Gerber Cutting Time.
Yuri reported on the time spent at Lab 8 for the machining and
handling of the P3 panels. For this production, the Lab 8 technicians were
handed a set of preliminary travelers where they wrote down the start and
end time for many operations during the panels production. The goal was
to compare the actual milling and handling time with expectations. 7 Anode
and 6 Cathode panels were machined. For the anode panels, the average time
spent on the Gerber machine was 190 minutes/panel, with 105 minutes spent
on the strips+HV artwork front side milling, 21 minutes spent on the re-milling
of the front side, 54 minutes spent on the HV artwork for the back-side
and 10 minutes spent on the re-milling of the back-side art-work (the re-milling
is typically due to "skips" in the art-work;i.e. regions where the copper
is not cut down to the glass fiber material).
On average the "full working time" (machining+handling) was approximately
50% longer than the actual "Machining Working Time". The cathode panels
took, on average, 90 minutes/panel, with 45 minutes for the strips, 21
minutes for the re-milling and the remaining for the fiducial mark work.
After we decided to cut the grooves 7 mils deep, the number of skips on
the strip cutting was greatly reduced (only one skip in 7 anode panels).
The biggest problem is coming from the HV artwork, where the amount of
debris created by the milling machine is not removed properly by the vacuum
system, and get stuck under the milling head preventing a proper machining
of the panels. There was not enough information on the preliminary travelers
to deduce the "full working time" for the Cathode panels.
In summary, the milling time for a P3 chamber is 3*190 min+3*90 min
= 12h30min, for a total of ~25-30 hours (4-5 days FTE) for the panels
for the 2 P3 chambers we machined. The estimate of full working time
is much more uncertain. The total time charged back to the project is ~
9 days, so there is still some work to do in order to stream-line the production
process in Lab 8, although the disagreement is not yet worrisome for a
prototype chamber. Yuri will suggest the information he needs to be included
in the travelers in order to get better estimates.
Issues on the PNPI Vertical Tension
Measurement Machine
Adam presented some issues on the setup of the Vertical Tension
Measurement Device.
1) The sensor head does not work for the P3 chamber. The sensor head
that Adam received was designed for a 10 degree chamber, and the wire spacing
in a 20 degree chamber is different. Classical Case of Miscommunication
! The present sensor head will use only ~10 of the 64 sensors.
A new head will have to be produced in time for the shipment of
all the tooling to PNPI.
2) 2 PC boards are needed to get the system operational. Adam will
prepare the PO.
3) The bolts for the panel brackets are interfering with the measuring
heads. Evgeni will design new bolts.
(MTTF) Mean Time to Failure Analysis
of ME23/2 Chambers
Giorgio presented an analysis of a Mean Time To Failure for the electrical
components hardwired in the chamber manufacturing. The analysis followed
the CERN Yellow Book 74-16: Introduction to Reliability Theory", by B.Schorr.
In our case, Giorgio relied on vendor data which provided information for
the failure rate (expressed as FITS, or failure probability for 10^9 hours
of operation, or % failures in 1000 hours. 1 FIT = 0.0001%/1000 hours)
for almost all the components hardwired in the chamber. Since our system
has no redundancy, the failure rate of a given system is just the sum of
failure rate of individual components. We received the following information
for the various failure rates:
Failure rate HV Sector = (14 x Failure Rate Carbon Resistor + 14 x Failure Rate Capacitors) = 0.02688%/1000 hours (pessimistic for Carbon Resistor) or 0.0003 %/1000 hours (optimistic for Carbon resistor).
Since we have 144 ME23/2 chambers x 6 layers/chamber x 5 HV sectors/layer
= 4320 HV sectors, and in 10 LHC years there are 90000 hours, at the end
of 10 years we would expect to see 4320 x 0.02688% x 90=104 failing HV
sectors (pessimistic) or 4320 x 0.0003% x 90 = 1 failing HV sector (optimistic).
