héctor alvarez pol 16 december 2002 on the multiwire drift chambers alignment of the hades dilepton...
TRANSCRIPT
Héctor Alvarez PolHéctor Alvarez Pol 16 December 2002
On the Multiwire Drift On the Multiwire Drift Chambers alignment of Chambers alignment of
the HADES dilepton the HADES dilepton spectrometerspectrometer
INDEX
Part I The HADES Physics
Part II The HADES spectrometer
Part III Overview of the drift chambers alignment
Part IV Alignment using hardware methods
Part V Alignment using software algorithms
Conclusions
PART ITHE HADES PHYSICS PROGRAM
THE HADES PHYSICS PROGRAM
Heavy ion collisions at SIS energies
THE HADES PHYSICS PROGRAM
Heavy ion collisions at SIS energies
THE HADES PHYSICS PROGRAM
Heavy ion collisions at SIS energies
THE HADES PHYSICS PROGRAM
Heavy ion collisions at SIS energies
Want to know! Observables
Behavior of the High Density Phase
• Partial restoration of the chiral symmetry
In-medium change of ρ, ω, φ masses
M (MeV/c2) Γ (MeV/c2) cτ (fm)
ρ0 769.3 150.2 1.3
ω 782.6 8.4 23.4
φ 1019.4 4.5 44.4
• Equation of State (EOS)
• Astrophysics: neutron stars and inner stars structure
In-medium dilepton decays are not affected by strong interactions!
THE HADES PHYSICS PROGRAM
In-medium vector meson decay
External vector meson decay
Dilepton invariant mass spectra
THE HADES PHYSICS PROGRAM
These series of experiments, exploiting the full range of primary and secondary beams available at GSI,
are expected to make important contributions to our understanding of Quantum Chromodynamics in the
non-perturbative regime and, in particular, will provide information on the origin of hadron masses.
From The HADES Physics Program, J. Friese and V. Metag
THE HADES PHYSICS PROGRAM
PART IITHE HADES SPECTROMETER
THE HADES SPECTROMETER
HADES installation at GSI
Flat acceptance in mass and in transverse momentum
Rejection of hadronic and electromagnetic background
which could obscure the dilepton signal
Low mass materials
are chosen in all detectors and support structures to minimize the multiple scattering
A selective trigger scheme
able to accept only those events with lepton pairs, mainly those of high mass
Excellent mass resolution ( m/m 1 % (σ) at ω mass) should allow the individual
identification of the vector mesons
Capability to deal with high count rates
and large particle multiplicities
to accumulate enough significant events in a finite time
Large dilepton acceptance
(~40% for lepton pairs), required because of the tiny dilepton branching ratio, of the order of 10-5. Allows the comprehensive studies of the behavior of vector mesons
in the nuclear medium
Large dilepton acceptance
Excellent mass resolution
High count rates
Large particle multiplicities
Rejection of hadronic and em. background
Flat acceptance in mass and in mT
Low mass materials
A selective trigger scheme
THE HADES SPECTROMETER
HADES features
Large dilepton acceptance
Excellent mass resolution
High count rates
Large particle multiplicities
Rejection of hadronic and em. background
Flat acceptance in mass and in mT
Low mass materials
A selective trigger scheme
THE HADES SPECTROMETER
RICH: Ring Imaging Cherenkov Detector
RICH
THE HADES SPECTROMETER
RICH: Ring Imaging Cherenkov Detector
RICH
THE HADES SPECTROMETER
RICH
MDCs: Multiwire Drift ChambersMDCs
MDCs
THE HADES SPECTROMETER
RICH
MDCs: Multiwire Drift ChambersMDCs
MDCs
THE HADES SPECTROMETER
RICH
MDCs: Multiwire Drift ChambersMDCs
MDCsMDC features:
• High position resolution
• Two track detection ability
THE HADES SPECTROMETER
• Operation on Isobutane-Helium mixture to reduce the multiple scattering
ILSE: Superconducting Toroidal Magnet
RICH
MDCs
MDCs
ILSE
THE HADES SPECTROMETER
RICH
MDCs
MDCs
ILSE
TOF: Time-of-Flight Detectors
TOF
TOF
THE HADES SPECTROMETER
Pre-Shower: Electromagnetic/Hadronic Shower Detector With Lead Converters
RICH
MDCs
MDCs
ILSE
TOF
TOF
Pre-Shower
Pre-Shower
THE HADES SPECTROMETER
RICH
MDCs
MDCs
ILSE
TOF
TOF
Pre-Shower
Pre-Shower
TOFino: lower angle Time-of-Flight
TOFino
THE HADES SPECTROMETER
PART IIIOVERVIEW OF THE DRIFT CHAMBERS ALIGNMENT
OVERVIEW OF THE DRIFT CHAMBERS ALIGNMENT
• Revision of the momentum reconstruction methods in the spectrometer
• Simulation of the misalignment effects on the reconstructed momentum
• Analysis of the architectural design and evaluation of the technical resources
• Definition of a specific alignment scheme
Steps towards the HADES alignment system
The tracking system
OVERVIEW OF THE DRIFT CHAMBERS ALIGNMENT
Opposite directions for e - and e+
Approx. linear behavior
Dependent on the misaligned MDC
OVERVIEW OF THE DRIFT CHAMBERS ALIGNMENT
• Maximum misalignment of the MDCs (according to Physics criteria):Δy ~ 50 μm along the particle magnetic kick direction
• Also allows the determination of maximum deviation in the tilt angles
Momenta between 400 and 600 MeV/c
Electrons Positrons
MDC Δp/p (%/100μm) Δp (MeV/c/100μm) Δp/p (%/100μm) Δp (MeV/c/100μm)
I -0.21 -1.1 -0.32 -2.2
II 0.13 0.7 0.19 1.3
III 0.34 1.7 0.47 3.3
IV -0.27 -1.4 -0.37 -2.6
OVERVIEW OF THE DRIFT CHAMBERS ALIGNMENT
Simulation results
After the analysis of the architectural design and the evaluation of the allowable displacements of the support structures and other constraints, the proposed and implemented alignment scheme consist of:
• Software algorithms, based on the analysis and minimization of residuals or other functions of the hits in the drift chambers, using data samples with the magnetic field off.
