The AMS-02 3D imaging calorimeter : a tool for cosmic

rays in space

C. Goy

To cite this version:

C. Goy. The AMS-02 3D imaging calorimeter : a tool for cosmic rays in space. XIII Interna-tional Conference on Calorimetry in High Energy Physics (CALOR 2008), May 2008, Pavia,Italy. 160, pp.012041, 2009, <10.1088/1742-6596/160/1/012041>. <in2p3-00287080>

HAL Id: in2p3-00287080

http://hal.in2p3.fr/in2p3-00287080

Submitted on 10 Oct 2008

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinee au depot et a la diffusion de documentsscientifiques de niveau recherche, publies ou non,emanant des etablissements d’enseignement et derecherche francais ou etrangers, des laboratoirespublics ou prives.

   

LAPP-EXP-2008-01 September 2008

  

The AMS-02 3D-imaging calorimeter: a tool for cosmic rays in space    

C. Goy on behalf of the AMS Calorimeter Group

LAPP - Université de Savoie - IN2P3-CNRS BP. 110, F-74941 Annecy-le-Vieux Cedex, France            

Presented at the XIIIth International Conference on Calorimetry in High Energy Physics, CALOR 2008 Pavia (Italy), 26-30 May 2008

  

in2p

3-00

2870

80, v

ersi

on 1

- 10

Oct

200

8Author manuscript, published in "XIII International Conference on Calorimetry in High Energy Physics (CALOR 2008), Pavia :

Italy (2008)"in

2p3-

0028

7080

, ver

sion

1 -

10 O

ct 2

008

Author manuscript, published in "XIII International Conference on Calorimetry in High Energy Physics (CALOR 2008), Pavia :Italie (2008)"

DOI : 10.1088/1742-6596/160/1/012041

The AMS-02 3D-imaging calorimeter: a tool for

cosmic rays in space.

Corinne Goy LAPP, IN2P3/CNRS, Universite de Savoie on behalfon the AMS calorimeter group.Laboratoire d’Annecy-Le-Vieux de Physique des Particules, Chemin de Bellevue, F-74941Annecy-Le-Vieux Cedex

E-mail: corinne.goy@lapp.in2p3.fr

Abstract. The AMS02 Electromagentic calorimeter (ECAL) is based on a lead/scintillatingfiber sandwich. The flight model was tested and calibrated during Summer 2007 in a testbeam at CERN, using 6 to 250 GeV electron and hadron beams. Preliminary results on themeasurements of ECAL parameters and performances are reported in this paper.

1. IntroductionThe AMS-02 experiment is a large acceptance spectrometer designed to operate on theInternational Space Station (ISS) for three years. The physics goals of the AMS experimentare:

• The measurement of the nature and fluxes of the cosmic rays up to the TeV range, includingthe identification of nuclei and light isotopes (3He, B, C, 9Be, 10Be) and the galactic andextragalactic gamma rays spectra to a few 100 GeV.

• The search for an additional component originating from Dark Matter annihilations on topof the antiproton, positron or gamma spectra.

• The search for nuclear antimatter with an expected sensitivity of 10−9 for He/He and 10−7

for C/C.

The experimental setup (Figure 1) consists of a 8 layers silicon tracker inserted in asuperconducting magnet providing a 0.8 Tm bending power. A total of 4 layers of Time OfFlight scintillating system situated on the top and the bottom of the tracker provides the maintrigger and gives the direction of the particles. A set of Veto-Counters surrounds the trackerto reject side tracks. A Transition Radiator Detector and the calorimeter allow for electron,positron and gamma identification, while the Nuclei species are identified by the RICH detectorin conjunction with the Tracker. The total acceptance amounts to 0.5 m2.sr.

2. The calorimeterThe aim of the electromagnetic calorimeter is to measure the energy of electromagnetic particlesfrom a few GeV up to 1 TeV, to produce a stand-alone gamma trigger, to give a rejection againsthadrons with respect to electrons and positrons of the order 10−4 combined with the tracker,and to identify gamma rays.

in2p

3-00

2870

80, v

ersi

on 1

- 10

Oct

200

8in

2p3-

0028

7080

, ver

sion

1 -

10 O

ct 2

008

3 m

3 m

Figure 1. Schematic view of the AMS detector.

The calorimeter is an imaging calorimeter consisting of 9 modules of lead with embeddedscintillating fibers corresponding in total to a thickness of 16.2 cm. Fibers are oriented alongone or the other horizontal direction, alternatively for each module, as depicted in Figure 2,the incidence of the particles being vertical. The calorimeter has a sensitive surface of 648×648mm2.

