QD0 Prototype Magnetic Measurements Michele Modena, TE-MSC QD0 Workshop 2013, 29 January 2013.

QD0 Prototype Magnetic Measurements

Michele Modena, TE-MSCQD0 Workshop 2013, 29 January 2013

Michele Modena, TE-MSC “QD0 Prototype Field Quality Measurements” CLIC Workshop 2013

Outline:

- The CLIC MDI and QD0 baseline design (for L* = 3.5 m optic)

- The QD0 magnetic measurements campaign in 2012

- The 2013 magnetic measurement plan

- Conclusions and Future activities on QD0


CLIC BDS/MDI layout with QD0s “embedded” in the 2 Experiments (SiD, ILD) End-Cups


Magnet design boundary conditions:

- As much as possible compact design (to be compatible with an L* of 3.5 m, so minimizing the solid angle subtracted to the experiment Detector)

- Compatible with magnet active stabilization (i.e. minimize magnet weight and vibration sources, ex. coil water cooling)

- Presence of the post-collision line beam vacuum chamber (in its closer position at 35 mm from beam axis)

QD0 study & design requirements

QD0 Baseline Parameter Value

Nominal target for field gradient 575 T/m

Magnetic length 2.73 m

Magnet aperture (required for beam) 7.6 mm

Magnet bore diameter8.25 mm*

* Including a 0.30 mm vacuum chamber thickness

Good field region (GFR) radius 1 mm

Integrated field gradient error inside GFR < 0.1%

Gradient adjustment +0 to -20%


QD0 in the MDI layout with L*=3.5 m layout (simplified view)


“Hybrid design” basic concept

-Comparing to a “classical” quadrupole design, the presence of PM wedges compensate strongly the saturation of the poles Max Gradient increase of a factor 1.5-1.68 Of course, tunability is the other most interesting aspects of this configuration.

NI= 5000 A

Grad [T/m] Sm2Co17 550

Grad [T/m] Nd2Fe14B 615

NI= 5000 A

Grad [T/m] Sm2Co17 531

Grad [T/m] Nd2Fe14B 599

-The presence of the “ring” casing the PM blocks cause a reduction of ~ 4-5 % in gradient performance but has big advantages for the manufacturing and assembly technical aspects!


Evolution of the prototype design, conceptual design of the full size magnet

Mode 1st 2nd 3rd 4th

Freq [Hz]

190 260 310 366


Final assembly of the prototype in Fall 2011:


Outline:






The only available measuring system to perform the magnetic gradient and magnetic field quality measurements are the “stretched wire” and the “vibrating wire” systems. This is due to very small magnet aperture (Ø: 8.25 mm!)

1) The “stretched wire” is a well know, state-of-the-art measuring method for gradient and magnetic axis:

- expected accuracy: 10 -3

2) The “vibrating wire” is a new CERN development (still in a prototype version!) but that is giving very good results in terms of easy, fast and magnet radius independency, field quality (multipoles or harmonics) magnetic measurements.

- when utilized for harmonics measurement: expected accuracy : < ±3 units ( accuracy for measurements done at 3 mm and then extrapolated at 1 mm radius as specified in our case)

- when utilized for magnetic axis measurement: expected resolution: 1 μm.

The QD0 magnetic measurements campaign in 2012

Some References for the two measuring methods:

1) D. Zangrando, R.P. Walker: “A Stretched Wire System for Accurate Integrated Magnetic Field Measurements in Insertion Devices” at: http://oldweb.elettra.trieste.it/organisation/accelerator/docs/st_m_96_01.pdf2) P Arpaia, M Buzio, J Garcia Perez, C Petrone, S Russenschuck and L Walckiers: “Measuring field multipoles in accelerator magnets with small-apertures by an oscillating wire moved on a circular trajectory” at: http://iopscience.iop.org/1748-0221/7/05/P05003

Reference for the CLIC QD0 Magnet Measurements campaigns results:

1) P. Arpaia et al: “Magnetic Measurement of the Model Magnet QD0 (Nd2Fe14B), designed fort the CLICX Final Focus Beam Transport Line”, EDMS: 11841962) P. Arpaia et al: “Magnetic Measurement of the Model Magnet QD0 (Sm2Co17), designed fort the CLICX Final Focus Beam Transport Line”, EDMS: 1262052

http://oldweb.elettra.trieste.it/organisation/accelerator/docs/st_m_96_01.pdf

http://oldweb.elettra.trieste.it/organisation/accelerator/docs/st_m_96_01.pdf

http://iopscience.iop.org/1748-0221/7/05/P05003


Two campaigns of measurements were done in 2012 with QD0 prototype in two different configuration:- In January 2012: the magnet equipped with the Nd2Fe14B blocks was measured with the “Vibrating wire system” (see after)- In August 2012: the same type of measurement was done for the configuration with Sm2Co17 blocks .

0 1 2 3 4 5 6 70

50

100

150

200

250

300

350

400

450

500

550

600

Prototype 100 mm, Nd2Fe14B, CALCULATEDPrototype 100 mm, Nd2Fe14B, MEASUREDQD0, Liron >300 mm, Nd2Fe14B

AMPERE-TURNS PER POLE [kA]

FIE

LD

GR

AD

IEN

T [

T/m

]

0 1 2 3 4 5 6 70

50

100

150

200

250

300

350

400

450

500

550

600

Prototype 100 mm, Sm2Co17, CALCULATEDPrototype 100 mm, Sm2Co17, MEASURED

AMPERE-TURNS PER POLE [kA]

FIE

LD

GR

AD

IEN

T [

T/m

]

1) Maximum achievable gradient:Here below the measurements of the MEASURED Gradients (red dots) (extrapolated from the INTEGRATED gradient effectively measured), plotted together with the COMPUTED Gradient (blue curves).