• Hardware sensors (RASNIK), monitoring the relative displacements of the external MDCs with respect to the inner ones, during the data taking period.
OVERVIEW OF THE DRIFT CHAMBERS ALIGNMENT
PART IVALIGNMENT USING HARDWARE
METHODS
ALIGNMENT USING HARDWARE METHODS
RASNIK: Red Alignment System from NIKHEF
Two emitters on the external MDCs frame
Camera and lenses fixed to the internal
MDCs frame
IR light path
ALIGNMENT USING HARDWARE METHODS
2. Aperture of the lens
ALIGNMENT USING HARDWARE METHODS
Parameters of the innovative setup 1. Angle between the sensor plane and the image plane
Experimental setup
ALIGNMENT USING HARDWARE METHODS
• The resolution improves for the smallest apertures
• The resolution is practically independent of the incident angle
• For α ≥ 30°, the analysis module starts to fail
Selected setupLens aperture 15 mmAngle with sensor plane 25°
Conclusions:Resolution analysis procedure
Second order polynomial fit
Binocular design
Epoxy Carbon Fiber: KT = - 0.5x10-6 K-1
ALIGNMENT USING HARDWARE METHODS
Binocular
ALIGNMENT USING HARDWARE METHODS
Optical axis adjustments Focus adjustment
Mask and LEDs supports
ALIGNMENT USING HARDWARE METHODS
IR LEDs Matrix Mask Mount
ALIGNMENT USING HARDWARE METHODS
Stable calibration terms in different parts of the mask
Stable calibration terms for different masks
Calibration
RAHAD online monitor
• Internal raw data check• ROOT graphics facilities
EPICS Operator Screen
• Distributed monitor screens• Archiver facilities
ALIGNMENT USING HARDWARE METHODS
Complete data sample
Reduced data sample
XMDC YMDC
ZMDC
Complete data sampleσ( XMDC ) = 3.86 μmσ( YMDC ) = 4.64 μm σ( Z MDC ) = 6.88 μm
Reduced data sampleσ( XMDC ) = 1.23 μmσ( YMDC ) = 1.55 μm σ( Z MDC ) = 2.5 μm
ALIGNMENT USING HARDWARE METHODS
Resolution estimation
Correlation with the magnetic field
Correlation with the temperature of the MDC frames
ALIGNMENT USING HARDWARE METHODS
Experimental results
PART VALIGNMENT USING SOFTWARE
METHODS
ALIGNMENT USING SOFTWARE METHODS
Coordinate transformations
BBB
AAA
TXRX
TXRX
MDC to Lab:
VXMX AB
MDC to MDC:
A1
B RRM
BA1
B TTRV
where
221202
211101
201000
MMM
MMM
MMM
M
and
for instance
sinsincoscoscosM00
Variables: ))y(S),x(S,y,x(X
Hit compatibility and sample selection
)X(V)X(
21
expV)2(
1)X(f 1T
2
Probability density function:
4
)y(Sy
2ySy
2)y(S
2
2y
2
2ySy)x(Sx
2xSx
2)x(S
2
2x
2
2xSx
)y(Sy2)y(Sy
11
)x(Sx2)x(Sx
11
Equiprobability volume (hyperellipsoids on α4):
ALIGNMENT USING SOFTWARE METHODS
Three MDCs alignment algorithm
Then, minimize
with:
should be zero for each track.
Tracks
Q)(sin
)(sin22
222
)y(yQ
)x(xQ
)y(yQ
)x(xQ
)y(yQ
)x(xQ
)(sin
C2
2
CC2
2
CB2
2
B
B22
BA2
2
AA2
2
A22
ba
basin
2A02z
A01y
A00x
222
xA01y
A00z
zA00x
A02yy
A02z
A01x44A
baMaMaMabbabMbMba
bMbMbabMbMbaba
2xQ
where, for instance:
a
b
BCBCBC
zyx
BABABAzyx
zz,yy,xxb,b,bb
zz,yy,xxa,a,aa
A B
C
ALIGNMENT USING SOFTWARE METHODS
If one parameter is fixed to the correct value
The problem reduces to find out a set of histograms which univocally defines thecorrect value of the fixed parameter.
