TWIST

1
TWIST

• The TRIUMF Weak Interaction Symmetry Test

Goal A high precision measurement of the
positron (Te) and (cos?e) spectrum from ?-decay
to test the SM predictions for the weak
interaction
2
Outline

• Physics motivation
• Discovery potential for TWIST
• Experimental method
• Analysis approach
• Systematics studies
• Timeline

3
(No Transcript)
4
TWIST physics motivation -- test the Standard
Model for ?-decay
Most general interaction does not presuppose
the W

• S,V,T scalar, vector or tensor interactions
• R, L right and left handed leptons (e, m, or
  t )

5
Couplings in the present Standard Model
6
Current measured couplings --
7
e spectrum in x, cos?e
Spectral shape in x, cos?e is characterized in
terms of four parameters -- ?, ? , ?, ?
P? is the muon polarization
(L. Michel, A. Sirlin)
8
e spectrum in x, cos?e
? is being measured at PSI
9
Current status --
TWIST will measure ?, ?, ? in two steps --
10-3 in 2004 3x10-4 in 2005/6
10
500 times TWIST sensitivity
Spectral effects with changes in ?, ?, ?, ?
11
Search for deviations from SM --
12
Search for deviations from SM --
and also for ?
13
Chirality of the muon decay
14
Coupling to right-handed muons
15
(No Transcript)
16
Left/Right Symmetric Model
Two weak bosons with mass eigenstates M1 and M2
? Left/Right mixing angle
17
(No Transcript)
18
The TWIST program

• Collect high precision data to obtain the e
  spectrum from ?-decay as a function of x and
  cos?e
• Detailed study of systematic errors in TWIST
• Extract the best values of the spectral
  parameters ?, ? , ?, ? simultaneously (the
  first time this has been done)
• Obtain a precision in ?, ?, ? (a) of 10-3 and
  (b) a few parts in 104 (10-3 precision for ?)
• Compare ?, ? , ?, ? from our fit with Standard
  Model values

19
Obtain high precision data on the e spectrum
Highly polarized ?
? stop in Al target(several kHz)
Unbiased ? (scintillator) trigger
20
Chambers half detector
Planar drift chambers sample positron track
Use 44 drift planes, and 12 PC planes
21
(No Transcript)
22
(No Transcript)
23
Typical decay event
24
TWIST Data

• High data rates (few kHz) ? Data sets of 109
  muon decay events in two weeks
• TWIST is systematics limited. (High data rates
  and computational resources are essential for
  studying systematic effects.)
• In 2002-03, 6 x 109 muon decay events on tape.
• Standard data set 300M triggers ? 58M useful
  events (smaller samples for some systematics
  studies)

Reconstructed muon decay spectrum
25
Determination of ?, ?, ?, ?
Accelerator data are collected
Monte Carlo data are generated

• Event Analysis
• Event classification (31 types)
• Helix fit to events within fiducial volume
• Extract e momentum and angle spectrum in
  bins of x and cos?e

Fit
Accelerator data spectrum
Monte Carlo data spectrum
??, ??, ??, ??
26
Determination of ?, ?, ?, ?
Data spectrum is fit to Monte Carlo spectrum --
From the fit ??, ??, ??, ?? are determined. Blind
Analysis ?o, ?o, ?o, ?o are generated randomly
(once) and remain hidden until the end of the
experiment.
27
Evaluating Systematic Errors
Accelerator data are collected
Accelerator data are collected

• Event Analysis
• Event classification (31 types)
• Helix fit to events within fiducial volume
• Extract e momentum and angle spectrum in
  bins of x and cos?e

Fit
Accelerator data spectrum
Accelerator data spectrum
??, ??, ??, ??
28
Evaluating Systematic Errors
Monte Carlo data are generated
Monte Carlo data are generated

• Event Analysis
• Event classification (31 types)
• Helix fit to events within fiducial volume
• Extract e momentum and angle spectrum in
  bins of x and cos?e