Fibers are read at one end by 4-anodes photomultipliers (PMt :R7600-00-M4 Hamamatsu)giving an ultimate granularity of 72 anode-cells in the horizontal plane and 18 pointslongitudinally, each cell collecting the light on a surface of approximatively 9×9 mm2. Thefour 30mm-long light guides, the PMt and the front-end electronics are inserted in magneticshielding iron tubes which are maintained around the calorimeter by an aluminium structure.

Figure 2. 3 modules of the calorimeterwith enlightened fibers to illustrate thestructure. There are 9 layers in total.

Figure 3. The calorimeter fully equippedwith front-end electronics.

in2p

3-00

2870

80, v

ersi

on 1

- 10

Oct

200

8in

2p3-

0028

7080

, ver

sion

1 -

10 O

ct 2

008

The flight model fully equipped with the frond-electronics is shown in Figure 3.Dynamically, the expected signal from cosmic rays ranges from a few photoelectrons for

Minimum ionizing Particles (non interacting protons) to about 105 for electromagnetic showersin the TeV range. To cope with this dynamic range, the front end electronics is designed withtwo gains and a ratio high to low of 33 on average.

3. Test-Beam setupThe flight model of the calorimeter was tested on a beam line of the CERN SPS accelerator ringin October 2006 and July 2007. Fully instrumented with the flight electronics, it was carefullywrapped in a plastic film to protect it from dust. It was fixed on a support structure whichcould move along X (horizontal) and Y (vertical) with respect to the beam direction (Z axis)and which could rotate around the vertical axis. With the test beam setup, 5 (resp. 4) layerswere measuring the position and energy in X(Y). Electron beams of energies between 6 to 250GeV and hadron beams with energy ranging from 20 to 100 GeV were used. Data were recordedessentially in the center of the calorimeter and scanning along the X and Y directions.

4. Slow ControlThe calorimeter must be able to survive extreme temperatures without irreversible damages, thereference temperatures being -30 ◦C and + 50◦C. Consequently, it should be able to operate fora wide range of temperature. During the test beam, thanks to sensors located close to the PMts’electronics on the aluminium back panel, the temperature was recorded in several points on eachface and could then be extrapolated in each point. The pedestal dependance with respect to thetemperature was studied and found compatible with the studies done in the laboratory. Figures4 et 5 show respectively a map of the temperature and a typical pedestal dependance. With anaverage dependance of 0.45 ADC counts per degree, 3 channels out of 2592 have a 0 (negative)pedestal at -20◦C.

TOP

BOTTOMSIDE

Cell Position

Layer

Figure 4. Map of the extrapolated tem-peratures on each face on the calorimeter.

T (C)

Pedestal (ADC counts)

Figure 5. A typical dependence of thepedestal as a function of the temperature.

5. Energy reconstructionSeveral corrections are applied to reconstruct the energy of electrons. They are detailed in thefollowing subsections.

in2p

3-00

2870

80, v

ersi

on 1

- 10

Oct

200

8in

2p3-

0028

7080

, ver

sion

1 -

10 O

ct 2

008

Entries 8961 / ndf 2χ 172.2 / 89

Width 0.066± 2.021 MP 0.11± 14.85 Area 98.1± 9117 GSigma 0.142± 5.196 B1 5.067e+00± 6.099e-09 B2 7.44415± -0.05224

ADC counts0 10 20 30 40 50 60 70 80 90 1000

100

200

300

400

500

Entries 8961 / ndf 2χ 172.2 / 89

Width 0.066± 2.021 MP 0.11± 14.85 Area 98.1± 9117 GSigma 0.142± 5.196 B1 5.067e+00± 6.099e-09 B2 7.44415± -0.05224

Figure 6. Example a MIP distributionfitted with a Landau convoluted bya Gaussian function + an exponentialbackground,

#C

ell

Mip - <Mip_PMt> (ADC counts)

Figure 7. Dispersion of MIPs withinPMts observed in test beam.

5.1. EqualizationThe first set of corrections is meant to equalize the response of each anode and to treat the highto low gain ratio individually. The equalization is performed using the MIP signal obtained ineach cell from a scan along the X and Y axis in the middle of the calorimeter by a 100 GeVhadron beam [1]. Each distribution is fitted by a Landau function convoluted by a Gaussianfunction plus an exponential background and an example is given in Figure 6. The MIPs weredistributed with an average of 15 ADC counts and a sigma of 2.3 ADC counts. The dispersionwithin a PMt as shown in Figure 7, amounts to 15% in agreement with the specifications.Figure 8 displays an example of the correlation between high and low gain measurements. Thedistribution of gains measured in the test beam is shown in Figure 9.

