1. JLAB 12GeV Operations Magnet Measurement Overview Ken Baggett, Joe Meyers, & Renuka Rajput-Ghoshal for the Metrology and Magnet Measurement Department
IMMW October 2015

2. What’s happening at JLAB
Jlab has completed installation of magnets that have increased the accelerator’s power output from 6GeV to 12GeV.
The experimental halls have also been upgrading operations to take advantage of the increased energy capabilities of the accelerator.
For magnet measurement this means a change of paradigm:
Move away from “production line” magnet measurement setups
Embrace “in-situ” magnet measurements
Larger magnets, higher power
MMF not equipped to handle (crane) or power (facility power supplies)
Superconducting magnets
Challenge for Magnet Measurement Group:
Develop “mobile” measurement systems
Ensure measurement specifications continue to be met

3. Design overview: 12GeV Upgrade Scope
Refurbish and Measure
Spreaders and Re-combiners,
Extraction, & BSY: ~70 Dipoles
Tagger Magnet
Location
Complete: Refurbish and measure arcs 2,4,5,6,7,8,&9

4. Magnet Measurement Lab - Measurement Stands
SW System
Stretched Wire was used for the excitation curve mapping (BL vs. I)
System core: 4 μm Newport stages / HP3458A voltmeter
Cross checked matching between the SW and Hall probe systems
Stepper Stand
Stepper Stand used for field quality measurements
Group 3 Hall probes
Linear Encoders
Automated Calibration Routine

5. 5 Probe Holder
Probe holder built for multiple Hall probes.

6. Dipole Grid Mapping 1m “H” dipoles mapping Solution
5 Probe Holder
(1 cm spacing)
Teflon Guide
G-10 probe cart
1m “H” dipoles mapping
Need lots of points
Single probe on a stick method is slow
Solution
Hall probe holder for 5 probes
G-10 mounting cart for probe holder
Synchronized x stages on both ends of dipole
Z axis stage pushes cart through the x stages

7. What we’ve done
Completed the measurement process from the 12GeV upgrade
~400 large dipoles
Spec for accuracy of absolute strength (BL) = 0.4%
Spec for precision = 0.03%
117 quads
>100 correctors
24 hall line dipoles
Approach used to meet specs and throughput:
Stretched wire used to get BL vs I for all magnets
Detailed maps with Hall probes and NMRs used to build map along curved trajectory.
Done for 10% sampling of arc dipoles.
Done for all spreader and re-combiner dipoles
Continued development of the linear stages
Integrate into super-conducting magnet measurements for experimental halls.

8. Model vs Real Machine Measured in real machine Model to fit BPM gains
LOCO method was used to consider the effect of quadrupole errors, dipole construction errors, and beam position monitors calibrations.
Model is expanded to first order in Taylor series of the errors.
A set of difference orbits obtained by exciting the correctors along the beamline is taken, yielding the measured response matrix.
An iterative procedure is invoked and quadrupole errors as well as beam position monitors (BPM) calibration factors are obtained.
Measured in real machine
Model to fit
BPM gains
Quadrupole and body gradient errors
from Y. R. Roblin, May 2012

9. Matrix formalism
from Y. R. Roblin, May 2012

10. Iterative Algorithm ,
from Y. R. Roblin, May 2012

11. ARC4 Fit to 2012 data. 2011 data 2012 data Fit to 2012 Design
from Y. R. Roblin, May 2012

12. ARC6 Fit to 2011 data. 2011 data 2012 data Fit to 2011 Design
Bad bpm
from Y. R. Roblin, May 2012

13. Measured Body Gradients
K1 ( 𝒎 −𝟐 )
𝑩 ′ 𝑳 𝑩𝑳 (%/cm)
(magnet stand)
ARC
Amps
1.26e-3 (2012)
0.026 ±0.002
0.027
ARC6
273
1.8e-3 (2012)
0.036±0.002
0.034
ARC4
183
1.53e-3 (2012)
0.031±0.002
0.029
ARC5
229
3.9e-4 (2011)
0.012± 0.006
0.017
ARC7
214
-7e-5 (2012)
-0.002± 0.006
0.013
ARC9
275
3.0e-3 (2011)
0.016± 0.004
0.014
ARC3
MBA 3m
MBB 2m
MBE 1m
from Y. R. Roblin, May 2012

14. Hall D Tagger Magnet
from D. Sober, Gluex Collab. Meeting 21 Feb. 2014

15. Analysis of tagger magnet field maps
Ultimate goal:
To allow energy determination by mapping to a precision of 0.1% for beam paths to focal plane and beam dump
To accomplish this, require as intermediate goals
Better than 0.1% accuracy in average field (15 gauss at 1.5 T)
~3 mm accuracy in the absolute positions of the magnet entry and full-energy exit edges
~0.5 mm accuracy in the absolute position of the long exit edge
from D. Sober, Gluex Collab. Meeting 21 Feb. 2014

16. Field map analysis
Important fact: We have a detailed Tosca field calculation (done by Guangliang Yang at Glasgow Univ.) of the magnet as built.
Comparison of the measured field maps with Tosca calculations gives a powerful mechanism for
Checking the alignment of the field map coordinate system
Checking the consistency of the 6 mapping configurations
Extrapolating to unmeasurable regions near the entry and exit beam pipes
from D. Sober, Gluex Collab. Meeting 21 Feb. 2014

