Systematic Errors Studies in the RHIC/AGS Proton-Carbon CNI Polarimeters Andrei Poblaguev Brookhaven National Laboratory The RHIC/AGS Polarimetry Group: I. Alekseev, E. Aschenauer, G. Atoian, A. Bazilevsky, A. Dion, H. Huang, Y. Makdisi, A.Poblaguev, W. Schmidke, D. Smirnov, D. Svirida, K. Yip, A. Zelenski 1 PSTP 2011, St. Petersburg 06/18/22
21
Embed
Systematic Errors Studies in the RHIC/AGS Proton-Carbon CNI Polarimeters
Systematic Errors Studies in the RHIC/AGS Proton-Carbon CNI Polarimeters. Andrei Poblaguev Brookhaven National Laboratory The RHIC/AGS Polarimetry Group: I. Alekseev, E. Aschenauer, G. Atoian , A. Bazilevsky , A. Dion, H. Huang, - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Systematic Errors Studies in the RHIC/AGS Proton-Carbon CNI Polarimeters
Andrei Poblaguev
Brookhaven National Laboratory
The RHIC/AGS Polarimetry Group:I. Alekseev, E. Aschenauer, G. Atoian, A. Bazilevsky, A. Dion, H. Huang,
Y. Makdisi, A.Poblaguev, W. Schmidke, D. Smirnov, D. Svirida, K. Yip, A. Zelenski
1PSTP 2011, St. Petersburg04/20/23
PHENIX (p)
AGS
LINACBOOSTER
Pol. H- Source
Solenoid Partial Siberian Snake
200 MeV Polarimeter
Helical Partial Siberian Snake
Spin Rotators(longitudinal polarization)
Siberian Snakes
Spin Rotators(longitudinal polarization)
Strong AGS Snake
RHIC pC PolarimetersAbsolute Polarimeter (H jet)
STAR (p)
BRAHMS(p)
AGS pC Polarimeter
Spin flipper
Layout of the RHIC facility
• H jet (pp) polarimeter provides absolute polarization measurements at RHIC• RHIC pC polarimeters provide polarization monitoring including polarization profile measurements• AGS pC polarimeter provides polarization monitoring (mainly used for technical control and special beam studies)
2PSTP 2011, St. Petersburg04/20/23
Proton-Carbon Polarimeter kinematics
Plan view
Event selection in RHIC/BNL pC polarimeters:
3PSTP 2011, St. Petersburg04/20/23
Polarization MeasurementSpin dependent amplitude:
Rate in the detector:
1. Spin Flip (one detector):
2. Left-right asymmetry (two detectors)
Square-root formula:
Combining “spin flip” and “left/right asymmetry” methods allows us to strongly suppress systematic errors
A theoretical model for AN(t) (a fit to the BNL E950 data)
4PSTP 2011, St. Petersburg04/20/23
AGS CNI Polarimeter 2011
PSTP 2011, St. Petersburg 5
1,8 - Hamamatsu, slow preamplifiers
2,3,6,7 - BNL, fast preamplifiers
4,5 - Hamamatsu, fast preamplifiers
3 different detector types:
Larger length (50 cm)
Regular length (30 cm)
Silicon Strip Detectors:
90 degree detectors (2,3,6,7)
45 degree detectors (1,4,5,8)
Strip orientationDead Layer
04/20/23
Schema of Mesurements
PSTP 2011, St. Petersburg 6
WFD α-source measurements (241Am , 5.486 MeV)
“Banana fit” t-t0 = tA(xDL,αA)
04/20/23
An example of data selection
PSTP 2011, St. Petersburg 7
Wrong determination of mean time
It must be a vertical line if detector is properly calibrated
If t0 is known, a model independent calibration can be done
04/20/23
The AGS pC polarimeter is succesfully used for the relative measurements
Study of Polarization dependence on beam intensity
8PSTP 2011, St. Petersburg04/20/23
Is absolute polarization measurement possible with a proton-Carbon polarimeter ?
A systematic errors study is necessary to answer this question.
• Are results dependent on detector configuration ?• Do we know the Analyzing Power AN(t) ?• Could we properly calibrate detectors ?• Do we understand energy losses in the target ?• Can we control rate dependence of polarization measurements ?• …
9PSTP 2011, St. Petersburg04/20/23
10PSTP 2011, St. Petersburg
Polarization measured by all 3 types of detectors is consistent within 1-2% accuracy !
Can we explain slope difference for 90 and 45 degree detectors by rate effect ?
