TPC track distortions Review of situation

1
TPC track distortionsReview of situation

• Alain Blondel, Silvia Borghi, Simone Giani,
  Simone Gilardoni and all the people that
  contributed with ideas or suggestions or comments
  during the analysis meetings and offline
• Thanks to Charles Pattison for the ntuple
  productions

2
Outlook

• Two distortions effect
• Electromagnetic effect induces a d0 shift
  dependent on z position (smaller effect, 10 mm)
• Blondels effect (larger effect 30-40 mm)

3
How d0 changes in case of distortions
Definition of d0
-
4
First Effect Presented at analysis meeting on
29th April 2003 and studied in Dydaks et al.
HARP Memo 03-002, 30 June 2003
5
A
29th April 2003

• We consider the following regions.
• the forward part (region A and E)
• the backward part (region C and B2)
• We observe a different effect in these two
  regions.

E
500 mm 285 mm 215 mm
500 mm
C
B2
C -51lt?lt0 B2 0lt?gt35
E 68lt?lt77 A 0lt?lt77 from end-cap
6
29th April 2003
MiniBoone at 8.9 GeV/c
E
C B2
7
No feedback from collaboration until
HARP Memo 03-001 6 June 2003

Therefore, during all of HARP data taking after 8
August 2001, the high voltage of the inner field
cage was lowered by 1.6 than the corresponding
high voltage of the outer field cage. 3 We
thank L. Linssen for having brought the high
voltage misalignment problem to our
attention, and for ongoing calibration
work to ascertain the precise size of misalignment

8
HARP Memo 03-002 30 June 2003

The effect is most prominent at small radius,
where near the target position displacements of
order 10 mm are expected. Positive tracks are
systematically shifted to smaller momentum,
negative tracks to higher momentum.



(As already shown in my previous first plot)

The effects of the magnetic and electric field
inhomogeneities have the same overall effect on
the distortion of track sagittas. Another
distinct feature is the strong dependence of the
displacements on the longitudinal position inside
the active TPC volume, especially around the
nominal target position. This suggests that the
position and the length of the target plays an
important rôle. The higher upstream the track,
the less it will be affected. Therefore, tracks
from thin targets will generally be less affected
than tracks emerging at the downstream end of
long targets. This may shed some light on
observations on TPC track distortions reported
recently 7.

F. Dydak, A. Krasnoperov, Yu. Nefedov
9
BUT
As you can see in the next plots the shift for
positive particles is constant in the direction
of negative d0, but for negative particles the
shift is not always in the same direction of
positive d0 and sometime the shift is equal to
one of positive particles
10
Be thick (1 l) at 12 GeV/c
C
E
11
Be thin (2 l) at 8 GeV/c
E
C
12
Be thin (5 l) at 15 GeV/c with inverted B
E
C
13
Conclusions (on 29th April 2003)
In the forward region (region A or E) the d0 peak
for positive particles is shifted to -10 mm and
sometime the d0 peak for negative particles is
shifted to 10 mm. If we consider the first part
of TPC (region C and B2) this effect disappears
and the d0 distribution is centered in 0. The
shift for positive particles is the same also
for inverted magnetic field.
14
On the same subject CONCLUSIONS Memo 03-002 The
problem of TPC track distortions due to magnetic
and static electric field inhomo geneities has
been adressed in a quantitative way. The high
voltage misalignment between the outer and inner
field cages is identified as the likely primary
cause of sagitta distortions of TPC tracks. The
position and the length of the target plays an
important rôle.


Our state of the art knowledge The Memo 03-002
could confirm the shift for positive particles,
but it could not explain, yet, why sometimes the
peak of negative particle is not shifted on the
positive d0. The last plot shows that a more
dominant effect of distortion exists.
15
Blondels Effect Already presented in analysis
meetings in April 2003
16
Where did we start?
Ta thick at 3 GeV/c
Negative blue Positive red
17
Ta thick at 3 GeV/c
From C. Morones Thesis
18
Ta thick at 3 GeV/c
1/ r
-
O
19
Ta thick at 3 GeV/c
-