17. Tagger Magnet Mapper

18. Tagger Mapper Arm

20. Field map analysis
Mapping was performed between 07 January and 11 February 2014
Measured B vs X and Y on a 1 cm-by-2.5 cm grid
5 X-values measured simultaneously using 5 Hall probes, 1 cm apart
4 cm advances in X (so that Probe 1 and Probe 5 repeat a point)
2.5 cm steps in Y (+ 1 cm steps near entry and full-energy exit edges)
Mapping apparatus was not large enough to cover the entire required region (~65 by 650 cm) in one setting:
6 configurations (shown on next slide), with survey required for each.
Full maps were measured at 3 excitations, to compare with the Tosca model and allow interpolation:
1.7 Tesla (E0 = 13.6 GeV)
1.5 Tesla (E0 = 12 GeV)
0.75 Tesla (E0 = 6 GeV)
from D. Sober, Gluex Collab. Meeting 21 Feb. 2014

21. Points measured in field mapping: Configurations 1-6 Grids: 1 cm in X, 2.5 cm in Y
from D. Sober, Gluex Collab. Meeting 21 Feb. 2014

22. Field map analysis For each mapping configuration The bottom line:
Plot B vs x and y for all data points
Calculate ratios of probe readings in uniform field region (to tweak the probe calibrations)
Calculate field averages in regions which overlap other configurations
Compare edges with Tosca to check map alignment
Compare magnitude and shape with other configurations
The bottom line:
No important differences from Tosca calculations.
Alignment seems (almost) good enough.
from D. Sober, Gluex Collab. Meeting 21 Feb. 2014

23. Raw Data
from D. Sober, Gluex Collab. Meeting 21 Feb. 2014

24. Using probe calibration factors
from D. Sober, Gluex Collab. Meeting 21 Feb. 2014

25. The entry field region (Configuration 6)
Single entry for all beam traces
Path tracing requires full XY grid to calculate B’(x) normal to the beam entry trajectory.
essential for focusing effects.
Impossible to measure a full grid in this region because the entry beam pipe is part of the vacuum chamber – already in place.
Solution: Use extrapolation and Tosca field, anchored by accurate measurements along the beam axis
from D. Sober, Gluex Collab. Meeting 21 Feb. 2014

26. Measured points in entry region (Config. 6)
from D. Sober, Gluex Collab. Meeting 21 Feb. 2014

27. Tagger Mapping Summary
The agreement of the measured field configurations with each other and with the Tosca calculations is mostly excellent (except for a few alignment glitches). We now know that we can trust Tosca for extrapolating the map into unmeasurable regions.
We still have a long way to go before we have final unified maps for the 3 excitations. A lot of work is needed to tune the probe calibration data, merge the configurations, create 2D grids in the entry and full energy exit regions, … (Months, not years.)

28. Moller Quad Mapping

29. Rotating Coil Circuit board probe
Continues to be the probe of choice for most rotating coil measurements.
Easy to adapt drawings
Cheap to purchase
Quick delivery
Great signal to noise for multipoles
Compared well to existing wound probes
Calculated absolute strength compares at the 0.06% level (DBUCK coil compared to SW)
4 Printed Circuit Layers
4 Coils
UBUCK, DBUCK, DQBUCK
100 m trace width
150 m trace spacing
DBUCK given by:
(0, y2)
(0, y1)
(0, y4)
(0, y3)
𝐾 𝑛 = 𝑅 𝑛 𝐿 𝑖𝑦 1 𝑅 𝑛 − 𝐿 𝑖𝑦 2 𝑅 𝑛 - 𝑅 𝑛 𝐿 𝑖𝑦 3 𝑅 𝑛 − 𝐿 𝑖𝑦 4 𝑅 𝑛

30. Mobile Rotating Coil Stand

32. Simple Radius Scaling Experiment

33. Harmonics
Harmonics of large 8” aperture “Moller Quad” magnet measured with circuit probe using fiberglass offset.

34. Hall B Torus Mapping (upcoming challenge)

35. Magnet in Hall – Mechanical Constraints
6 meters
Coils electrically connected in series and assembled around a central hub, magnet will be mapped after magnet is trained to full field.
Axis of the torus is 6 m above the floor, limited room between the coils and around the magnet.

36. Torus Mapping: Physics Requirement
Best 𝒅𝒑/𝒑 resolution is from particle trajectories near the beam-line (highest B-field)  low-angle elastic scattering strictest requirement
Here is the trajectory of an elastically scattered electron from 11 GeV beam
(p = 10.1 GeV/c, q = 7o)
For this track, ideal 𝒅𝒑/𝒑 ≈𝟎.𝟑%
 need to know 𝐵 𝑑𝑙 to ~ 0.1%
The fractional momentum resolution (dp/p) is best at high magnetic fields, and the physics requirement
for good resolution is most stringent at high momentum. If we have ideal tracking chamber performance we expect
0.3% for dp/p at small angles and small radius. This is the most stringent requirement for knowledge of the
B-field strength. To not significantly worsen the ideal case of 0.3%, we need to know B to 0.1%

37. Field mapping locations
Mapping Plan- Where to Map
Field mapping locations
We want to measure at small radius to be sensitive to coil placement. We want another measurement at
~80cm radius because there is a ‘sweet spot’ (a certain value of radius and z) where the field is purely
azimuthal.

38. Summary & Moving Forward
12GeV accelerator upgrade measurements have been completed.
~400 large dipoles
117 quads
>100 correctors
Tagger magnet mapping completed successfully. Preliminary beam based observations confirm measurement validity.
Experimental Hall magnet mapping is primary focus.
Continue development of the “in-situ” mapping abilities
Continue integration into super-conducting magnet measurements for experimental halls.

39. Thank For Your Attention! Questions?