All 2011 data was included in the fit. Results of the fit
should be used for comparison only
Polarization, P(1.2) , is given for intensity 1.2×1011
Polarization vs Beam Intensity (Late CBM),Vertical Target3, all 2011 runs
Polarization dependence on detector type
04/20/23
Hamamatsu (45 degree) vs. BNl (90 degree) detectors
PSTP 2011, St. Petersburg 11
Polarization dependence on detector type
No visible variations of the polarization ratio during 4-month Run 2011!
04/20/23
AN measurement for assumed 65% polarization
PSTP 2011, St. Petersburg 12
Analyzing Power AN(t)
• Poor consistency between theory and measurements• Wrong energy calibration and energy losses in the target may contribute to the discrepancy• Results depend on the target (rate ?, energy losses ?)
Potentially, analyzing power may be measured by the pC polarimeter (up to a normalization constant)
04/20/23
Dead-Layer corrections
Stopping range parametrization:
“standard parametrization”, p=1/dconstant energy loss, p=Eloss polinomial
Carbon Energy from measured amplitude:13PSTP 2011, St. Petersburg
Enrgy Calibration
04/20/23
L0 is stopping range derived from MSTAR dE/dx (used in “standard” calibration)
Inverse task:If E(αA) is known then we can determine L(E) and dE/dx
If t0 is know then we can measure Carbon energy as a function of the amplitude αA
and thus we can measure dE/dx (in deadlayer length units)WARNING: In such a way we measure effective dE/dx which may be different from ionization losses dE/dx.
If t0 is unknown we can make a fit, that is to try all possible t0 and select one which provides best data consistency. It might provide us with value of t0 and calibration of the measured amplitude ECarbon = E(αA) .
WARNING: the fit may work incorrectly if parameterization of stopping range L(p, αA) can not approach well true effective dE/dx.
14PSTP 2011, St. Petersburg
A model independent calibration of the amplitude
Enrgy Calibration
04/20/23
New calibration method vs standard one
• The function L(E) = p0L0(E) + p1L02(E) fits data much
better then “standard” calibration function p0L0(E)• Significant difference in the value of t0
• Significant difference (up to 15% ) in the energy scale
Better fit of data does not guarantee better calibration !
15PSTP 2011, St. Petersburg04/20/23
Comments about t0 determination in the fit
Including t0 to the fit:
(τ is time of flight for 1 MeV carbon )
If then (good calibration)
However, if may be approximated by variations of the
then result of the calibration is unpredictable
may be masked by faked correction
16PSTP 2011, St. Petersburg04/20/23
More realistic example is rate of good events is total DAQ rate
Simplified exampleOnly one carbon signal may be taken by the DAQDetection efficiency:
where r is average rate per bunch.
An estimate of the rate effectRate effect
- is a strip pair number - is average rate per strip (millions events per spill) - is rate in strip i (events per bunch), n = 0.0528 - is relative rate in the strip I
assume factor k is the same for all strips
Rate contribution
Machine contribution
17PSTP 2011, St. Petersburg04/20/23
Polarization dependence on beam intensity (averaged over all 2011 runs) :
The measured value of the rate effect factor
agrees well with a pileup based estimate
Vertical Target3, all 2011 runs: Strip Pairs
18PSTP 2011, St. Petersburg
Rate effect
04/20/23
Target dependence of the Polarization measurements
PSTP 2011, St. Petersburg 19
AGS pol., during H-jet meas. at injection
Intensity -1.5
Polarization vs intensity, Horiz. target #1, JQ-on Polarization vs intensity, Vertical target#3, JQ-on
• Slope difference is consistent with our estimates• We can explain 4±1 % of polarization difference by rate effect. Where the rest 4.6±1.7% come from?
Enrgy losses in the target
04/20/23
Energy losses in the target
20PSTP 2011, St. Petersburg
φ Target
Beam
Energy range 400-900 keV
Calculation
AngleTarget Thickness (μg/cm2)
4 8 16
0 0.991 0.982 0.965
45 0.987 0.975 0.951
80 0.950 0.903 0.825
85 0.903 0.802 0.610
0 - 360 0.970 0.948 0.911
Measured/True Polarization
Results are independent on target width !
125 μm target
Effect of energy losses in the target• may be significant• may be unpredictable
Enrgy losses in the target
04/20/23
dE/dx
AN(t)
(d ~ 30 nm)
Summary
• Different types of detectors were tested in the Run 2011• Results of polarization measurements were consistent within 1-2% accuracy• No significant variation of the results of measurements were observed during the whole 4 month run.• The polarimeter has a capability to measure analyzing power up to the arbitrary normalization factor, but accurate study of the systematic errors is needed for that.• Standard energy calibration method was found to be unreliable, new method of calibration are suggested but more development is still needed.• Experimental evaluation of the rate effect is consistent with estimation of pileup contribution.• More accurate control of energy losses in the target is needed.