Clearly the ignorance on the failure rate for the carbon resistors introduces
a big effect for the HV sectors. The failure rate appears to be bounded
by 3.4% (pessimistic) and 0.02% (optimistic).
Failure Rate Anode Board = (18xFailure Rate Carbon Resistor + 16 x Failure Rate surface Mounted Resistor + 14xFailure Rate Switching Diode + 4 x Failure Rate Surface Mounted Caps) = 0.077%/1000 hours (pessimistic for Carbon Resistors) or 0.043%/1000 hours (optimistic for Carbon resistors).
Since we have 144 ME23/2 chambers x 24 boards/chamber= 3456 boards,
and 90000 hours in 10 LHC years, the failure rate is
3456x0.077%x90=240 failures (pessimistic) or 3456x0.043%x90 = 134 failures
(optimistic). In this case the estimate are much closer because the relative
weight of our ignorance on the Carbon resistors failure rate is lower.
Failure Rate Anode Channel = (2xCarbon Resistor Failure Rate + 2xHV Caps Failure Rate + 2x Carbon resistor Failure Rate + 2x Surface mounted resistor Failure Rate + 2x Switching Diode Failure Rate) = 0.012%/1000 hours (pessimistic for Carbon Resistor) or 0.005%/1000 hours (optimistic for Carbon Resistor).
Since we have 144 ME23/2 chambers x 6 layers/chamber x 64 Anode channels/layer=55296
Anode channels, and 90000 in 10 LHC years, the total number of anode channel
failures is 55296x0.012%x90=597 failures (pessimistic) or 55296x0.005%x90=
250 failures (optimistic).
Status of P2' and P3
Oleg presented the status of P2' and P3 in MP9. P2' has been
assembled and framed by Nov. 5. The top panels bows by approximately 30
mils (good improvement from the 70-80 mils of the dry stack experiment,
before changing the design of the buttons, but still not perfect).
There is a plan to reassemble the chamber (changing the procedure for tightening
the bolts or changing the panel order) to see if anything changes, although,
engineering wise, nothing is expected to change and the present bowing
is probably due to small thickness variations summing up in the 7 layers.
P2' is waiting for the assembly of the chamber cart to proceed with the
gas sealing and gas leak test. At that point it will be completed and will
be moved to Lab 7 for the HV training and Cosmic Test. Before the
transfer, the alignment marks will be measured and the inter strip, strip-to-ground
and wire group-to ground capacitance will be measured (Last hour news:
The hardening steel for the cosmic test is installed, Yuri will coordinate
the assembly of the cosmic stand starting next week. The goal is to see
cosmic muons in P2' before the end of the year).
7 panels for P3 have been cleaned and are ready to be assembled (glue
gap bars and anode bars). The gluing machine is ready, Jerry will participate
in the gluing starting on Monday.
Two Week Schedule
All the items in the previous 2 week schedule have been accomplished
except the assembly of the sealing chamber cart, for which we were waiting
parts that arrived this week. This is the schedule for the next two weeks:
P2’
==
Chamber Cart Assembly
G.Smith, E. Borissov
Chamber Sealing
O.Prokofiev, B.Jensen
Chamber Leak Meas.
O.Prokofiev, B.Jensen
Vertical Mounting Test
G.Smith
Lab 7
====
Steel Installation
Y.Pishalinikov
Cosmic Stand Assembly
Y.Pishalnikov, Pam, B.Koecher
Gas System Resurr.
D.Northacker
P3 - MP9 Tooling
=============
Vertical Tension Machine
Purdue
Gluing Machine Setup
J.Zimmermann
Glue Supply
G.Smith, B.Koecher
Panel Cleaning P3
D. Northacker.
Parts Procurement
G.Apollinari, A.Korytov
P3 - MP9 Assembly
==============
Glue Cathode Gap Bars
O.Prokofiev, PNPI, J.Wittenkeller
Glue Anode Bars
O.Prokofiev, PNPI, J.Wittenkeller
Buttoms Procurement
V.Razmyslovich, R.Evans
Dry Assembly Stack-up
O.Prokofiev, PNPI