[G12] Track Inefficiencies - Possible Drift Chamber Issue

[Lei Guo]

The TOF paddles are knocked out manually, so you need the paddle IDs…. In fact, rafael just produced another list of additional new bad paddles. MK is rederiving his track efficiency correction map.

To check if the drift chamber wire efficiency map has been applied correctly, you need to compare the lab theta, and phi distributions by sectors. 
Do you account for the number of good events you throw away with the kinematic fitting CL cut? That percentage that we throw away with CL cut is not entirely consistent with the simulation, and you have to correct  for that as well. Or maybe you have already done that?
Another thing one might have overlooked is that if one uses start counter timing. I don’t know if MK’s map is derived after applying the |Stvtime-RFvtime|<1ns cut. SO, it may or may not affect you. The 1ns cut will throw away good events (ST resolution is not ~400ps), and since the ST resolution is not simulated, you have to account for that. However, it really depends on whether the map was derived with the ST timing cut or not.

Lei


Lei Guo
Assistant Professor
Physics Department
Florida International University
Miami, FL

email: leguo at fiu.edu <mailto:leguo at fiu.edu> or lguo at jlab.org
Office:305-348-0234

> On Jul 27, 2015, at 9:53 AM, Michael Paolone <mpaolone at jlab.org> wrote:
> 
> Hey Volker,
> 
> The behavior is odd indeed.  I have some thoughts, but before I jump into
> them, let me know what your channel is exactly and how you indentify it. 
> Is it pi+, pi- and p detected, with a missing pi0 you cut on?  Is it a
> kinematic fit to the missing pi0?
> 
> An overall thought: Have you looked at the distribution of each detected
> particle in the DC (in x,y) at that specific energy to see if you can
> "see" any strange holes or behavior?  Then you could see how it compares
> to simulation.
> 
> -Michael
> 
> 
>> Hi Everybody,
>> 
>> I know that we are all working on finalizing various g12 analyses.
>> However, we found a serious issue with our γp → pω cross section that
>> currently prevents us from moving on. We are somewhat stuck and it may
>> affect the whole run group.
>> 
>> The attached pictures show the 3Ï€ invariant mass for the energy range
>> 1650 - 1700 MeV and for forward angles of the 3π system. A nice ω peak
>> is visible and a massive hole on the right side of the peak. This hole is
>> not supposed to be there (unless somebody has a good physics argument).
>> The energy range is probably very low for most of the g12 analyses.
>> However, the hole will slowly move to higher masses with increasing photon
>> energy but it will not disappear. The other two pictures show the same
>> distribution if one (1) uses events where only sectors 1, 3, 5 triggered
>> or alternatively, (2) only sectors 2, 4, 6 triggered.
>> 
>> We assume the effect is based on track inefficiencies, perhaps dead
>> regions in the drift chamber. In principle, Michael Kunkel’s "trigger
>> map" should account for this since his approach is based on comparing two-
>> and three-track events, i.e. it combines trigger and track inefficiencies;
>> the idea is good. In our analysis however, this trigger map leads to an
>> overall disagreement with the g11 ω cross section, whereas Zulkaida's
>> current cross section is in fair agreement with g11 but exhibits certain
>> problematic regions, e.g. the forward direction. These holes in the mass
>> distributions are not accounted for by the Monte Carlo simulations and we
>> assume the effect is not in the MC.
>> 
>> We have a few questions we would like some help with (and need to find an
>> answer for).
>> 
>> 1) Since it is still possible that the problem is at our end, would
>> anybody be able to reproduce this problem for us? The effect is so big
>> that even a quick and dirty look at it, will probably work.
>> 
>> 2) We tried to knock out dead TOF paddles as suggested in the analysis
>> note. The paddle numbers are available in the data. However in the Monte
>> Carlo, the numbers appear to be available only for the proton and not for
>> the pions. Has anybody else noticed this issue? How do others knock out
>> the paddles in the MC? Or is this done automatically? It is difficult to
>> do this based on measured angles since two particles with the same polar
>> and azimuthal angles, one produced at the beginning of the target and one
>> at the end, may hit different TOF paddles. The g12 target was very long.
>> 
>> 3) The Monte Carlo “gpp" options given in the analysis note do not
>> reproduce the holes in the mass distributions. For this reason, we do not
>> know if dead wires are actually simulated and to what extent. Can anybody
>> comment on this?
>> 
>> This problem shows up in the γp → pω channel but we are concerned that
>> it may also affect the two-pion channel we are analyzing, perhaps not as
>> holes in mass distributions but as general track inefficiencies in certain
>> regions of the drift chamber. In the latter case, it would extremely
>> difficult to notice. If so, it can potentially affect any reaction that
>> uses Monte Carlo for the acceptance correction. My understanding is that
>> we partially use the pω and the KΛ cross sections to make sure that the
>> g12 MC, trigger, etc. is working correctly.
>> 
>> Best wishes,
>> 
>> 	Volker
>> 
>> 
>> 
>> _______________________________________________
>> G12 mailing list
>> G12 at jlab.org
>> https://mailman.jlab.org/mailman/listinfo/g12
>> 
> 
> 
> _______________________________________________
> G12 mailing list
> G12 at jlab.org
> https://mailman.jlab.org/mailman/listinfo/g12

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/g12/attachments/20150727/797d3630/attachment.html>