SHMS NGC Cerenkov. Donal Day University of Virginia. Hall C Readiness Review August 24 & 25, 2016 Newport News

Transcription

1 SHMS NGC Cerenkov Donal Day University of Virginia Hall C Readiness Review August 24 & 2, 216 Newport News 1

2 Outline HMS Recent Performance SHMS Nobel Gas Cerenkov How it got done Design Principles and Constraints Expected Performance Construction Details Status Extra Parts Acknowledgements Extra Slides 2

3 We have done this before FRONT VIEW +Y +Z +X 12/8/94 +Y TOP VIEW +X +Z 1 cm 1/2/1 hydroformed windows rohacell mirror phototube magnetic shield base cm BNC HV 3

4 How it got done Design and Simulation UVa Tank Drawings and Fabrication JLAB Design Procurement/Fabrication of everything inside the tank; mirrors, pmts, mirror mounts, pmt mounts, etc: UVa Glass blanks: 43 by 43 cm by 3mm with R = 13 cm Rayotek Scientific, Inc Measure Shapes Apex Metrology Roughness Examination UVa's Zygo white light interferometer Glass Coating: Al followed by MgF 2 UV reflectance to exceed 7% (at 1 nm) CERN EP-DT Group PMTs and and bases: 14 stage, low noise, inch quartz window, positive HV: 9823QKB4 (PMT) and C643KFP-1 (divider) ET Enterprises Ltd, Electron Tubes, UK Magnetic Shields Ad-Vance Magnetics Gas Handling JLAB Tedlar Dupont Window and Port Foam Seals Precision Sheet Metal Fabrication 4

5 Recent HMS Cerenkov Performance Nuclear Dependence of R: E Vahe Mamyan UVa inch Burle PMT 884 coated with WLS. Atm of C 4 F 1 SANE E Anusha Liyanage HU.42 Atm of C 4 F 1 π Threshold: 4 GeV/c "#$%&'! Efficiency dip at mirror overlap "#$%&'! ())*+*,%+-!./!1#1,%&$1!2++,3&2%+,4!!678! 1% effect confirmed at hallcweb.jlab.org/ elogs/ Jan+Experiments/ 384 (#&)*!%$+,-.!#/!1#&#!-*-2&.#%'!!./!1#1,%&$1!2++,3&2%+,4!!678!

6 SHMS PID Requirements : negative polarity Experiment p (GeV/c) Req d e - :π - Disc. E (Fpi-3) :1 Spec d NG Cerenkov Spec d Calorimeter Total Expected E (σ L /σ T ) :1 E (pion factorization) (d) :1 :1 (HMS Cerenkov E (c) :1 gives up to 3:1 now) E (x>1) :1 E (g 2 n, d 2n ) >1 2 :1 >2:1 (1:1 above >1 4 :1 6 GeV/c) 4 overlapping spherical mirrors R = 13 cm, 43 by 43 cm 2 m of active length Noble gas at 1 Atm Page

7 Threshold condition : (1 β) < (n 1) cos = 1 n Choice of gases Argon up to 6 GeV and a mixture of Argon and Neon up to 11 GeV

8 Full Featured Geant 4 Simulation V Mamyan, M. Yurov

9 Photoelectron Yield Reflectance % QE % λ (nm) λ (nm) Tube () A N e λ 1 λ 2 R184 UV glass: Neon ET 9823QKB Quartz: Neon 2 cm active length, 8% of vendor s QE Argon 4x

10 Photoelectron Yield Reflectance % QE % λ (nm) λ (nm) Tube () A N e λ 1 λ 2 R184 UV glass: Neon ET 9823QKB Quartz: Neon 2 cm active length, 8% of vendor s QE Argon 4x

11 Mirror Installation 11

12 12 Overlap with beveled edges

13 PMTs 13

14 Tuning Green laser illuminating whole of acceptance Red pencil laser probing range of angles 14

15 Materials in path of electron Materials in path of electron in NGC Cerenkov Item Material Z Atomic mass density (g/cm^3) RL (g/cm^2) RL (cm) Thickness(in) Thickness (cm) # RL Source Entrance Window^* Tedlar ((CH2CHCl)n) Z/A = PDG Gas Ar E E PDG Ne E E PDG Glass SiO PDG Exit Window Tedlar ((CH2CHCl)n) Z/A = PDG Total RL (with argon).42 *See Percent from mirror 8.2 Magnetic Field at PMTs Steve Lassiter, Hall C SHMS Detector Working Group Meeting Aug. 26, 21 Max B x is 16.2 Gauss Max B x is 2.9 Gauss 6-inch shield made of.4 Ad- Vance Ad-Mu-8 1

16 Detector Efficiency We can assume that the photoelectrons have a Poisson distribution W(N, N) = N N e N N! for registering N photoelectrons when N are expected. If by P (N) we denote the probability for the detector (PMT and associated circuitry) to record the pulses due to N photoelectrons, we can write the efficiency of the detector as = P 1 ( N= W(N, N)P (N). Let us assume that P (N) is of the form P (N, N apple N 1; ) = 1, N i.e.: there is a threshold for the detection of N N. photoelectrons. Then the efficiency is of the form =1 e N Hence, we have the efficiency functions 1+ NX 1 N =1 N N N!!. 1 = 1 e N, 2 = 1 e N (1 + N), 3 = 1 e N (1 + N + N 2 /2), 4 = 1 e N (1 + N + N 2 /2+ N 3 /6), etc. 26% 32% 92% 1% 16

