[BTeam] suggested gradient distribution

[Jay Benesch]

In column I of the attached I suggest a gradient distribution for the 
500 MeV/linac run.  The fudge factor in the SL is 1.043 so I run the 
modules there a bit higher.  The first six columns were extracted from 
the CED this morning.  Tripoffset is the gradient for 8 hour fault 
interval.  Tripslope is the slope in the fit
log(1/interval)= intercept + slope*gset
log(1/28800s)=-10.268 = intercept  +TripSlope*TripOffset

Using Arne's MC I calculate about 3800W total including static heat 
loads for modules turned off, FEL and its transfer line.  I have the C50 
modules off because the probe Qs are higher and the external attenuators 
were never adjusted to compensate; it follows that the C50s won't 
regulate well at 5-6 MV/m.  Also, they have lower Qs.

----------------

stats on probe Qs in RFCM download files; there's a factor of 10^5 
missing from real value.

mean, stdev for C25: 1.54E7, 0.17E7  (spec 1.5E12 +-20%)
mean, stdev for C50: 3.75E7, 0.68E7

Both distributions are consistent with normality.

--------------------

Jay


-------------- next part --------------
A non-text attachment was scrubbed...
Name: 7apr2015_ced.ods
Type: application/vnd.oasis.opendocument.spreadsheet
Size: 31940 bytes
Desc: not available
URL: <https://mailman.jlab.org/pipermail/bteam/attachments/20150407/64c64262/attachment-0002.ods>