SIDIS/PVDIS Cherenkov

From Hall A Wiki
Jump to: navigation, search

Tuesday 10/21/14

Some rates from Zhiwen's wiser code. I have refined my trigger code to now work more like n actual trigger (i.e. it organizes hits in time, then opens logic gates to form the trigger per sector). This helpd the code be human readable, and makes the rates more accurate; however, the overall change is on the order of a percent or so.

In general, the Wiser Rates are about twice the Geant4 rates. Almost everything comes from the Pi0's, and 3/4 of that rate comes from leptons entering the Cherenkov window (in other words, 1/4 of the rate comes from conversions inside the cherenkov... Most of that is in the gas and the mirror. Only a small amount seems to come from the entrance window, the PMT-glass, or in the cone.

The rates from Zhiwen's pions are as follows (For PVDIS config):

Pi0

Rates per sector
Trigger config Rate per sector (MHz)
2 or more pe's in at least 1 PMT 7.70
1 or more pe's in 2 different PMTs 10.91
2 or more pe's in 2 different PMTs 4.77
3 or more pe's in 2 different PMTs 2.80
4 or more pe's in 2 different PMTs 1.66
2 or more pe's in 3 different PMTs 2.76


Pi-Minus

Rates per sector
Trigger config Rate per sector (MHz)
2 or more pe's in at least 1 PMT 0.17
1 or more pe's in 2 different PMTs 0.21
2 or more pe's in 2 different PMTs 0.11
3 or more pe's in 2 different PMTs 0.067
2 or more pe's in 3 different PMTs 0.046


Pi-Plus

Rates per sector
Trigger config Rate per sector (MHz)
2 or more pe's in at least 1 PMT 0.093
1 or more pe's in 2 different PMTs 0.15
2 or more pe's in 2 different PMTs 0.062
3 or more pe's in 2 different PMTs 0.046
2 or more pe's in 3 different PMTs 0.031


Some more information about the pi0 events: Zhiwen's root file has 61k events. Of those, 3k have at least one optical photon that makes it into a pmt. From there, ~900 events pass the 2x2 trigger!

Tuesday 9/30/14

List of questions

  • Uniform signal and background event generation across a "final" SoLID detector configuration (We have this now, I think).
    • I need to run those files through my simulation and generate "finalized rates"
      • Possible issue: At the moment I am using an older version of GEMC / geat4 with older physics databases. Can cause discrepancies between simulations done by others.
  • Background rate issues:
    • High rates from secondary / cascading reactions
      • These seem to come from hadronic reactions
        • Geant vs Wiser
    • All Cherenkov background eventually comes from Geant based EM process
    • Tracking back to exact processes are difficult in geant (Rich has done the most work on this)
      • We can say a few things, rates can be divided into 3 general categories:
        • Before PMT glass -- Most of rate.
        • in PMT glass -- Rate ~ 10%
        • after PMT glass (leptons incident on anodes) -- Rate = ?
    • Sources of optical photons have reaction chains including:
      • Moller
      • charged pion knock-ons
      • neutral pion decays
      • neutron scattering
  • Increasing signal/background
    • Wavelength shifting plus higher photoelectron cut should help.
  • Problems with high rates:
    • DAQ rate
      • Dead-time
    • Longevity of PMT
  • What needs simulation / what needs real prototype?

Tuesday 11/26/13

SIDIS backgrounds

Using the Geant4 physics lists with the SIDIS geometry configuration and a 15uA beam on the He3 target, I get a backgroud rate through the LGC window:

Fig. 1: Low energy background flux through the LGC window
Fig. 2: Low energy background flux through the LGC window with energies high enough to create optical photons

This results in rates of at least 1 photoelectron per sector as follows:

Rates per sector
Trigger config Rate per sector (MHz)
1 or more pe's total 0.319
2 or more pe's total 0.219
4 or more pe's total 0.134
1 or more pe's in 2 different PMTs 0.128
2 or more pe's in 2 different PMTs 0.084

Wiser rates

Using Zhiwen's code, I have found the associated hadron rates in the SIDIS configuration.

Rates per sector
Trigger config pi0 target (kHz) pi0 up-win (kHz) pi0 down-win (kHz) pi- target (kHz) pi- up-win (kHz) pi- down-win (kHz) pi+ target (kHz) pi+ up-win (kHz) pi+ down-win (kHz) total (kHz)
1 or more pe's total 39.71 4.35 18.66 5.04 3.06 3.47 6.38 3.14 3.93 87.74
2 or more pe's total 31.21 2.16 13.63 3.57 2.11 2.48 4.37 2.18 2.67 64.38
4 or more pe's total 22.0 0.84 9.00 2.09 1.23 1.41 2.53 1.32 1.38 41.80
1 or more pe's in 2 different PMTs 27.45 1.47 11.83 2.60 1.56 1.90 3.13 1.71 1.96 53.61
2 or more pe's in 2 different PMTs 21.8 0.79 8.69 1.93 1.11 1.34 2.36 1.25 1.30 40.57

Direct on PMTs

I've also looked what particles pass the PMT face (after the PMT glass). For the PVDIS configuration, this is what I see:

Fig. 3: PVDIS PMT-face flux
Fig. 4: PVDIS PMT flux gamma vertex
Fig. 5: PVDIS PMT flux electron/positron vertex

For SIDIS:


Fig. 6: SIDIS PMT-face flux
Fig. 7: SIDIS PMT flux gamma vertex
Fig. 8: SIDIS PMT flux electron/positron vertex

Tuesday 9/10/13

Some comparisons with the latest baffle design (tentatively named "more1"), and how it is different than the previous design with less baffles but an inner radius for the first baffle of 5cm.

