Minutes of the geometry meeting 18/3 2005

_______________________________________________________________________

Present:

J. Beauquis, J.-P. Koutchouk, F. Savary, R. Chamizo, N. Emelianenko,

W. Scandale, M. LaChina, G. Berard, Z. Gao, E.Wildner

Activities in geometry:

The first point on the agenda was to go through the list of activities related to geometry.

The aim of this was to give a global picture of the work that is being done in the domain of geometry, as far as it is known to me. See slides “activities1”.

Comments on this activity list was that the holding point for itp15/20 must be applied from now on and a meeting with project engineers will be held asap to formalize the information flow via MAS/MA in case of non conformities (out of tolerance). We have to evaluate the extension of this holding point in terms of manpower available for tracing in data base etc. (compare the magnetic holding point system), on top of the effort to do the holding point itself.

A holding point at WP08 is ongoing since the new fidualization baseline was put into operation. Elena gets the non conformity and gives new adjustment values. Zhengwei also checks all magnets after fiducialization and prints documentation. To be formalized (Elena, Davide and Jean-Pierre).

The report for itp20 November/December should be finished by Friday 18/3. Next report, January/February can be already generated including modifications made by Natalia. For the report Mars/April we should also include the positioning of cold feet pads.

The holding point software and the automatic report writer (presentation by Natalia Emelianenko, see slides “next_steps_PDA_tr1.ppt”):

Natalia presented new features in the automatic report creator for itp20: Moving average (present period and actual period on the same graph, with average and standard deviation calculated separately for the two periods). She has also included the possibility to have a statistics summary automatically generated and included in the report. The user interface for the generation is also improved.

She discussed her involvement in our contribution to the CERN wide J2EE public services (where the holding point for magnetic measurements and the geometry viewer serve as prototypes), updates on the automatic holding point for security and she showed examples from the cable data base for the mailing lists. She also discovered a problem with the concurrency of sessions, which may explain some problems that we have encountered in the past.

Correlation between sagitta and the “positioning triangle” (presentation by Jerome Beauquis, see slides “CFP&Sag&Shape1.ppt”):

The “positioning triangle” is the triangle formed by the position of the three cold feet pads. At itp20 the centre of the pads is derived from fitting Leica measurements at 4 points on the circumference of the pads to a circle. Jerome studied, according to a discussion between Marta, Frederique and Jean-Pierre, the correlation of the sagitta taken from the database (in the database the sagitta is calculated and stored at each upload of a new measurement) and the height of the positioning triangle at itp20. The result is that this correlation seams to be too week to conclude a sagitta value from the measurements of the positioning triangle with the precision we need for being able to guarantee a sagitta within tolerances by controlling the position of the cold feet pads in industry. We have in general differences of around 1 mm in the sagitta for the same measurement value of the cold feet pad centres (maximum values of 2 mm found). The correlation is different for different firms but for all firms differences of the order of mm are found. Jerome also looked at the correlation for values of the sagitta where the calculation of the sagitta in the data base gave an error (also stored in the database, see slides Zhengwei) less than 0.15. Even in this case we see that the we have up to 1.5 mm of variation of the predicted (according to best fit) and measured sagitta. In the discussion it was suggested that the reason for this lack of correlation may be measurement errors, an imprecise positioning of the cold feet pads or local deformations on the cold mass.

The change in sagitta and the change in positioning triangle for two measurements (itp20 and WP01) were compared. Even for the change of shape (comparison between two measurements), no correlation could be deduced. Possible reasons may be local deformations or measurement errors.

The triangle of the shape, measured as the centre of the cold bore tube in the position of the cold feet pads, and the sagitta seems, according to Jerome’s presentation, not correlated either. The reason for this could be the same as for the lack of correlation between the sagitta and the positioning triangle: measurement errors, imprecise positioning of the cold feet pads or local deformations. The triangle formed by the position of the flanges and the position of the cold bore tube centre at the position of the mid cold foot pad and the sagitta seem to be better correlated. The reason may be related to the use of the data from the shape measurement, fit to tenth order polynomial.

Three point based sagitta adjustment (presentation by Marco La China, see slides “threepoints1”):

With the assumption that there are no local deformations and that the shape of the magnet is circular there seems to be a good correlation between the sagitta and the positioning triangle. Here the shape measurement has been fitted to a tenth order polynomial and the positioning triangle has been taken as the values of this polynomial at the nominal position of the cold feet pads. The theoretical value of the relation between the positioning triangle and the sagitta is 1.75. Marco found by calculating the relation for 99 magnets a coefficient of 1.46. This difference could not be explained.

The good correlation could be due to the use of the theoretical positioning of the cold feet pads. The use of the measurement of the cold bore tube centre may also have an influence (Smoothing?).

Dependence of the sagitta on the fit errors and of position of  the “half sagitta” (presentation by Zhengwei Gao, see slides “DSagitta with RMS etc1.ppt”):

The suspicion that the shape change between different measurements is not the difference between two circles, triggered this study. The fit error of the shape to a second order polynomial may be related to the sagitta if there is some non circular shape change. There seems to be no such dependence, the fit error is equally small for large and small calculated sagitta values.

The value of the half of the sagitta with respect to the mean plane (“residue”) is an indication of the form of the shape measured. If this value is correlated to the calculated sagitta we have an indication that the shape is of similar “form” if we compare two measurements. This seems to be the case.

Further study: The change of the positioning triangle between itp20 and WP01 (Davide’s jig measurement) and the change of sagitta between itp20 and WP01 should be estimated. The sagitta is not known at WP01 but may be deduced from WP08 by knowing the adjustment at the central foot and the sagitta at WP08 (in database). The adjustment procedure at fiducialization works very well in production and uses the sagitta calculated in the database to calculate the adjustment of the central support post at WP08.

The problems with the correlation of the sagitta and the measurements of the cold feet pads may be a problem (nominal shape cannot be guaranteed) if we want to use the method of “pre-shaping” the magnets in industry to adjust spool pieces and flanges. To be followed up.

Reminder:

These presentations were made to support two different proposals.

Method 1: Pre-shaping the cold mass at industry to nominal cold feet pad positioning.  At  CERN the shape should be reproduced at cryostating by adjusting the cold feet pads to the position in industry.

Method 2: Measuring the positioning triangle in industry and at CERN and deduce the sagitta change. At fiducialization the sagitta change should be corrected and the shape should be close to the shape in industry (itp20).

Pen:  Elena Wildner