The measured Gradient in the configuration with Sm2Co17 blocks it is in very good agreement with the FEA computation. This is not the case for the Nd2Fe14B blocks were a difference of ~ -6% is noticeable. This could have 2 possible explanations:

-The Permendur part saturate at lower level than expected. to check that we re-compute the performances by FEA utilizing the exact permeability as measured on a production sample; this possible cause was then excluded.-The quality (magnetization module and/or direction) of the Nd2Fe14B PM blocks is not the expected one we should get an answer on this possible cause measuring the PM blocks with a new measuring device (by Brockhaus Messtechnick GmbH) expected at the MM Section for late Spring 2013.


0 1000 2000 3000 4000 5000 60005

10

15

20

25

30

35

40

45

50

55

Nd2Fe14B: GdL [T] The 2 graphs show, for both QD0 configurations, the measured INTEGRATED Gradient in function of the magnet powering (total ampere-turn in the coils).

A minimum hysteresis effect is visible. Due to the “hybrid” configuration, with a PM fix contribution at zero current, the hysteresis loop are not center to the axes intersection and is not necessary to perform negative current cycle to pre-cycle the magnet, but only to zero the current.

From this point of view the gradient setting is more straightforward comparing to a classical EM magnet design.

2) Magnet cycles repeatability:

0 1000 2000 3000 4000 5000 60005

10

15

20

25

30

35

40

45

50

55

Sm2Co17: GdL [T]


2) Magnet cycles repeatability:

Nd2Fe14B

Sm2Co17


The below histograms provides for both QD0 configurations the magnetic harmonic content (multipoles) versus the magnet powering ; upper graph for Nd2Fe14B, lower graph for Sm2Co17. For comparison: the first computed “permitted” mutipole at NI=5000A is : b6=1.4 units (integrated) for Nd 2Fe14B and b6=0.7 units for Sm2Co17. The concerned measured components are the one circled in red.

3) Magnetic Field Quality:

IMPORTANT: remind that the measurements global accuracy is estimated at: ±3 units!


Outline:






1) Investigation of the performance with Nd2Fe14B magnet configuration:

In order to understand the discrepancy between computed and measured maximum gradient, we propose to perform an individual measurements of each inserts and compare their characteristics.

Being each of the four PM inserts obtained by gluing 4 single PM wedged block with two different magnetization direction (see left figure below), we can only do an average measurements of each insert (no possible to easily detect which single blocks is not conform in magnetization module or direction). We can perform a relative measurement, comparing the magnetic moments (average magnetic flux density) of each insert. This measurement can be done with a special measuring device based on 3D pairs of Helmholtz coils. One of these system is under procuring from industry (Brockhaus Messtechnik GmbH, see right figure below) and the system is expected at CERN in March 2013.

Depending by the results of these measurements, we could decide to eventual order some new set of insert to check the final performances of the quadrupole.

The 2013 magnetic measurement plan

2) Cross-check of the harmonic measurement with a new “Rotating Coils” system:

The new system (probe diam. Ø: 7.7 mm!) is actually under development (see figures below).

(NOTE: All these new MM systems will be very proficient to projects like CLIC (we ask for them as one of the main new “client”) and a very good collaboration with MM Section is in place on these subjects; they develop/procure new tooling and methods and we provide good “test cases” !)


Outline:






Magnetic Measurements Conclusions:1) QD0 maximum Gradient measured in the configuration with Sm2Co17 inserts is in very good agreement with the FEA computation. The measurements configuration with Nd2Fe14B inserts show a difference of -6% respect to the computed values. This effect is probably due to not correct magnetization (module and/or direction) of some of the Nd2Fe14B PM blocks.

2) Magnet cycling seems very precise and repeatable, a minimum hysteresis effect is present.

3) The measured magnetic field harmonics at the required (maximum) gradient are in the range of the required tolerances (10 units).

4) Considering the vibrating wire method measurement accuracy, the measured magnet field quality is in good agreement with the computation. The presence of normal and skew not permitted harmonics depends by the manufacturing tolerances of the Permendur part and by the PM inserts magnetization module and/or direction.

5) By the expected measurements with the 3D Helmholtz coils system, we expect better information on the PM blocks quality and their influence on the overall field quality.

6) The stability of the magnetic axis was confirmed inside ± 2.5μm all along the powering curve of the quadrupole.

Future activities on QD0 prototype measurements & other: 1) PM blocks measurements by 3D Helmholtz coils system.

2) Measurements with a new dedicated rotating coils system under development.

3) Other measurements/activities could be eventually planned to better discriminate the contributions to the field quality by the quadrupolare structure in Permendur and by the PM inserts.

4) Other type of measurements to be eventually planned.

5) Manufacturing of a longer prototype is under evaluation/discussion. This will mainly depends by the evolution of the CLIC/MDI layout studies and R&D. As CERN we are strongly interest to proceed in this direction if useful, anyway this will be an activity planned for after the LHC “LS1”.

Acknowledgments:

Thanks to: A.Vorozhtsov, E.Solodko, P. Thonet, C. Lopez for the magnet design, procurements of components and assembly

Thanks to: C.Petrone, J.Garcia Perez, A. Bartalesi, M. Buzio for magnet measurements and data analysis

Thank you for your attention

QD0 Prototype Magnetic Measurements Michele Modena, TE-MSC QD0 Workshop 2013, 29 January 2013.

Documents

qd0 slide

temsc qd0 workshop

qd0 baseline design

prototype design

magnetic gradient

conceptual design

magnet weight

layout simplified view