The SuperNEMO Tracker Manchester status

1
The SuperNEMO TrackerManchester status
Steve Snow Ray Thompson Stefan Soldner-Rembold
Irina Nasteva James Mylroie-Smith Nasim
Fatemi-Ghomi
2
Outline

• Electrostatic simulations of Geiger cells
• Comparison between Garfield and FlexPDE
• Results for 9-cell prototype layouts
• Results for different layouts of the SuperNEMO
  tracker
• To do
• Construction of 9-cell prototype
• Status of first 9-cell prototype
• Status of second 9-cell prototype
• Single Geiger cell

3
Electrostatic simulationsof Geiger cells
4
Simulations of 3x3 cells

• The 9-cell prototype is simulated with
• X pitch 30 mm
• Y pitch 30 mm
• Gap 10 mm
• Cathode diameter 50 mm
• Anode diameters 50 and 30 mm
• Possible layouts
• Basic octagonal cells
• Octagonal cells with 4 extra wires around mid
  cell
• Octagonal cells with 4 extra wires around all
  cells

ground plane
extra cathodes
5
Garfield and FlexPDE

• Garfield
• electrostatic simulations of wire chambers in 2D
• makes use of symmetries
• can simulate gases with Megaboltz
• used and tested in many gas detector simulations
  (NEMO3)
• FlexPDE
• finite element analysis
• user supplies differential equations to be
  solved (programme knows nothing about the
  physics)
• can do simulations in 3D
• easy to use

6
Applied and effective voltages
In a wire chamber we have -
An arrangement of wires with voltages applied to
them.
A resulting field distribution that can be
calculated with Garfield or FlexPDE.
Very near the wires, the field always has the
form EA/r. Equivalently, the potential contours
are circles centred on the wire.
It is the strong E field within 1.5 mm of the
anode wire that determines the avalanche gain,
which in turn drives the Geiger plasma
propagation. It is the strong E field at the
surface of the cathode wire that can drive
electron emission processes, leading to
self-sustained discharge.
So the electrostatics of a wire chamber is
characterised by the A values near each of the
wires. Instead of quoting A directly, we usually
convert it to the effective voltage the voltage
necessary to produce the same value of A when the
wire is in the centre of a 30mm tube Veff ?
A/r dr A ln( rtube /rwire )
7
Gain versus Voltage and Anode radius
To compare layouts with different anode diameters
we need to know how the Townsend coefficient a
varies with E. We used predictions from
Magboltz for the NEMO-3 gas mixture. The
avalanche gain is given by integration of a(E) in
the high field region Gain exp( ? a(A/r).dr )
The result of the integration for 30 and 50
micron wire diameters, at a range of effective
voltages, is shown in this plot. This shows that
a 50 micron wire with Veff 1700 V will give the
same gain as a 30 micron wire with Veff 1654 V.
8
Basic octagonal 3x3 cells - results
On the anodes we show the applied voltage,
necessary to produce a gain equivalent to 1700 V
on a 50 micron wire in a 30 mm tube. On the
cathodes we show (-1x) the effective voltage.
50 micron anodes 30 micron anodes
9
Octagonal4 (mid cell) - results
On the anodes we show the applied voltage,
necessary to produce a gain equivalent to 1700 V
on a 50 micron wire in a 30 mm tube. On the
cathodes we show (-1x) the effective voltage.
50 micron anodes 30 micron anodes
10
Summary of 3x3 cells results

• Garfield and FlexPDE agree to within 0.4
• We can go on to use FlexPDE for 3D simulations
  (wire ends)
• Adding extra cathodes around mid cell reduces
  Veff on cathodes
• Decreasing anode diameter to 30 mm gives a
  higher gain at a given voltage

11
SuperNEMO module assumptions

• We assume that
• There will be a continuous block of Geiger cells
  filling nearly all the space between the source
  foil and the calorimeter.
• All cells have the same layout except for
  possible minor variations on the surface layers.
• The space between foil and scintillator must be
  gt30 cm for TOF to work. But total module
  thickness should be kept down.
• The structure will be 9 cells deep in the X
  direction and very large in the Y direction. So
  the unit cell for electrostatics is the pink
  area.
• X pitch and Y pitch need not be identical.

