CHESS upgrade with compact undulator

magnets: Operating experience and first

results
Cite as: AIP Conference Proceedings 1741, 020003 (2016); https://doi.org/10.1063/1.4952782
Published Online: 27 July 2016

A. Temnykh, D. Dale, E. Fontes, A. Lyndaker, Y. Li, J. Ruff, P. Revesz, and A. Woll

ARTICLES YOU MAY BE INTERESTED IN

Latest experiences and future plans on NSLS-II insertion devices

AIP Conference Proceedings 1741, 020004 (2016); https://doi.org/10.1063/1.4952783

Overview of the superconducting undulator development program at ANKA

AIP Conference Proceedings 1741, 020002 (2016); https://doi.org/10.1063/1.4952781

Magnetic field optimization of a novel hybrid permanent undulator

AIP Conference Proceedings 2054, 030029 (2019); https://doi.org/10.1063/1.5084592

AIP Conference Proceedings 1741, 020003 (2016); https://doi.org/10.1063/1.4952782 1741, 020003

© 2016 Author(s).
 CHESS Upgrade with Compact Undulator Magnets:
Operating Experience and First Results

A. Temnykh a), D. Dale, E. Fontes, A. Lyndaker, Y. Li, J. Ruff, P. Revesz

and A. Woll

Cornell High Energy Synchrotron Source

a)
Corresponding author: [email protected]

Abstract. In November 2014 two in-air 1.5m CHESS Compact Undulator (CCU) magnets built by KYMA S.R.l. were
installed in Cornell Electron Storage Ring (CESR) in canted arrangement and after few days of commissioning their
regular operation started. CCU magnets are compact, lightweight and cost efficient devices. They have very stable
magnetic field integrals independent of deflection parameter value. This feature greatly simplifies the storage ring
operation. The CCU concept was developed at Cornell in 2011 and the first 1m in-vacuum CCU magnet was beam-
tested in 2012. The article presents CCU concept and some details of the design. It describes also the layout of CCUs
installation in CESR, their performance and characteristics. The current status of operation and future plans are
discussed as well. Presently, at CHESS two CCU magnets provide radiation for 5 out of 11 experimental stations.

INTRODUCTION: UPGRADE CONCEPT

Cornell High Energy Synchrotron Source (CHESS) is based on Cornell Electron Storage Ring (CESR) which
was commissioned in 1979 as electron-positron collider with counter rotating electron and positron beams. For
approximately 30 years CESR has been used for high energy physics (HEP) study. During this period of time
CHESS operated in parasitic regime receiving radiation generated by both electron and positron beams in bending
magnets and in two permanent magnet (PM) wigglers. In 2010 HEP program was discontinued and CHESS
operation became premier mission for CESR. As the HEP program finished, some constrains on CESR operation
were removed enabling CHESS upgrade with undulator magnets.
Before the upgrade, three beam lines in “G” area, G1, G2 and G3, and two beam lines, A1 and A2, in “A” area
were receiving radiation generated by positron and electron beams in a PM wiggler. The wiggler radiation fan was
wide enough to feed beam lines separated by ~1mrad angle in the horizontal plane as shown on the top of Fig.1.
In the upgrade, this wiggler was replaced by two canted compact undulators see Fig.2. The length of undulator
magnets was chosen to fit the space available after wiggler removal, and canting angle was defined to send
undulator radiation along the existing beam-lines, as shown in the bottom of Fig.1. Because of undulator magnet
compactness, light weight and simplicity of control, they were easily fitted into existing infrastructure minimizing
down time and cost.
The following sections describe performance of two critical components of the upgrade: the CCUs and the thin-
wall vacuum chamber designed to operate along with the CCUs. Experience obtained in the course of 9 months of
the CCUs operation and improvement of upgraded beam lines will be discussed as well.

Proceedings of the 12th International Conference on Synchrotron Radiation Instrumentation – SRI2015

AIP Conf. Proc. 1741, 020003-1–020003-4; doi: 10.1063/1.4952782
Published by AIP Publishing. 978-0-7354-1398-6/$30.00

020003-1
 FIGURE 1. West CHESS upgrade illustration. On the left - A1, A2 and G1 and G2/3 beam lines fed by radiation from G-line
wiggler before upgrade and by two CCUs after the upgrade. On the right – two KYMA CCUs installed in CESR.

