Mar 24 Telecon discussion of the Tower to Grid Integration Requirements
The following is my version of the subject telecon minutes:
The telecon opened with Mike explaining the email from Robert to
Sandro. Sandro and Nanda had questions regarding the tolerances of the
Flight grid and the grid simulator and requested that the simulator be
made to a better tolerance so as to be able use it as a reference for
the Tower CMM measurements.
AIT was present at the Telecon and indicated their desire to
participate in the process of defining the interface design and
integration concept.
Sandro requested that the eccentric washers be made with a locking
mechanism such that during the integration process the determined
position is not lost. The concept of using a clamp was discussed. It
was not clear if that concept is the baseline. There was some
misunderstanding of the request by SLAC. Mike indicated that the
current procurement/qualification is going ahead without the locking
mechanism and that there is no man-power available for any parallel
effort such that the washers with the locking mechanism could also be
qualified. There was some discussion of the efforts that are
necessary from INFN in order to provide the necessary data to SLAC for
the implementation of the integration concept that is being considered.
Mike was able to explain the concept further such that there was a
thorough understanding. The details of how and what of this concept is
yet to be defined. INFN made a statement indicating the efforts that
they have foreseen in their scheme of things : they are expecting a
grid simulator with the plane of the hole pattern well defined with
respect to the outside corners of the simulator. The simulator will be
measured and verified. The Holes of the corner flexure will be measured
with respect to the perpendicularity of the tower in its integration
configuration. The washer positions will be calculated and fixed using
the tools provided by SLAC and the grid simulator will be integrated to
the tower. The Tower will be turned over and the position of the tower
will be measured with respect to a defined point on the grid simulator.
The Tower will undergo the environmental tests and prior to shipment to
SLAC the position of the tower will be remeasured with respect to the
reference position on the grid simulator.
There was no discussion of the requirement of an optical alignment
verification system at INFN. SLAC AIT stated that they will remeasure
the Tower in its delivered configuration and a verification of the data
from INFN will be performed prior to any action of the Tower. There
some discussion about the reasons to discard the flight washers that
are used by INFN for the integration of the Tower to the grid simulator
and using a new flight set for the final integration of the Tower to
the grid. SLAC AIT has baselined this as their standard procedure and
will be developing all the necessary actions to implement it.
Mike Menning’s addition:
I have some clarifications which I will try to state below:
First of all, the only cones for which rotational position is important are those at the three reference corner positions for each tracker.
Back in February at Pisa we identified these positions as the +X+Y, the -X-Y, and the +X-Y. As Sandro has correctly pointed out, there are many other choices but we arbitrarily established these as our baseline.
The rotations of the cones at the three reference corners must be the same at Pisa and at SLAC whenever alignment information is being gathered or verified. It would be best to keep these cones in the same position at all times.
The rotational position of the cones at the three reference positions will be determined by analysis performed on the CMM data collected on a tower in Pisa without a grid simulator attached.
The position of the cones at the three reference positions will be adjusted by rotating the cones within the flexures until index marks on the cones line up with the desired angular positions on an installation template indexed to the flexure.
The cones at the three reference positions can be removed or repaced at any time but they must be reinstalled to the proper angular position according to the installation template whenever alignment verification is required.
The rotational positions of the eccentric cones at the 9 remaining locations are not important as long as the eccentric cones have enough range or motion to position the center of the hole in the inner cone over the center of the bushing in the grid or grid simulator.
The rotational positions of the eccentric cones at the 9 remaining positions will be different when mounted on a grid simulator than when mounted to the flight grid.
The clamping device discussed at the teleconference is used to apply a seating force to the eccentric cones in the flexures when a tracker is being installed on the grid or grid simulator. The eccentric cones need to be properly seated in order to determine the proper shim thickness ( with a feeler guage ) between the flexure and the grid or grid simulator. If the nut on the stud were torqued down to seat the cones without a proper shim in place the bottom tray could be distored and damaged.
The clamping device can be used on other occasions to retain the eccentric cones in the flexures. In this case the clamp will inhibit rotation of the cones but will not " lock " them in place. Rotation is inhibited due to the force from the clamp which presses the cones into the tapered hole in the flexure. If ever the clamps allow the cones to rotate they could be easily returned to their proper angular position.
The initial CMM measurement on the tower alone ( without grid simulator
) allows the tower shape to be referenced to the plane defined by the center of the studs at the flexures at the three reference corner locations. By doing this we bypass any errors in the grid simulators and we can minimize the tracker shape errors.
Sandro's desire and suggestion to measure the alignment of the tower relative to the grid simulator as a method for verifying that no subsequent change in alignment has occurred is a good one. Otherwise we would have to take the tower off the grid simulator each time to check alignment. Alignments measured relative to the grid simulator will be influenced by the shape of the grid simulator but the alignment measurements in this configuration should be repeatable and can be verified without removal of the grid simulator.
Nanda...your comment that INFN "is expecting a grid simulator with the plane of the hole pattern well defined with respect to the outside corners of the simulator" needs to be adjusted a bit since only three of the mounting holes will define a plane. We would probably pick three holes as a reference ( the +X+Y, -X-Y, and the +X-Y ) and reference that plane plus some information to locate the axes in the plane to three tooling balls mounted to the external surface of the grid simulator. We need to get our alignment specialists involved to work out details at this level.
Robert Johnson’s
addition:
1. There is no requirement for an optical alignment system at INFN. All
that we want to require are the CMM measurements discussed. SLAC I&T wants
an optical alignment system at SLAC, but that goes beyond Tracker
responsibility (they have to tell us what they require in terms of
references on the top tray).
2. SLAC would prefer NOT to remeasure the tower upon delivery to SLAC.
Instead Mike has proposed to send the SLAC alignment engineer to INFN to
participate in the measurements there. We don't believe that the shipping
itself has much risk of changing the alignment, compared with the vibe test
(unless it is severly mishandled, in which case we will see it on the
accelerometers of the shipping data logger).
3. My understanding is that the retaining clamp IS baselined. Regarding the
locking feature, I felt like there may have been a misunderstanding of
terminology during the meeting. To me a locking feature would be something
built into the cones that lock them after installation such that they cannot
rotate (during vibration, for example). However, I thought I heard Sandro
say that this was not really the issue that he was concerned about. I could
be wrong, but it seemed to me that his concern was to keep them in place
during installation, up until time that the nut is tightened. In that case,
why doesn't the retaining clamp specified by Mike and Martin do the job?