simulation study of ion back flow for the alice-tpc upgrade

34
Simulation study of Ion Back Flow for the ALICE-TPC upgrade Taku Gunji Center for Nuclear Study University of Tokyo 1 RD51 Collaboration Meeting at SUNY, 5.10.2012

Upload: laasya

Post on 23-Feb-2016

41 views

Category:

Documents


0 download

DESCRIPTION

Simulation study of Ion Back Flow for the ALICE-TPC upgrade. Taku Gunji Center for Nuclear Study University of Tokyo. RD51 Collaboration Meeting at SUNY, 5.10.2012. Outline. ALICE GEM-TPC upgrade Measurement of IBF in RD51 Lab . at CERN Measurement of IBF at TUM - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Simulation study of Ion Back Flow

for the ALICE-TPC upgradeTaku Gunji

Center for Nuclear Study University of Tokyo

1

RD51 Collaboration Meeting at SUNY, 5.10.2012

Page 2: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Outline• ALICE GEM-TPC upgrade • Measurement of IBF in RD51 Lab. at

CERN• Measurement of IBF at TUM• Status of IBF simulations on June 2012• Update since then

– IBF vs. charge up of GEM– IBF vs. space-charge above GEMs

• Summary and Outlook

2

Page 3: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

ALICE GEM-TPC upgrade• LoI of the ALICE upgrade

– https://cdsweb.cern.ch/record/1475243/files/LHCC-I-022.pdf• High rate capability

– Target: 2MHz in p+p and 50kHz in Pb-Pb collisions• Plan for the ALICE-TPC upgrade

– No gating grid and continuous readout• Inherited the idea from ILC/PANDA GEM-TPC [arXiv:1207.0013]

– MWPC readout will be replaced with GEM. – Keep current gas composition: Ne(90)/CO2(10)

• Issues for the GEM-TPC upgrade – Stability of GEM operations (gain, charge up, discharge, P/T)

• Prototype GEM-TPC will be installed/tested at ALICE in 2012.– Good dE/dx resolution for the particle identification

• ~5% for Kr by PANDA GEM-TPC. Comparable to the current ALICE-TPC.• Prototype will be tested in 2012 at CERN-PS T10 beamline

– Ion back flow to avoid space-charge distortion• Requirement < 0.5%• Measurement using test bench in CERN, Munich and Japan• Simulations to search for the optimal solutions

3

Page 4: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

IBF measurements at CERN

Ampteck Mini X-ray tube

Ag target: Ka=22KeVRate (Ar(70)/CO2(30))= 5e7 estimated by Id

4

C. GarabatosY. Yamaguchi

• Systematic measurement is on-going at RD51 lab.

Page 5: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Rate, # of seeds/hole5

• Estimation of rate/hole, # of seeds/hole in the lab test and Pb-Pb 50kHz collisions.– Lab. test at CERN

• X-ray rate: ~105Hz/mm2,# of seeds: ~1000• # of seeds/hole (4cm driftdiffusion~500um)

– 1000/(500um)2*(100um)2 = 40 (or less ~ 20)• Rate/hole: ~105Hz/mm2 x (0.5mm)2 ~ 25kHz (40usec)

– Pb-Pb 50kHz• Occupancy (IROC:4x7.5mm2) = 50%• Seed electron density (Nch*100*dR/S)= 150e/cm2 • # of seeds/hole (1m drift diffusion~3mm)

– 15(e/(3mm)2) / (3mm)2 x (100um)2 = 0.02• Rate/hole (with seed):50kHz*50%*0.02 = 0.5-1kHz

(~msec)• Much relaxed conditions compared to lab. test at CERN.

Page 6: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Results of IBF at CERN6

• Extensive study for the parameter dependence– 0.25% can be achievable.

• Comparable to ILC/PANDA GEM-TPC– Study for Ne/CO2 (90:10)– Strong VGEM dependence

Ne/CO2/N2

M. Killenberg et al. NIM A530, 251 (2004)B. Ketzer et al. arXiv:1207.0013

Page 7: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Rate/Position dependence7

• X-ray rate dependence and tube position (from GEM1) dependence– Current of primary ions is linear to tube current.– IBF strongly depends on:

• (VGEM)• Rate • Position from GEM1

– Less diffusion for 1.5cm.– IBF is strongly affected by local charge density???• space-charge/recombination

Caveat: The conditions of this measurement are far away from the conditions expected in 50kHz Pb-Pb.

Page 8: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

IBF Measurements at TUM8

• Systematic and simultaneous studies of IBF, gain, and energy resolution.