- -
-
1/(charge Pt) (1/MeV/c)
20
Miniboone at 8.9 GeV/c
-

- -
-
21
One possible explanation
Distortions due to ExB effects are the cause of
this effect. This effect may not be constant in
time and in space.
22
Different settings
target type beam p (GeV/c) lambda length in z (mm) trigger type data taking pos wrong/ total neg wrong/ total Effect on d0?
be thin 8 2 8 ThinGe8/14 2001 no
be thin 12 2 8 Central/11 2001 31.8 3.2 yes
miniboone 8.9 5 20 Thick /14 2002 no
be thick 8 1 400 Thick /14 2002 no
be thick 12 1 400 Thick /14 2002 no
C 3 2 8 ThinLe3 /14 2001 20.4 6.3 yes
al thick 12 1 400 Thick /14 2002 28.5 6.8 yes
k2k 1 lambda 12.9 1 400 Beam /14 2001 no
k2k replica 12.9 2 800 Thick /14 2002 28.4 9.0 yes
ta thick 3 1 112 Beam /14 2001 35.6 7.0 yes
ta thick 12 1 112 Thick /14 2002 31.8 4.5 yes
ta thick 15 1 112 Thick /14 2002 36.3 4.9 yes
23
Summary of table

• The banana effect does not depend in conclusive
  manner on
• the target (al and k2k 1 lambda)
• the interaction length of target
• (k2k 1 lambda and k2k replica)
• beam momentum (Be thin at 8 GeV/c and 12 GeV/c)
• the data 2001 or 2002 (C and Al).

24
What happens if B is inverted?
Be 5l at 15 GeV/c B neg
C 2l at 3 GeV/c B pos
25
What happens if B is inverted?
If the beam is positive the banana has the same
position also when B is inverted. Compatible
with ExB effect. Not yet explained Why is the
d0 peak at 0 always present? (demonstrates that
this ExB effect is not constant ) Why not in
coincidence with understood spill time structure
or run time structure?
26
Another possible explanation (additional or
alternative)

• Beam effects
• Rate
• Size
• Optics.

27
Does the banana depend on the beam?
C 2l at 3 GeV/c B pos no beam interruption
between runs with banana and runs without banana
First run 9377-9381
28

• After 7 runs
• On 15th October 2001 at 12.39
• vertical collimator settings 53.8
• events per burst 60
• On 15th October 2001 at 17.56
• Beam expert has finished tuning the beam.
• On 15th October 2001 at 18.13
• the vertical collimators to - 52.0.
• events per burst is 80

? ABRACADABRA
29
Does the banana depend on the beam?
C 2l at 3 GeV/c B pos no beam interruption
between runs with banana and runs without banana
First run 9377-9381
First run 9389-9449
30
What happens? (preliminary)

1. In C setting (and also in a Be setting) the
   banana disappears when the beam spot size
   decreases.
2. The spot size alone does not create the banana,
   in fact in Be 2 lambda 15 GeV/c a smaller spot
   size shows the banana effect
3. We never observed the banana with negative beam.

31
Preliminary conclusions
In these cases the spot size and the focusing of
the beam seems to play a important role. The
rate of T9 protons seems not to matter.
32
Space dependence of the banana effect

• The banana is always present in all regions and
  no evident changes are noted.
• The tracks belonging to the banana have no strong
  dependencies by
• ?0
• tan(?)
• z0

33
Conclusion

• The banana effect does not depend in conclusive
  manner on
• the target (al and k2k 1 lambda)
• the interaction length of target
• (k2k 1 lambda and k2k replica)
• beam momentum (Be thin at 8 GeV/c and 12 GeV/c)
• the data 2001 or 2002 (C and Al).
• setting (C)
• magnetic field polarity (Be at 15 GeV/c)
• rate of protons in T9 (C)
• on spill time during the same run
• the spatial regions
• the track parameters ?0, tan(?), z0

34
?
Under investigation

• Does the banana effect depend on
• the beam spot size
• the time in which way
• the ionization of the chamber that
• creates a ExB effect
• If we do not use the settings affect by the
  banana, which amount of data we lose
• If we do not understand really this effect, how
  can we be sure that the effect (in a smaller way)
  is not present in the other settings
• Why does the banana effect changes in time

?
?
?