Fit
Monte Carlo data spectrum
Monte Carlo data spectrum
??, ??, ??, ??
29
Evaluating Systematic Errors
Monte Carlo data are generated
Monte Carlo data are generated

• Event Analysis
• Event classification (31 types)
• Helix fit to events within fiducial volume
• Extract e momentum and angle spectrum in
  bins of x and cos?e

• Event Analysis
• Event classification (31 types)
• Helix fit to events within fiducial volume
• Extract e momentum and angle spectrum in
  bins of x and cos?e

Fit
Monte Carlo data spectrum
Monte Carlo data spectrum
??, ??, ??, ??
30
Evaluating Systematic Errors

• Methodology
• Exaggerate possible sources of systematic error
  -
• Take accelerator data sets under a different
  conditions
• Generate Monte Carlo runs with different
  settings
• Analyze same data with different calibrations
• (Use full (or nearly full) data set for each
  test)

• Fit two data sets measure the effect on ??, ??,
  ??, ??
• Scale the effect by the exaggeration factor

31
Evaluating Systematic Errors
Examples

• Chamber gas density muon stopping distribution
• Different magnetic field energy calibration
• Magnetic field shape
• Alignments
• Beam properties
• Detector response
• STR HV, drift cell geometry
• Efficiency, fiducial region
• Resolution
• Cross talk
• TWIST simulation (GEANT) and more

32
Evaluating Systematic Errors
Present Status
Preliminary
and many more at this level
No show-stoppers!
33
TWIST Simulation Validation
Accelerator data are collected
Monte Carlo data are generated
Monte Carlo data are generated

• Event Analysis
• Event classification (31 types)
• Helix fit to events within fiducial volume
• Extract e momentum and angle spectrum in
  bins of x and cos?e

Monte Carlo data spectrum
Monte Carlo data spectrum
34
TWIST Simulation Validation
Accelerator data are collected
Accelerator data are collected

• Event Analysis
• Event classification (31 types)
• Helix fit to events within fiducial volume
• Extract e momentum and angle spectrum in
  bins of x and cos?e

Accelerator data spectrum
Monte Carlo data spectrum
Accelerator data spectrum
35
TWIST Simulation Validation

• Test the simulation independently of ?, ?, ?, ?
• Take accelerator data sets under a different
  conditions
• Generate Monte Carlo data sets with same
  conditions
• Analyze data sets with the same analysis package
• Compare the differences --

• Determine the sensitivity for each
  physics/detector effect in ??, ??, ??, ??

36
TWIST Simulation Validation
Examples

• Chamber gas density muon stopping distribution
• Different magnetic field energy calibration
• pmax vs ?e
• ?2 and confidence level distributions
• hits per plane
• muon stopping distribution
• delta production cross-section
• energy loss
• multiple scattering
• and more

37
Compute power - WestGrid

• At University of British Columbia
• 504 dual-3Ghz Xeon nodes
• 10 TB global disk storage
• Robot tape archiving system
• Many tens of 108 events analyzed
• Many tens of 108 events simulated analyzed
• 70 ms/event (simulation)
• 30 ms/event (reconstruction)
• gt5000 CPU days used
• (www.westgrid.ca)

Funded by the Canada Foundation for Innovation,
Alberta Innovation and Science, BC Advanced
Education,and the participating research
institutions.
38

• 2004
• Data in hand for measurement of ?, ? to 10-3
• Study of systematic errors (for 10-3) nearly
  complete
• Publish measurement of ?, ? at 10-3 in 2004.
• Take data for measurement of P?? - for precison
  of 10-3 - publish 2004/05
• 2005/06
• Take data for measurement of ?, ? , ?, ? to a
  precision of a few parts in 104 (10-3
  precision for ?)
• BOTTOM LINE Compare ?, ? , ?, ? from our fit
  with Standard
  Model values ? New Physics?