1. 1 Light Yield results from the KEK tracker test using G4MICE M. Ellis Tracker Phone Meeting 25 th January 2006

2. 2 Outline  Status/Disclaimer  Light versus L1A timing – L1A timing cut  Light yield distribution vs 3HF concentration  Space point distributions for each station  Light yield vs Fibre bundle number  Light yield vs Position along fibre (X planes)  Conclusions so far  Next steps  Timing of potential further analyses

3. 3 Status / Disclaimer  Mapping information used is that provided by Osaka group (same version used by Hideyuki).  Last time I checked, there were still residual alignment/mapping discrepancies.  In order to get results for this meeting, have left these problems for later and widened road- width for tracking and only use B=0 data.  N.B. This means that there is a larger amount of pedestal tail in the tracks than should be the case.

4. 4 Reconstruction  G4MICE reconstruction: Doublet clusters (one or two neighbouring channels. Individual signals must be > 1.0 PE, total cluster size must be > 2.0 PE). Doublet clusters (one or two neighbouring channels. Individual signals must be > 1.0 PE, total cluster size must be > 2.0 PE). Space points (triplets and duplets). Space points (triplets and duplets). Straight tracks fitted from space points. Straight tracks fitted from space points. TOF and VLPC timing information reconstructed in G4MICE (Aron). TOF and VLPC timing information reconstructed in G4MICE (Aron). Timewalk/etc not implemented yet, so PID is not used in this analysis (in any case most data is at 2 Gev/c). Timewalk/etc not implemented yet, so PID is not used in this analysis (in any case most data is at 2 Gev/c). Events are only considered if a track is found. Events are only considered if a track is found.  Residual and chi2 distributions confirm that the decoding/alignment is not yet correct.

5. 5 Data sample  Three sets chosen (Magnet off): 250 MeV/c: Runs 1426 and 1427 – 4972 events. 250 MeV/c: Runs 1426 and 1427 – 4972 events. 325 MeV/c: Run 1422 – 5574 events. 325 MeV/c: Run 1422 – 5574 events. 3 GeV/c: Runs 1291,1292,1293,1294,1297, 1300,1302,1303,1304,1306,1307,1309,1311,1 257,1258,1259,1272 and 1273 (run numbers in yellow still running). Total events from this sample processed for this talk - 76,967 events. 3 GeV/c: Runs 1291,1292,1293,1294,1297, 1300,1302,1303,1304,1306,1307,1309,1311,1 257,1258,1259,1272 and 1273 (run numbers in yellow still running). Total events from this sample processed for this talk - 76,967 events.  Most of the data used is from 3 GeV/c so the analysis will use all runs as one set.

6. 6 Light Yield vs L1A Blue: 2500 ppm 3HF, Red: 3500 ppm 3HF, Black: 5000 ppm 3HF

7. 7

8. 8 Light Yield Distributions  One plot per 3HF concentration divided into old and new stations.  Pedestal much more pronounced in plots than cosmic analysis due to decoding/alignment issues, hence log scale.  Two histograms shown for each plane: Black histogram for all hits Black histogram for all hits Blue histogram for hits that pass the L1A cut Blue histogram for hits that pass the L1A cut

9. 9

10. 10

11. 11

12. 12

13. 13

14. 14 Space Point Distributions  Each plot shows 2D histogram of the position on the station of the points used in tracks for this analysis.  Several areas on each station show hot spots and lighter areas.  Several of the hot spots are due to low signal (~ 2 PE) channels that frequently enter the track fit due to the loose roadwidth and decoding/alignment problems.

15. 15

16. 16

17. 17

18. 18

19. 19 Light Yield vs Fibre Bundle  Profile histograms used to plot the average light yield (not most probable!) vs the fibre bundle number.  The L1A cut was applied for each plot.  Once QA equipment arrives and can make quantitative measurements can compare light yield with QA studies.

20. 20

21. 21

22. 22

23. 23

24. 24

25. 25

26. 26

27. 27

28. 28

29. 29

30. 30 Light Yield vs Position on Fibre  As for previous set of plots, profile histograms used so mean light yield, not most probable!  Only used X planes, so plot light yield versus the Y position (i.e. Position along the fibre).  Looking for attenuation or discontinuities that would indicate a problem in the actual scintillating fibre.

31. 31

32. 32

33. 33

34. 34

35. 35 Conclusions so far  Absolute light yield values (most probable) for the old stations are not the same as those obtained in the cosmic ray test.  This could still be due to the analysis so far (i.e. strongly depends on tracking selection).  General quality is reproduced (i.e. 2500 ppm best in old stations).  Variation in light yield across fibre bundles is seen in some planes.  No substantial variation along the length of a fibre is seen for any of the X planes.

36. 36 Next Steps  Really need to sort out decoding and alignment before repeating analysis.  This will also allow us to use the much larger data set with the magnet on!  TOF/ACC PID would be useful, especially at the lower momenta. Further work needed to make use of this data.  Once we have the final decoding in the native G4MICE format (instead of the Osaka file) it will also be possible to study light yield as a function of D0 connector, station connector, internal waveguide, etc...  Need to feedback results of tracker “autopsy” in order to set analysis priorities.

37. 37 Timing of potential further analyses  Due to the need for this analysis by today several other tasks have been set aside and are now becoming urgent.  Depending on needs at Imperial for tracker QA/autopsy work I might be able to work further with Aron next week.  If geometry/decoding can be improved, then a second processing run can be attempted in early February, perhaps results before the DAQ workshop.