Low Gain (ADC counts)

High Gain (ADC counts)

Figure 8. Relation between High to Lowgain for a single anode.

Entries = 1296 Mean 32.89RMS 0.4607

/ ndf 2χ 73.54 / 51Constant 2.53± 70.24 Mean 0.01± 32.92 Sigma 0.0092± 0.4164

Gain (High/Low)30 31 32 33 34 35 36

Nb

of c

ells

0

10

20

30

40

50

60

70

80

Entries = 1296 Mean 32.89RMS 0.4607

/ ndf 2χ 73.54 / 51Constant 2.53± 70.24 Mean 0.01± 32.92 Sigma 0.0092± 0.4164

Figure 9. Distribution of the PMts’ gainsmeasured in test beam.

5.2. AttenuationThe signal is attenuated along the fibers and needs to be corrected depending upon the originalposition along the fibers. Scan along X and Y with electrons at 10 and 30 GeV and with hadrons

in2p

3-00

2870

80, v

ersi

on 1

- 10

Oct

200

8in

2p3-

0028

7080

, ver

sion

1 -

10 O

ct 2

008

Entries 8085

/ ndf 2χ 92.92 / 91

p0 2.6± 116.6

p1 17.1± 2434

p2 15.0± 624.1

p3 0.055± 5.284

p4 0.000102± 0.001361

ADC counts0 500 1000 1500 2000 2500 3000 3500 40000

20

40

60

80

100

120

140

160

180

200

Entries 8085

/ ndf 2χ 92.92 / 91

p0 2.6± 116.6

p1 17.1± 2434

p2 15.0± 624.1

p3 0.055± 5.284

p4 0.000102± 0.001361

h_anode_h1_7_35 Entries 4539

/ ndf 2χ 67.06 / 66

p0 2.75± 94.23

p1 14.6± 1642

p2 12.3± 391.2

p3 0.080± 5.212

p4 0.000192± 0.002083

ADC counts0 500 1000 1500 2000 2500 3000 3500 40000

20

40

60

80

100

120

140

160

180

Entries 4539

/ ndf 2χ 67.06 / 66

p0 2.75± 94.23

p1 14.6± 1642

p2 12.3± 391.2

p3 0.080± 5.212

p4 0.000192± 0.002083

h_anode_h35_7_35

Figure 10. 10 GeV Electron signal measured close (left) and far (right) from thephotomultiplier.

are used to measure the attenuation. An example of the signal of 10 GeV electrons recordedclose and far from the PMt is shown in Figure 10. Similarly the MIP peak evolves along thefibers as shown in Figure 11. All attenuation curves are fitted simultaneously by the followingfunction f × exp(−x/λfast) + (1 − f) × exp(−x/λslow) letting free an individual amplitude percell.

An example of the result of the fit is given in Figure 12. The correction is applied such thatthere is no correction in the center of the calorimeter: the effect of the correction on the energymeasured in one layer is depicted in Figure 15.

ADC counts0 10 20 30 40 50 60 70 80 90 1000

50

100

150

200

250

300

350

Mip Evolution CentreOne EndOther End

Mip Evolution CentreOne EndOther End

Mip Evolution CentreOne EndOther End

Mip Evolution CentreOne EndOther End

Figure 11. MIP signal measured at three differentpositions along the fiber.

mm0 100 200 300 400 500 600

1600

1800

2000

2200

2400

2600

ElectronX10GeV 6Cell35

Figure 12. Example of fit of theattenuation.

in2p

3-00

2870

80, v

ersi

on 1

- 10

Oct

200

8in

2p3-

0028

7080

, ver

sion

1 -

10 O

ct 2

008

5.3. Impact correctionThe energy measurement is sensitive to the impact of the particle with respect to the boundariesof the cell as illustrated in Figure 13. To correct for this, a method adapted from [2] was used.It relies on the fact that the ratio S1/S3 of the energy in the cell with the maximum energy tothe energy of the neighboring ones per layer and summed over all layers in one direction hasa better sensitivity to the impact point (Figure 13). A simple dependance of this ratio to theenergy can be observed as shown in Figure 14. The normalized inverse shape of this function isused to correct for the impact dependance and the result is given in Figure 16.

X barycenter (Cell Unit)33.5 34 34.5 35 35.5 36 36.5 37 37.5 38

Ene

rgy

in X

(A

rb. U

nit)

0.4

0.45

0.5

0.55

0.6

0.65

0.7

0.75

0.8

30 GeV ElectronsR

atio S1/S3

Figure 13. Sensitivity of the energyreconstruction to the impact positiongiven by the barycenter (small black dots),Sensitivity of the ratio S1/S3 (See text) tothe impact position (large red dots)

S1/S3

Figure 14. Dependance of the energyas a function of S1/S3 (See text).

