm
deflection for this round, up to the DBD...

          Happy Thanksgiving,

              Ron

........................
             Ron Settles
             Max-Planck-Institut fuer Physik
             Foehringer Ring 6
     D-80805 Muenchen
             +4916090710273


On Wed, 24 Nov 2010, Daniel P Peterson wrote:

> Jan and Ron,
> I will not be at the meeting tomorrow but would like to mak this brief
report.
>
> At wpmtg117, we discussed the possibility of increasing the endplate
thickness beyond 100mm. I have run the FEA for thicknesses
30,50,75,100,125,150 and 175. The short stoy is that there is little
advantage to increasing the thickness beyond 100mm.
>
> In the range from 30 to 75mm thickness, the deflection varies closely
to a power (t**-1.5), slightly different from my expectation of (t**-2).
At 100mm, the deflection already deviated from following the power; the
deflection at 100mm is more than expected from (t**-1.5). The
improvement from 125mm to 150mm is quite small.
>
> The the root cause of the breakdown of the (t**-1.5) variation can be
seen in my presentation, wpmtg117, 20101111, page 4, bottom figure. The
elements of the back-plate do not remain straight in rows 1 and 8; they
are buckling between the spaceframe-separator-plates. This only get
worse at larger thickness. At 175mm thickness, although the FEA solution
looked to be following the trend, the FEA complained that the solution
might be garbage.
>
> The increase in thickness comes with an increase in material because
the  spaceframe-separator-plates are getting longer. Alternatively, the
extra material can go into the back-plate. Increasing the backplate
thickness from 3mm to 4mm provides a decrease in deflection similar to
increasing the thickness to 150mm. This is preliminary and not optimized.
>
> Dan
>