Manchester Status

1
Manchester Status
Stefan Soldner-Rembold Steve Snow Ray
Thompson Nasim Fatemi-Ghomi Irina Nasteva -
new RA
Ray Thompson /Steve Snow
Supernemo meeting MSSL 4-6 April 06
2
SuperNEMO preliminary design
Plane and modular geometry 5 kg of enriched
isotope per module
1 module source (40 mg/cm2) 4 x 3 m2 Tracking
volume drift wire chamber in Geiger mode, 3000
cells Calorimeter scintillators PMTs
20 modules 100 kg of enriched isotope
60 000 channels for drift chamber
20 000 PMT if scint. block 2
000 PMT if scint. bars
4 m
1 m
5 m
1 m
Side view
Top view
3
3cm f 2.7m
Basic Nemo3 Geiger Cell
9 cell prototype aims - Technology Transfer
- understand Nemo 3 cell parameter space
Gas composition He 1Ar 4 Ethanol 1.
Extension of cell to 4m Efficency of Geiger
propagation Plateau width? 2. Ageing - high
gain small fraction continuous afterpulsing
3. Reduce scattering - Wire
diameter - looks unlikely -
Cell size optimisation why 3cm ? 2 track
resolution - channel count 4. Test ideas for
modular cell construction/wiring ultrasonic
welding to avoid crimps ??
4

• Possible end cap concepts
• Classic large machined end plates
• - individual crimps
• - well understood
• - difficult to wire
• - large
• - dont really need the mechanical
• accuracy
• - hard to test
• Modular cells
• - cell units wired individually
• - stored on simple frames
• - tested in small units
• - shipped to final assembly site then
• mounted on big frame
• Plastic moulded endcap

5
Endcap molding front
6
ENDCAP MOLDING BACK
7
Cell
Cell stacking
8
Mechanics

• Gathering tools
• New 150m2 clean room suite now ready can start
  wiring
• Kit of parts to construct test cells
• Clean gas tanks 2 x 2m x 200mm f
• Or 1 x 4m
• Up to 5x 5 cells
• Endplates to suit cell designs
• Clean gas system
• revamped gas microstrip system ethanol
  cooler

9
Software tools
MAGBOLTZ (Steve Biagi NIM A 421 1999 234-240) is
a programme that calculates properties of
electron drift in most wire chamber gasses and
their mixtures. I have got a copy and tried it on
the Nemo 2 and 3 gasses. FlexPDE is a programme
that solves systems of differential equations
using the finite element method. I am already
familiar with it, so I have tried it on the 2D
electrostatics of the Nemo 3 cell. It also works
in 3D but I have not tried this yet. GARFIELD is
well known wire chamber simulation software,
already used for Nemo 3. It includes interfaces
to MAGBOLTZ and other useful tools. It solves 2D
electrostatics of wires and planes, but no 3D.
10
MAGBOLTZ
Actual velocity plateau measured in Nemo is about
20 lower than MAGBOLTZ prediction
Calculates drift velocity, diffusion and Townsend
a, starting from electron-molecule collision
cross sections and the Boltzmann transport
equations. Examples ? ?
11
FlexPDE
Used to simulate this wire layout ? Minimal cell
to represent 3-layer plane. Symmetry boundary
conditions on dashed line. Value boundary
conditions in coloured areas.
? FlexPDE output example. Potential map, contours
from 60 to 400 V Anode at 1840 V, all wires 50
micron.
12
Interpretation of potential maps
The thing that matters is the E field strength
near to ( lt 1mm ) the anode wire. In this region
the field is extremely close to E A/r, where A
is a constant to be determined by the
simulation. Instead of plotting A directly, I
interpret it as the equivalent voltage on a 50
micron wire at the centre of a 30 mm tube
Some contours field is slightly stronger on one
side of the wire than the other. Effective
voltage is 1605 V.
No contours field is very close to 1/r.
Effective voltage is 1560 V
Conclusion middle plane needs to be run 45 V
higher to give the same gain as edge
planes. Becomes 20 V when extra ground wires are
added.
13
Nemo 3 test cell equipotentials
14
Comparison with other experiments
There are plenty of other helium based drift
chambers out there. None use Geiger mode
so comparisons with Nemo may be unfair but ...
Other experiments find it necessary to have
Øcathode few Øanode , otherwise high field
on cathode surface causes continuous discharge.
Nemo has unusually thick anode and equal cathode.
Survives because of low rate ? Improved stability
or transparency if anode could be reduced ? Nemo
has unusually low proportion of quencher ? good
X0. But a further compromise of 50 on X0 would
allow use of Kloe gas and proportional mode.
Worth investigating for Supernemo ?
15
Something I don't understand
The very tight tolerance on the wire diameter, 50
microns 1, "is necessary to ensure proper
plasma propagation."
There should be a simple relation between wire
diameter and operating voltage that give
equivalent gain. Use Magboltz value of a vs E and
basic electrostatics to give E vs radius for wire
in centre of Ø30 mm tube. Integrate a dr to get
total gain
1 on wire diameter corresponds to 1.3 V on wire
potential
15 on wire diameter corresponds to 20 V on wire
potential
16
Another thing ...
In Nemo 3 the sagitta of tracks bending in the
magnetic field is only a factor 2 greater than
the sagitta error caused by multiple scattering.
Naive calculation using 50 cm electrons ? to 25
G field ?
In SuperNemo, could we make better use of the
good intrinsic resolution of the tracker ( 0.4
mm ) by using a higher magnetic field and shorter
path length ? Better rejection of wrong-sign
tracks. Independent measurement of electron
energy. Compromises rejection of crossing
tracks by timing. Reduces thickness of module.
17
Summary

• Various tools / components in place
• New RA Irina 100
• Can now start serious work..
• Set up test cells
• - Test endcap designs---
• Simulations - Electrostatics
• - Physics optimisation,
  multiple scattering
• 2 track resolution
• Timescale

18
Geiger mode
Steve Snow 3/4/06

• trying to -
• Understand the self-limited Geiger mode in
  general,
• Collect the software tools that will be
  necessary to understand our test cells,
• Understand how Nemo 3 has been optimised,
• Think about optimising SuperNemo.

19
Self limited Geiger mode

• I waded through the series of papers by
  D.H.Wilkinson.
• As far as I understand, he says there are just
  two key parameters that control the plasma
  propagation
• Gain excluding feedback the average number of
  secondary electrons produced by one electron
  drifting in to the anode from 'far' away.
• Photon feedback probability the probability that
  a UV photon is produced in the avalanche and then
  travels far away from the anode and liberates a
  photoelectron in the gas. Assumed proportional to
  the number of secondary electrons.

Neither parameter is directly measurable,
but Gain can be measured at lower voltage and
extrapolated upwards, or else simulated with
existing tools (later slide). The onset of Geiger
mode is when gain x feedback gt 1. We
can measure this and infer the feedback
probability per secondary electron.