0 10 20 30 40 50 60 703500

4000

4500

5000

5500

6000

X-Barycenter(cell unit)

Ener

gy

(AD

C c

ou

nts

)

Layer 10

Figure 15. Energy measured in layer10 before (black) and after (grey/red)correcting for the attenuation as afunction of the X position.

Y Barycenter (cell unit)33.5 34 34.5 35 35.5 36 36.5

Ener

gy

in Y

laye

rs (A

DC

co

un

t)

0

5000

10000

15000

20000

25000

30000

Figure 16. Energy before (black) and after(grey/red) impact correction.

5.4. Rear Leakage correctionThe last correction will correct for the energy leakage. The relative leakage energy isapproximated by a linear function of the fraction of energy measured in the latest layers ofthe calorimeter [3]. Figure 17 exemplifies this relation with 100 GeV electrons and in Figure 18,

in2p

3-00

2870

80, v

ersi

on 1

- 10

Oct

200

8in

2p3-

0028

7080

, ver

sion

1 -

10 O

ct 2

008

it can be seen that the linear relation scales with the beam energy. Therefore this dependancewas taken into account using the energy divided by the longitudinal barycenter which is roughlyproportional to the incident energy . Figure 19 shows the reconstructed energy before and afterleakage correction.

Figure 17. Example of the lineardependence of the relative leakage energy,given by the ratio of the reconstructedenergy Erec, (including all precedingcorrections) over the beam energy as afunction of F, the fraction of energy in thelast two layers for 100 GeV Electrons

0 0.05 0.1 0.15 0.2 0.25<

Erec

/Eb

eam

>0.4

0.5

0.6

0.7

0.8

0.9

1

30 GeV150 GeV250 GeV

F

Figure 18. Average relative leakageenergy as a function of the fraction ofenergy in the last two layers for 3 incidentbeam energies.

6. PerformancesPreliminary results on the resolution and the linearity are shown in Figures 20 and 21. Theresolution was fitted to be σ/E = 10.3%/

√E⊕1.6% while the linearity is kept within 2%. Using

the same complete set of corrections, the linearity and the resolution are not degraded whenthe incidence angle of the beam is 15 degrees. Using the fit to the longitudinal profiles of allenergies, displayed for 3 energies in Figures 22, the radiation length X0 can be deduced: it wasfitted to be 1.07(0.02) layer unit. Therefore, the calorimeter consists of 16.9 radiation lengths.

(GeV)beamE0 50 100 150 200 250

(G

eV)

mea

sE

0

50

100

150

200

250 With Leakage Correction

Without Leakage Correction

Figure 19. Energy before and after leakage correction in function of the beam energy.

in2p

3-00

2870

80, v

ersi

on 1

- 10

Oct

200

8in

2p3-

0028

7080

, ver

sion

1 -

10 O

ct 2

008

(GeV)beamE0 50 100 150 200 250

be

am/Eσ

0.015

0.02

0.025

0.03

0.035

0.04

0.045 / ndf 2χ 137.8 / 9

p0 0.0007137± 0.1021 p1 8.982e-05± 0.01596

/ ndf 2χ 137.8 / 9p0 0.0007137± 0.1021 p1 8.982e-05± 0.01596 Preliminary

Figure 20. Preliminary results on theenergy resolution.

(GeV)beamE0 50 100 150 200 250

be

am/E

mea

sure

dE

0.86

0.88

0.9

0.92

0.94

0.96

0.98

1

1.02

1.04

Figure 21. Preliminary result on thelinearity.

Layer Layer Layer

Figure 22. Longitudinal electromagnetic profiles fitted from left to right at the energies of 10,50 and 250 GeV.

7. ConclusionThe beam tests of the flight model of the AMS calorimeter were successful. No channels werefound dead or noisy. Preliminary results for the resolution and linearity are very encouraging:a constant term of 1.6% was obtained while the linearity is kept within 2% in the energy rangefrom 6 to 250 GeV. Further refinements on the different corrections are expected to improve theconstant term. The calorimeter was found to be 16.9 radiation lengths thick.

References[1] AMS note 2008-01-01[2] The L3 Collobration,Nucl. Inst. Meth. A 289(1990) 35

L3 note 1712, (1995)[3] Loic Girard. PhD thesis(2004) and reference therein.

in2p

3-00

2870

80, v

ersi

on 1

- 10

Oct

200

8in

2p3-

0028

7080

, ver

sion

1 -

10 O

ct 2

008