12
Octagonal layouts

• Simulated with
• X pitch 30 mm
• Y pitch 30 mm
• Gap 10 mm
• Cathode diameter 50 ?m
• Anode diameters 50 and 30 ?m

13
Hexagonal layouts

• Simulated with
• X pitch 30 mm
• Y pitch 30 mm
• Gap 10 mm
• Cathode diameter 50 ?m
• Anode diameters 50 and 30 ?m

14
Basic octagonal layout - results
On the anodes we show the applied voltage,
necessary to produce a gain equivalent to 1700 V
on a 50 micron wire in a 30 mm tube. On the
cathodes we show (-1x) the effective voltage.
15
Octagonal2 layout - results
On the anodes we show the applied voltage,
necessary to produce a gain equivalent to 1700 V
on a 50 micron wire in a 30 mm tube. On the
cathodes we show (-1x) the effective voltage.
50 micron anodes
Field lines are no longer shared equally between
these four cathodes. We could benefit by
increasing the separation of the closest pair.
30 micron anodes
Edge effects are negligible except for the last
cell
16
Octagonal4 layout - results
On the anodes we show the applied voltage,
necessary to produce a gain equivalent to 1700 V
on a 50 micron wire in a 30 mm tube. On the
cathodes we show (-1x) the effective voltage.
17
Hexagonal4 layout - results
On the anodes we show the applied voltage,
necessary to produce a gain equivalent to 1700 V
on a 50 micron wire in a 30 mm tube. On the
cathodes we show (-1x) the effective voltage.
50 micron anodes
30 micron anodes
18
Hexagonal6 layout - results
On the anodes we show the applied voltage,
necessary to produce a gain equivalent to 1700 V
on a 50 micron wire in a 30 mm tube. On the
cathodes we show (-1x) the effective voltage.
50 micron anodes
30 micron anodes
19
Figures of merit
We want wires/cm to be small for
transparency, cathode Veff to be small for
stability. The dominant parameter is
cathodes/cell, followed by anode diameter, then
Hex/Oct. We choose the octagonal as our baseline
design.
20
To do

• Geiger cells
• Simulate the NEMO3 cell to give an estimate of
  tolerable cathode Veff.
• Check a handful of points on the gain versus
  voltage and anode diameter plots by operating a
  test cell in proportional mode.
• Check whether Geiger propagation depends only on
  gain, as assumed, or whether it has some extra
  dependence on wire diameter.
• Find out experimentally the highest tolerable
  cathode surface field a) on a fresh wire,
  b) after some ageing.
• Physics simulation
• Are 40mm cells are acceptable for two-track
  resolution?
• Which of the following have most influence on
  acceptance of 0vbb events? energy loss or
  multiple scattering, in the gas or in
  wires, wire length, source foil area,
  foil-to-scintillator distance,
• This was partially studied by Darren Price,
  could be a new student project

21
Construction of 9-cell tracker prototype
22
First 9-cell prototype

• 3x3 cells (as in simulation)
• X,Y pitch 30 mm
• Length 2 m
• Cathode diameter 50 mm
• Anode diameter 50 mm
• Wires from NEMO3
• Gas system, He-Ar, ethanol cooler
• Trigger system for cosmics 2 scintillators in
  coincidence

23
Status of the first 9-cell prototype
Prototype was wired
24
Status of the first 9-cell prototype
and closed in the vacuum vessel
25
Second 9-cell prototype
Based on Forget concept of separate stackable
cells
26
Some alterations to allow prototype to be
fabricated by CNC machining rather than molding.
Wireclamps screwed rather than ultrasonic welding.
27
Single Geiger cell

• A single Geiger cell was constructed to study
  plasma propagation
• Single anode inside a tube
• Diameter 26 mm
• Length 3 m

28
Single cell tests

• He-Ar gas mixture, no alcohol yet
• Trigger on 2 scintillators
• We have seen the first signals

29
Status Summary
Single long tube - Pulses. (simulation
reference) First 9-cell prototype - Wired, in
clean vacuum vessel. conventional crimp
design awaiting cleaning of gas
piping. ready to switch on after Dubna
meeting. Second 9-cell prototype Most of endcap
components CNC Forget concepts machined.
Awaiting side closure pieces. 2nd vacuum
vessel ready. Need to build wired cell
carrier. Readout Currently using a scope and
LabView. Need a multichannel ASIC readout
card (LAL). We have bid for H1 ADC boards
after decommissioning (2007).