• Reading out currents from all electrodes.Caveat:Rate of X-ray is ~10% of that at CERN

B. Ketzer, A. Honel from TUM

Page 9: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Results of IBF at TUM9

• VGEM dependence – Gain increases as expected. Resolution gets better

as higher VGEM – No VGEM dependence of IBF • Due to smaller rate (10%)?

resolution

gain

IBF

Page 10: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

IBF studies in simulations• First results were presented at the last meeting on

June 2012.• Discrepancy between measurements and

simulations.– Measurements at CERN: strong VGEM dependence– Simulations: No VGEM dependence (agree with TUM results)

10

2 GEMs2 GEMs

Page 11: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Possible Reasons• Charge-up on the Kapton surface– CERN GEMs (bi-conical shape) are used in the

measurement.– Measurement with cylindrical GEM holes will be

done in the lab.• Space-charge– Clear VGEM, rate and position dependence at CERN

• Larger local electrons/ion density leads to smaller IBF?

• Since x-ray rate/hole is ~25kHz at CERN Lab., remaining ions in the space affects IBF?

11

Page 12: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Charge-up simulations• Simulation setup

– 1 GEM (50um. Bi-conical). HV=400V (gain=50). Ar/CO2=70/30

– Kapton surface area is divided into 16 segments. • Procedure

– 1: Generate 100 avalanches. Then calculate # of ions and electrons absorbed in each segment.

– 2: Put (these # of electrons and ions x 5000) into Kapton surface and calculate electric field.

• Equivalently, 100x5000 (=5x105) seeds in one cycle

– 3: Repeat 1 & 2 for many times.

12

Page 13: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Accumulation of charges in Kapton

• Nelc-Nions at each segment vs. iteration cycle• Nelc/Nions are saturated at 0.5-1x107 seeds at

the gain of 50 (HV=400).

13

Lower GEM

Ed

0/16

16/16

Seeds

electrons electrons

ions ions

Page 14: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Accumulation of charges in Kapton

• Nelc-Nions at each segment vs. iteration cycle• Nelc/Nions are saturated at >1x108 seeds at the

gain of 50 (HV=400).

14

Lower GEM

Ed

0/16

16/16

Seeds

electrons electrons

ions ions

Upper GEM

Page 15: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Gain vs. cycle• 10-20% increase of Gain is seen.

– Due to many electrons at the bottom of Kapton, potential around there gets lower and electric field gets larger.• Gain increases. Avalanches happen more at the bottom.

15

Average electron creation Point in z [cm]

Total electrons Electrons in induction

Page 16: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

IBF vs. cycle• No big change of IBF with charge up– Ions escaping to drift space come from the center of the hole in R.– No big change of <R> of creation points.

16

Average avalanche pointsIn R [cm]

IBF

RZ

Page 17: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Space charge simulation• Very simple simulation for space-charge• Strategy– Make volume to put ions.

• Volume : 70 (pitch/2, X) x 70*sqrt(3) (Y) x 100 (Z) um3

– 100um is chosen since the spread of ions (after avalanches) is ~100um (more or less) above GEM.

• Replica of this volume by mirror symmetry• 10 volumes above GEM covering [0, 1mm] from top

of GEM– Put ions (from 0-105~106) in one of 10 volumes– Electric field is calculated.– Make avalanches

17

DZ=100um

1mmions

Page 18: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Electric Field above GEM• Examples on the change of the field – 0, 105 and 106 Ions in [0, 100um] above GEM.

• 10usec after avalanches– Field Strongly depends on the number of ions.– More ions are curled up and absorbed at the

electrode with larger Nions??

18

Nions=0 Nions=105 Nions=106

Ed=0.4kV/cm

Page 19: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Gain and IBF vs. VGEM(1GEM)• Ions at [0, 100um] above GEM (Ed=0.4kV/cm)• Gain and IBF vs. VGEM for various Nions

– Gain doesn’t change. IBF does, especially Nions>104

– IBF doesn’t change for Nions<104.• Good direction to the meas. (VGEM Nions IBF)

19

Nions=104

Nions=105

Page 20: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Gain and IBF vs. VGEM(1GEM)• Ions at different location above GEM

(Ed=0.4kV/cm)• IBF is drastically changed with Nions>104

• Less effective if ions are on more upper of GEM.

20

Nions=0, 102, 103, 104, 2x104

Page 21: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Gain and IBF vs. VGEM (2GEM)• 2 GEMs (Ed=0.4kV/cm, Et=3.5kV/cm). • Ions at [0, 100um] “only” above GEM2• Gain changes by 20% (not understood)• IBF changes for Nions>105

21

Nions=105

Nions=4x105

Page 22: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Gain and IBF vs. VGEM (2GEM)• 2 GEMs (Ed=0.4kV/cm, Et=3.5kV/cm). • Ions at different locations “only” above GEM2• IBF changes for Nions>5x104

– Onset depends on the underlying electric field

22

Nions=0, 103, 104, 5x104, 1x105,

1.25x105, 1.5x106

Page 23: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

More dynamical simulations(?)• So far, ions are put in [Z, Z+100um] above

GEM1/GEM2.• Make spatial ion profile for each10usec, 30usec,

60usec, 100usec steps after avalanches. • Ions are swept away from T1 quickly (40usec) and

stays above GEM1 due to lower electric field.

