Problems with applying TPC alignment
I ran a test (in SL09e) where I made the alignment of one sector grossly wrong. In this case, I set the outerSectorRotationAngle of sector 8 to about -7 degrees (normally these angles are less than a tenth of a degree). In the following sets of plots pairs, the left plot is shown without modifying the alignment values, and the right plot shows the effect of this gross modification (performed by having a local CINT file in
The first pair of plots show all 24 sectors, with sectors 1-4 going from left-to-right across the top row, 5-8 from left-to-right across the second row from the top, etc. It is apparent from these plots that something happened to the reconstructed outer sector alignment in both sectors 8 and 16!
Looking more closely at sector 8 shows that its outer sector did not get rotated, but instead was shifted (translated) by about -15 cm (I thought it might be -14, as in 2 x -7, but it really looks more like -15):
Looking more closely at sector 16 shows that its outer sector did get rotated!
One might wonder whether I've got data from one sector incorrectly drawn in another sector (even though that would not explain exactly the different effects seen), but the dead regions of the sectors match up well with where they are supposed to be, so I do not think this to be a problem.
I will not show it here, but a similar test which used an outerSectorRotationAngle of -7 degrees on sector 11 caused a translation of about +15 cm in outer sector 11, and a rotation in outer sector 13 (this other test did not use CINT files but actually had an erroneous number in the offline database). This gives the impression that sectors which are at the same global phi but on opposite halves of the TPC seem to be paired up in this issue.
It appears there are some errors in applying the alignment values obtained from the database.
-Gene
- genevb's blog
- Login or register to post comments