Comparison of total rate through cherenkov window:

Fig. 1: Flux rates for previous baffle (blue) and latest baffle "more1" (green). This is the rate over the entire cherenkov window.
Fig. 2: The ratio of the rates (new/old)
Fig. 3: Flux rates above threshold for previous baffle (blue) and latest baffle "more1" (green). This is the rate over the entire cherenkov window.
Fig. 4: The ratio of the rates above threshold (new/old)

A comparison of PMT rates (> 1 pe) for the newer and older designs:

Rate comparison between baffle designs
Previous Baffle (MHz) "More1" (MHz)
1 or more pe's per sector 4.94 2.99
2 or more pe's per sector 3.44 1.93
1 or more pe's in two different PMTs 2.50 1.56

Tuesday 8/6/13

Baffle inner radius at 5cm / no beam line shield / no absorber.

Fig. 1: The Z-R vertex distribution of electrons and positrons immediately after the cherenkov window
Fig. 2: The Z-R vertex distribution of electron and positron mothers immediately after the cherenkov window
Fig. 3: The pid spectrum for electron / positron mothers
Fig. 4: The rate versus photoelectron cut for sum (blue) or 2 pmt coincidence (red)

Total statistics:

95 out of 2e8 events have at least 1 photoelectron.

64 / 95 come from electrons/positrons (0 / 95 come from leptons with less than 10 MeV Energy)

5 / 95 come from pions

the remaining 26 are still a ??

Friday 7/26/13

I've completed a new study of the low energy background's effect on the LGC rate.

Initial results

We have in this case the PVDIS baffles (with a 4cm inner radius at the first baffle), the standard exit beampipe, and the tungsten absorber. Electrons produced from the low energy photons seem to be the primary source of background signal. The Z-R spectra of photon events that cross the Cherenkov window are:

Fig. 1: The Z-R distribution of photon events crossing the LGC window

Notice most events come from the inner radius of the first baffle and the front of the absorber.

For this configuration, my calculated rate WITH NO TRIGGER CUTS is 61 MHz per sector (or 61/9 = 6.8 MHz per PMT). If we put a low-level trigger to restrict us to 2 or more photo-electrons per sector, the rate drops to 33 MHz per sector (or 3.6 MHz per PMT). There are many logical combinations for triggers, but another example is to have at least 2 PMTs fire per sector, each with at least 2 photo-electrons; in this case the rate drops to 22 MHz per sector (2.5 MHz per PMT).

I've started running new events with a larger inner radius for the first baffle (5cm as opposed to 4cm), also with no absorber or exit beampipe. I'll post resulst as soon as I have them.

Tuesday 6/24/13

A look at the trigger efficiencies for one PMT sum vs multi-pmt.

Fig. 1: SIDIS trigger efficiency comparison.

Tuesday 4/30/13

To expand to lower angles of SIDIS coverage, an additional 3rd mirror was added to the simulation.

  • Details and concerns:
    • The middle set of mirrors share the overlap coverage of PVDIS 22-33 deg and SIDIS 8-15 deg, The PVDIS angles > 33 deg are handled by the outermost of mirrors, and the SIDIS angles from 7-8 deg are handled by the innermost set of mirrors.
      • This is needed if we want to incline the middle mirror for use in both SIDIS and PVDIS.
    • The middle mirror quickly looses efficiency due to a reflection offset for SIDIS low angles. This is due to the size of the mirror and an inefficiency of a simple spherical mirror to uniformly reflect the incident rays onto the PMT face, due to the transport of the radiating particle in the magnetic field.
    • The new set of mirrors sits unrealistically close to the border of the tank. Would probably need to extend tank or reduce angle coverage of mirrors.
    • If we wanted SIDIS < 7 deg, we would need to not only change the diameter of the snout, but also adjust the winston cones between SIDIS and PVDIS configurations; the cones aren't wide enough to accept SIDS < 7 deg and PVDIS > 34 deg simultaneously.
Fig. 1: Cross section of 3 mirror design
Fig. 2: Reflections vs Angle
Fig. 3: P.E's vs Angle
  • Some possibilities:
    • Redesign for a larger inner most and smaller middle mirror, have both mirrors incline when switching to PVDIS.
      • Expected result: larger SIDIS p.e.'s at small angles, less PVDIS p.e.'s at small angles.
    • Redisign for separate mirrors and cone positions for PVDIS and SIDIS:
      • Expected result: Better overall acceptance at edges (small angle / large angle). Large increase in mirror cost and difficulties in tank design.