23

Et = 3kV/cm Ed = 0.4kV/cm Et = 3kV/cm Ed = 0.4kV/cmIon profile per one seed (Ar/CO2=70/30, Gain~1000)

electronelectron

Page 24: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

More dynamical simulations(?)• Ion spatial distribution for 10usec (100kHz) and

100usec (10kHz) separated avalanches– Many ion clouds on T1/drift space for the case of avalanches

at every 10usec. No ions in T1 for the avalanches at every 100usec.

• Make new field with these profile * Nseeds

24

Ion profile per one seed (Ar/CO2=70/30, Gain~1000)Et = 3kV/cm Ed = 0.4kV/cm Et = 3kV/cm Ed = 0.4kV/cm

electronelectron

10usec spacingfor avalanches

100usec spacingfor avalanches

Page 25: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

IBF vs. rate• Lab. test conditions (20-40 seeds/hole and ~25kHz

rate/hole)• IBF vs. time spacing between avalanches (rate/hole)• Clearer rate dependence for higher gain

– IBF gets smaller with higher rate/higher gain

25

Seed/hole=3 Seed/hole=10 Seed/hole=25

Page 26: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

IBF vs. Nseed26

30usec spacing

60usec spacing

100usecspacing

• Lab. test conditions (20-40 seeds/hole and ~25kHz rate/hole)

• IBF vs. Nseed (related to diffusion)• Clearer Nseed dependence for higher gain

– IBF gets smaller with higher Nseed/higher gain

Page 27: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

30usec spacing (30kHz) 60usec spacing (16kHz) 100usec spacing (10kHz)

Nseed=10Nseed=20Nseed=15 Nseed=20

Nseed=40

IBF vs. VGEM• Lab. test conditions (20-40 seeds/hole and ~25kHz

rate/hole)• No influence for smaller gains (HV=350)• Steep change for Nseed with higher gain• Trend is ok. But still difference in magnitude

27

Page 28: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Summary and Outlook• IBF studies have been conducted at CERN/TUM.

– 0.25% can be achievable. – More studies on rate and position (spread of seed)

dependence are on-going.• IBF simulation studies are on-going.

– Still not yet understood the discrepancy between measurements (lab. test at CERN) and simulations.

– Charge-up and space charge are accounted. – Space-charge has influence on IBF, especially N ions>104

(Ed=0.4kV/cm) and 105 (E=3kV/cm).• Clear rate / gain dependence. • Partially explain VGEM dependence of IBF measured at CERN

– More dynamical simulations• Currently ions in the space contribute “only” to the field.• Recombination with the seeds? More precious dynamics.

28

Thanks to C. Garabatos, Y. Yamaguchi, B. Katzer, V. Peskov, R. Veenhof, and all of ALICE-TPC upgrade team

Page 29: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Backup slides

29

Page 30: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Gain and IBF vs. VGEM (2GEM)• 2 GEM configurations (Ed=0.4kV/cm,

Et=3.5kV/cm). • Now, I put ions on the upper of both GEM1 and

GEM2.– Ions are put [0, 100um] above GEM1 and GEM2.– Assumption:

• Nions at GEM1 = 0.2*Nions at GEM2. – Assuming that most of the ions are from GEM2.– 0.2 = IBF of single GEM with Ed=0.4kV/cm.

– Distance between GEM1-GEM2 = 2mm• vd for Ions ~ 5um/usec. • 2mm spacing => 400usec. • If one seed come at 2.5kHz per hole, ions above GEM1 and GEM2 are distributed with 2mm spacing.

30

DZ=100um

2mm

DZ=100umGEM1

GEM2

Page 31: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Gain and IBF vs. VGEM (2GEM)• 2 GEMs(Ed=0.4kV/cm, Et=3.5kV/cm). • Ions at [0, 100um] above GEM2/GEM1• Gain changes by 20% and IBF changes for

Nions>5x104

31

Page 32: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Gain and IBF vs. VGEM (2GEM)• 2 GEMs (Ed=0.4kV/cm, Et=3.5kV/cm). • Ions at different locations above GEM2/GEM1• IBF changes for Nions>5x104

32

Nions(GEM2, GEM1)=(0,0), (104, 2x103), (5x104, 104), (105, 2x104)

[100, 200um] above GEM1/GEM2 [500, 600um] above GEM1/GEM2

Page 33: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Play with the numbers -I• Qualitatively, space-charge can explain steep

dependence of IBF vs. VGEM as seen in the measurements. – Higher VGEM higher Gain higher space-charge effects less

IBF.• Quick play with the numbers

– Gain=400 (M~800) at VGEM=400– # of seeds = 700 (22keV/30eV)– Spread due to diffusion

• 600 um for 4cm drift– (300 um/sqrt(cm))

• 5% in 100um x 100um?• 35 seed/hole?

– # of ions = 35 x 800(M) = 3x104?• This is between red and green…• Nions = 2.5x105 is unrealistic in the measurements?• Rate? (Rate=5kHz/mm2? 200usec/avalanches per hole?)

33

Page 34: Simulation study of  Ion Back Flow  for the ALICE-TPC upgrade

Movement of ions 34