There was a one centimeter discrepancy between the UTC geometry center and the magnet center in E787.
In E949, this discrepancy still persists. Using a B=0 cosmic ray data set, one finds the expected  T-counter, I-counter
(in which the Z positions are the UTC track extrapolation to the center of the hit counter) and  B4  (where the
UTC track is extrapolated to get the Z position for r=0 in r-z plane if B4 is hit)  positions are offset by one centimeter.
For example, the ends of T-counter should be at -26 cm and +26 cm and B4 location should be at -12 cm.
This offset dose not affect the data analysis as long as the cuts for matching UTC to these counters are adjusted.
In E949, for the sake of convenience, the UTC geometry center will be shifted to match the magnet center. A simple -1 cm adjustment to the cathode strip Z positions makes the  T-counter, I-counter  and  B4  stand at  the expected Z positions.  Previous alignments for the target and RSSC etc. should be re-studied again.

UTC cathode strip position alignment

Using B=0 cosmic ray data,  we can run a special job with  the source codes ( cosmic_utc.F ,  cosmic_linfit.F  and hbkini.F )
the unix script ( make_cosmic_utc ) for creating an executable file and a job running script  (run_cosmic_utc)  using the input parameter file  (cosmic.kcm) . A straight line fit is performed with up to 12 Z points. In this study, a requirement of at least
10 Z points is imposed to increase the precision for checking the present UTC cathode strip position.   In the created hbook
file, ID=1131, ... 1136 are the plots for the means of Z residuals (Zmeas - Zfit) vs. Zmeas  which indicate somewhat Zmeas dependence.   A linear fit to the Z residuals vs Zmeas. The parameters of A0 and B0 can be used to correct the parameters in the UTC geometry file ut_geo.***** as given below
...
    0.3599679E+00 1.0047923  -1.2668037
  -0.3441262E+00 1.0187366  -1.1418120
    0.3110332E+00 1.0055414  -1.0907803
  -0.1215098E+00 1.0076092  -1.0607214
    0.1697496E+00 1.0020679  -0.9697966
  -0.0023262E+00 1.0064859  -1.0361597
...

where the second column is the correction to the strip tilt angle tan(theta) and the third column tran_z is for the foil translation
corrections. The new correction constants can be obtained by tan(theta)/(1-B0) for the strip tilt angle and
tran_z - (L/2+A0) for the foil translation. Where L is the effective foil length. Since few Z points are used in the fit,
resulting in somewhat correlation among the foils. It is necessary to have an iteration procedure by using the new ut_geo.****
as an input, repeat the above procedure until the means of Z residuals vs. Zmeas looks satisfactory (see  these plots  for example).