17 π Rejection 2 Pions Electrons Cerenkov cut Calorimeter cut Rosen7 R N π/e from.1 to 3 Electrons N pe > 2, ShrTrk > Photo-electrons E track /E 8 3 No cut 2 Cer cut > 2 1 re 3: HMS calorimeter track energy hsshtrk/e vs number of Cerenkov phototrons. The magenta line shows the Cerenkov Pion background 1 cut, the red line shows the Calorimecut. etectors are constructed to decrease the number of background events in order to eve the experiment s specific precision. goals. For 1 this experiment 1. there are two ctors used for particle identification, a threshold Cerenkov detector described in Cerenkov efficiency ion 2.7.3, and an electromagnetic lead-glass calorimeter described in Section main background in this experiment is from negative pions produced by charge ange reactions..99 Cut out electrons Calorimeter energy / HMS momentum 1.1 χ / ndf 19. / 13 p.9971 ± 8.276e- p ±.8 1 p2 82 ± The ratio of pions to electrons varied from.1 to 3 for all Cer + Calo cut Calorimeter cut >.7 Pion contamin E track / E. In this experiment the Cerenkov cut required that number of photoelectrons be Figure 32: HMS central momentum is.71 GeV. Top plot: HMS calorimeter track er than 2 (hcer npe > 2), and the ratio E (GeV) of the Calorimeter track energy divided energy E track /E (hsshtrk/hse) distribution without Cerenkov cut (the blue line) and with Cerenkov cut > 2 (the red line). Bottom plot: The E track /E distribution after Figure 31: First plot: HMS calorimeter total energy hcal et/e Cerenkov distribution when cut > 2 and E track /E >.7 cut (the red hatched area). The solid blue area is the pion contamination. number of photoelectrons are higher than but less than 2. Second plot: The Cerenkov cut e ciency as a function of scattered energy. 17

18 NGC Gas System Ar/Ne mixed using its own MFC system in gas shed Very similar to wire chamber gas system 1 atm, slow flow rate to maintain gas mix purity Initial fill done using high-flow circuit (~1 scfh) switch to low-flow circuit to maintain (~6 sccm) System protected against overpressure by pop-off valves, multiple overpressure/relief bubblers, and automated valve attached to Photohelic switch/ gauge Gas flows electronically monitored and logged Brad Sawatsky, Responsible 18

19 Backup & Status Two HMS mirrors Three NGC mirrors Four inch Hamamatsu UV glass (suitable for coating) One inch ET Quartz tube Huge inventory of experience Assembled and tuned detector in ESB with nitrogen flowing since October 21 Ready for installation and checkout 19

20 Acknowledgements Howard Fenker Brad Sawatsky Bert Metzger Vahe Mamyan Nicholas Philips Mikhail Yurov Steve Greco Dagmawi Abede David Wimer Garth Huber Dan Abrams Jixie Zhang Dien Nguyen Matt Caplan Tyler Cody Matt Biondi Thomas Schneider Tosh Rijal Thanakorn Iamsasri Cole Smith Matthew Nelson Melissa Goldman Stephen Washington 2

21 Extra Slides 21

22 Faro Arm in Astronomy at UVa Coordinate measuring machines (CMM) APEX Metrology, Zeiss G2 Calypso

23 Mirrors - the glass We worked with Rayotek of San Diego which claimed great experience in slumping glass. They were 1 year late - the shapes were very good. We specified R = 13cm Sheet3 Rating Mirror # Z O radius, cm A B C D dev_min, um dev_max, um dev_sig, um Z O A B C D Z O A B C D Z O A B C D

24 Y axis overlap Q. C Top Left Mirror # Position 3Z Q. B Q. C Top Right Mirror # 4 Position 4Z Q. B Spherical Fit - Data Height (um) -2-4 X axis overlap Q. D Q. A Q. D Q. A y (cm) 1 1 x (cm) 1 2 Q. B Q. A Q. B Q. A Conical Fit - Data Elliptical Fit - Data 4 4 Height (um) Height (um) Q. C Bottom Left Mirror # 1 Position 1Z Q. D Q. C Bottom Right Mirror # 7 Position 2Z Q. D y (cm) 1 1 x (cm) y (cm) 1 1 x (cm) 1 2 Red line is laser pointer path for PMT position tuning Spherical Fit - Data Spherical Fit - Data Height (um) y (cm) x (cm) Conical Fit - Data Height (um) y (cm) Elliptical Fit - Data x (cm) 1 Conical Fit - Data Elliptical Fit - Data Height (um) y (cm) x (cm) 1 2 Height (um) y (cm) x (cm) Height (um) 1 y (cm) x (cm) 1 Height (um) y (cm) x (cm) 1 24

25 vertical angle horizontal angle 2 MC produced electron vertical and horizontal angles as a function of X and Y at the front of the NGC window