ALIGNMENT USING SOFTWARE METHODS
Convergence inside the allowable error
Below 50 μm
Simulation results
How to fix the angular parameter
ALIGNMENT USING SOFTWARE METHODS
b b
a
b
c c
c
a a8.6x10-5
2.06x10-3
-1.93x10-3
Abscissa for y=0-6.6x10-5
November 2001 alignment: three MDCs algorithm
ALIGNMENT USING SOFTWARE METHODS
• The uncertainties in the calibration procedures and hit fitting tasks lead to hits with incompatible slopes on the MDCs.• As a consequence, the uncertainty intervals for the alignment results in Nov01 are slightly larger than expected (~100 μm for MDCs I-II, ~300 μm for MDCs II-III).
Differences in mrad
Two MDCs alignment
ALIGNMENT USING SOFTWARE METHODS
Minimization of the residuals:
2
)y(xS
2
)y(yS
2
)x(xS
Tracks
2
)y(S
2
)x(S
2
y
2
x
2
)y(SxW
2)y(Sy
W2
)x(SxW
2
)y(SW
1)x(S
W1
yW1
xW1
Q
Analytical minimization with respect to the components of the
translation vector:
0W
)y(S2W
)y(S2W
)x(S2Wy2
Wx2
VQ
0W
)y(S2W
y2VQ
0W
)y(S2W
)x(S2W
x2VQ
i )y(xS
xii
)y(yS
yii
)x(xS
xii
y
yii
x
xii
2
2
i )y(yS
i
y
i
1
2
i )y(xS
i
)x(xS
i
x
i
0
2
The solution is the relative translation
vector V=(V0,V1,V2)
ALIGNMENT USING SOFTWARE METHODS
Geometrical determination of the relative rotations,
for instance, in-plane rotations:
Two MDCs alignment
cosysinx'y
sinycosx'x
θ
ALIGNMENT USING SOFTWARE METHODS
Two MDCs alignment
Iterative approach to the solution:
1. Sample selection2. Analytical minimization of the
translation (vector V)3. Geometrical correction of the
rotation (rotation matrix M)
Below 50 μm
Simulation results
Convergence inside the allowable error
ALIGNMENT USING SOFTWARE METHODS
The Target Finder algorithm1. Analytical minimization of:
2. Iterative approach to the solution using bi-squared Tukey
weights
)z,y,x(dwQ ttt2i
ii
2
ALIGNMENT USING SOFTWARE METHODS
November 2001 alignment: two MDCs algorithm
Mean:-7.8x10-3
Mean:-4.5x10-3
ALIGNMENT USING SOFTWARE METHODS
Beam line reconstruction after alignment
Beam line (Z)
Track
ρ
θ
ALIGNMENT USING SOFTWARE METHODS
November 2002 “Last minute” result
Double target reconstruction
20mm
Very preliminary alignment
CONCLUSIONS
In this work, several tools and methods have been developed to obtain the relative alignment of the Multiwire Drift Chambers (MDCs), the main tracking detectors in the
HADES spectrometer.
• In a first step, the requirements on the resolution in the reconstructed momentum and the invariant mass of the lepton pair, have been expressed as maximum deviations in the knowledge of the relative displacements and rotations of the MDCs.
• A set of RASNIK devices has been considered as optimal solution for the hardware monitoring and a specific RASNIK configuration has been developed.
The influence on the resolution of both the light incidence angle onto the camera and the lens aperture have been studied.
CONCLUSIONS(1)
• The implementation of the RASNIK devices in the spectrometer has required the design of custom-made pieces. This task has been accomplished from the mechanical design of all pieces up to the final installation in the spectrometer.
•A complete monitoring program (RAHAD) has been developed. It performs a data calibration and transformation, according to the coordinate systems of the MDCs, as well as the interface with the EPICS “HADES Slow Control System”.
• Once the RASNIK setup was installed on the spectrometer, its performances below the requirements were confirmed.
The RASNIK results have been successfully correlated with temperature changes and with the magnetic field forces. The RASNIK monitoring results have been used to correct the alignment parameters obtained by software methods.
CONCLUSIONS(2)
• Regarding the software methods, several iterative algorithms have been developed in order to obtain the relative alignment parameters between MDCs.
Two different algorithms has been developed, for those sectors with three or two MDCs.
The use of the “Two MDCs algorithm” includes the determination of the target position, implemented in the so-called “Target Finder” algorithm. The “Three MDCs algorithm” has been chosen as the main method to obtain the position parameters.
• The different algorithms have been first tested under simulation, checking their convergence to the correct parameters. The errors have been estimated and the resolution in the determination of the relative alignment parameters fulfils the requirements.
• A set of data has been analyzed (Carbon beam at 1 GeV on a Carbon target, November 2001 run) using the alignment algorithms. The alignment parameters have been estimated, including their uncertainty intervals.
CONCLUSIONS(3)