Agenda
Oleg                            Status of P2', P3 and Aging Test Chamber
Giorgio                        Summary of Panel Measurements
Nelson                        Use and Necessity of DR (Discrepancy Reports)
 

Status of P2', P3 and Aging Test Chamber.
Oleg reported on the status of the various chambers.

• P2' is under High Voltage and taking data. All layers can take 4.3 kV. Above 4.3 kV, one layer shows signs of electrostatic instability. In that particular layer we know the wires are not in the center of the gas gap due to some panel bowing that was overlooked during production time and caused the wires to be glued in the wrong position (i.e. not in contact with the pads of the wire fixation bar, but rather approximately 1 mm over them). This problem will be fixed in the all the next chambers production. On a different HV section (out of 30), some instabilities arises above 4.3 kV, whose source is unknown.
• P3 has 2 panels ready. The last panel is being assembled and will be ready by January 11. The chamber will be assembled by January 13-15. The frame will be mounted after that.
• One Anode panel for the aging test chamber is ready. The second panel will be done during next week. The Anode panels should be completed by Jan. 15. The cathode panels still need some working for the source holes. A TD Technician is able to achieve the limited thickness required for a proper functioning of the Fe55 source (6 mils) and should be able to drill the panels before the end of the month. It was mentioned that if, during the drilling, the copper is broken through, we will repair it with a piece of mylar.

Two kinds of measurements were performed.

• "string-bowing" measurement, where one string is strung between the extremities of the panels and the sagitta between the panel and the string is measured. This is a quick-and-dirty check on the panel bowing. For a simple bowing, this measurement, performed over a dimension of the panel (typically 5 or 12 feet) can be used to interpolate the sagitta that we would observe over the "unsupported" distance in our chambers (typically 2 feet). This extrapolation is applicable only for "simple bowing". For compounded bowing (i.e.. a panel that goes up and down more than once across a given dimension) the interpolation does not work.
• Laser measurements, were a laser beam defines a plane, and a probe is placed on the panel (which, in turn is place on a "given surface") to measured the variation of the panel surface respect to the ideal plane defined by the laser. As a given surface we used a granite table (i.e. no "forcing" of the panel against a flat surface, to check the overall panels bowing and properties) and the Gerber and Axxiom and Vacuum tables (i.e. forcing of the panels against a given Datum). We started using the Gerber table whose vacuum didn't prove strong enough to hold the panel down, and we moved them to the Axxiom vacuum table.

• These measurements were performed both by Plascore and Fermilab. The agreement is rather good between our measurements and Plascore measurements. Typically we observe a sagitta of 30-40 mils across the 5' dimension (width) , 70-100 mils across the 12' dimension (length) and 200-300 mils diagonally. Unfortunately the setting of the panel in an un-stressed vertical position is critical and rather difficult. Gravity can easily influence the results. Based on these numbers, almost 50% of the panels would be rejected because they do not satisfy our requirements of less than 7 mils sagitta over 2 feet.

• The first laser measurements, performed on a flat granite table in November , were done to measure the intrinsic characteristics of our panels. The panels vary, in height, by 80 mils (2 mm) and present a characteristic "double-ridge" bowing due to the presses used at Plascore as shown in the "surface plot" on the following figure.   The "double-ridge" problem was present in all the 5 panels measured on the granite table, and the problem is very consistent on the two sides of a given panel (i.e. what is measured to be a "double-ridge" on the top surface, turns out to be a "double-valley" on the bottom surface of a panel). This fact was notified to Plascore and in the December production the lowered the pressure on their platens to check if the problem was due to excessive pressure.
• The second laser measurement were performed on the panels procured in December '98. The first set of measurement was performed on the Gerber vacuum table. First we measured the table, and then the panels over the table to be able to normalize the measurements. The reproducibility of the measurements has a spread of 1.9 mils over distances of 12 feet.    It turned out that the vacuum on the Gerber machine was not strong enough to hold the panel completely flat against the surface (bad) but these measurements proved to be useful (good) to see that the "double-ridge" problem was still there, suggesting that the original problem is in the flatness of the plates used at Plascore rather than  the pressure with which the panels are assembled.
• Finally we moved to the Axxiom table where the vacuum was good enough to keep the panels flat. We had no way to check "how flat" the panel was. However  the "flatness" induced by the atmospheric pressure is definitely better than what the frame and buttons will do in the final chamber, so we used these measurements as a "limit case".   Under pressure, the height of the 2 panels we measured varies by 35 mils (minimum to maximum). No "double-ridge" feature is present (i.e. the Axxiom vacuum is enough to push the panel flat against the datum). 70-80% of the measured points (out of a total of 275 points) lay within an envelope of 10 mils (+/- 5 mils).  The out-layers are not distributed according a specific pattern.  Both the top and bottom surfaces of a given panels show the same features.

• It appears like the "double-ridge" problem has to be solved by flattening the plates used at Plascore to manufacture the panels. We are presently in contact with them to assure this will be done.
• We will order another set of 14 panels to check any improvement, before going to the full production.
• It looks like the manufacturing process has an intrinsic "non-flatness" with a sigma of 5-6 mils, and therefore we may have to start thinking in terms of specifying a distribution for the flatness and allow few out layers.

• When is it Created: whenever a condition occurs that does not conform with the established specifications, engineering drawings, manufacturing practices or otherwise the normally expected conditions or results.
• Why is it Created: this is probably the most important reason, since a DR is needed to identify a condition and communicate it to all those that "have a need to know" that there is a deviant condition and what is being done about it. As an example, Nelson presented a DR produced during the assembly of the P3 panels, where one of the guard strips was either glued in the wrong position or moved during the 24 hr curing period.  The DR is now an historical document which can be easily found and provides the essential information about the discrepancy  and how it was dealt with (for this particular example, we can change the manufacturing procedure or engineer a fool-proof way of locating the guard strips).
• Who fills the Paperwork: The first hand observer; typically a technician or floor supervisor.  The responsible authority (either N.Chester or G.Apollinari in this project) will make the decision about the disposition of the discrepancy and implement the fix to prevent the problem recurrence.