Minutes of MAS-MA geo-meeting

 06/10 2005

_______________________________________________________________________

 

 

Present: Marta Bajko (only first point), Jerome Beauquis, Karl Martin Schirm, Patrick Winkes, Elena Wildner

 

Pen:  ElenaWildner

 

 

Subject:

Continuation of the work on the SSS:

 

Data Base:

The data from ACCEL have now been arranged in a template as discussed in preceding meetings (Jerome and Patrick) so that data can be uploaded to the MAS database. As discussed in one of the preceding meetings, the measurement is now also a complete measurement of the geometry just before delivery (in the 2*6 bosses system). This concerns data from August and onwards. These measurements will be put in the data base. Jerome will provide the script so that SU can make the uploads to the data base directly after the measurements for future measurements.

As decided in meetings already held, all data before the new template was made will be put in the data base when resources are available (a data base exists at ACCEL and will be transferred to CERN after the project is finished). Karl will in the meantime get an EXCEL extraction of all data (he needs around 2 weeks) with tracing of the evolution and decisions concerning the measurements, references etc.). Marta agrees that this gives her the necessary information about the situation at ACCEL.

 

 

Simplified experiment concerning the movements of the lines:

Extra reference points (D-points) were welded on the flat side of th SSS (M.Modena, E.Wildner, http://lhc-div-mms.web.cern.ch/lhc-div-mms/MMSPAGES/MA/GeoMeetings/02-05-05.htm) to check the movements of the end cover and the lines during cold test. The results are now available and show significant movements of up to 1mm. See table below. The movements are expressed in a local coordinate system on the ends. Coplete measurements will be made to complete the results.

 

Recalculation of mean plane on magnetic part of quadrupole:

Patrick was asked to make a best fit of the magnetic data of the quadrupole, instead of the mechanic over the whole SSS, on 3 SSS. The aim of this was to check the data in a coordinate system close to the assembly system (inertia tube).

Patrick very quickly produced the results and they show that we have a significant contribution from the mathematical treatment of the data. Patrick did not have the time to check the contribution of the quadrupole only, which would be better. He will do this exercise on the 3 magnets.

 

The best fit of the data on the whole tube was decided in a WGA with strong objections from MAS/MA who recommended best fit on the QUAD part for several reasons:

  1. Theoretically the quadrupole should sit on the theoretical axis as base line.
  2. The quadrupole magnetic axis has the closest correlation to the inertia tube by construction. The tolerance on the CBT is also smaller over the QUAD.
  3. Doing the best fit on the undulating tube clearly introduces difficulties of data comparison with ACCEL.
  4. It was agreed that aperture is primordial for the machine and that by shifts the aperture criteria could be satisfied with this method.

 

It is clear that if we compare tolerances in two systems we will have problems. The situation can only be improved by changing the best fit (this may solve a large part of the problem (to be investigated in detail by more calculations) or by measuring using a full fididucalisation in ACCEL. It is clear that if we have real movements of the flanges a mathematical manipulation will not solve anything, but we would know better from where the apparent movements come.

 

The data base contains now data with best fit on the CBT and shifts to arrange that the QUAD is correctly positioned on the theoretical axis.

 

Summary:

Data extractions needed will be provided from ACCEL and CERN databases.

Local movements of up to 1 mm on the flanges have been seen between before and after cold test.

The best fit procedure seems to contribute significantly to the problem with tolerances.