CHESS COMPACT UNDULATORS

CHESS Compact Undulators were developed at Cornell in 2011 [1]. Undulators consist of two planar pure
permanent magnet (PPM) arrays supported by side plates. Bottom arrays are stationary, top arrays are moving along
electron beam axis while the gap between arrays is fixed. The motion provides adjustable phase (AP) magnetic field
amplitude control developed in [2].
In the process of preparation of CHESS undulator upgrade, we built and beam-tested two CCU prototypes. The
first was 1 m long “in-vacuum” device; the second was 30 cm long “in-air” undulator with vacuum chamber fitted
inside. After considering performance, cost, and ease of operation, the “in-air” version was chosen for the upgrade.
In 2014 KYMA S.R.l adopted all necessary technologies, built and delivered two “in-air” 1.5m long CCU magnets.
The technology transferring and construction process are described in [3].

FIGURE 2.Cross-section, on the left, and 3D view, on the right, of CCUs built by KYMA for CHESS upgrade.

CCUs built by KYMA, see 3D view and cross-section in Fig.2, have 28.4 mm period, 6.5mm gap, 0.93T peak
field which provides Kmax= 2.5. In the K operating range which extends from 2.5 to 0.8, RMS phase error is
varying from 2.5 degree (at K=2.5) to 5.0 degrees (at K = 0.8). This phase variation does not affect the spectral flux
density and brightness of undulator radiation because large K variation is required only to control energies of low
undualtor harmonics. The brightness of these harmonics has low sensitivity to the undulator phase errors. The
brightness of high order harmonics is more sensitive to phase errors, but their energy control requires just small K
change which keeps undulator phase errors at low level.
The predicted increase in spectral flux and the gain in brilliance due to upgrade were confirmed by
measurements.
For the upgrade we chose the “in-air” CCU version. That simplified the magnet fabrication, but for operation in
storage rings that required very specific vacuum chamber, which is described in the next section.

020003-2
 THIN-WALL UNDULATOR VACUUM CHAMBER
The vacuum chamber designed and fabricated to operate along with “in-air” CCU: a) is fitted inside of the
undulator and does not interfere with movable undulator array; b) is mechanically stable under atmospheric pressure
load in a temperature range from the room to 140 0C needed for vacuum baking; c) has 4.5 mm vertical aperture
providing enough space for circulating beams; d) has 1.2 nTorr pressure and less than 50C temperature rise at
200mA of total beam current; e) minimally contributes to the total storage ring beam impedance.
The vacuum chamber with cross-section depicted on Fig. 3 (top left) was machined from aluminum (6061-T6)
extrusion showed by dashes.

FIGURE 3. On the left top, thin wall undulator vacuum chamber cross section, dashed line indicates aluminum extrusion profile.
On the left bottom, FEM analysis result showing thin wall deformation under atmospheric pressure load. On the right,
deformation of thin wall under atmospheric pressure measured across the groove.

The machining was done in-house under tight control. The most delicate machining of the thin wall was done in
several steps. After each cut, we measured the wall thickness by ultrasonic wall thickness gauge (Model TI-PVX)
and introduced correction into the next cut according to the measurements. In this way we achieved required 0.5 mm
wall thickness with -0.00+0.100 mm precision without damaging the wall.
The chamber is 3.3 meter long and extends through the both undulators. It has transitions from 50 mm to 5 mm
vertical aperture on both ends and smooth 5 mm aperture inside. The end transition profile was optimized to
minimize contribution to the beam impedance following reference [4].
Mechanical stability of thin walls was evaluated with FEA at design stage and validated with measurements. The
left-bottom plot in Fig. 3 displays the wall deformation FEA result. The 0.2 mm predicted deformation is in good
agreement with observed, right plot on Fig. 3.
Vacuum chamber has three sources of heating. The first is synchrotron radiation (SR) from adjacent dipole
magnets. This power is removed by the water flowing in 12.7 mm diameter channel on outer side of the ring. The
second source is Higher Order Modes heating at the end transitions. Based on the other facilities experience we
expected few tens of Watts and put few cooling channels in the transitions bodies providing sufficient cooling
capacity. Third source is a heating caused by a beam image current. Although calculation predicted just few Watts,
this source was of the most concern. Because the energy is depositing directly to thin walls with low heat
conductance, the walls could be easily overheat and lose mechanical stability. To resolve this concern, full scale
aluminum thin-wall vacuum chamber prototype was tested with electron and positron beams. The thin walls
temperature rise at 200 mA of beam current in all possible beam configurations was less than 5 degrees C. That
ensured safe operation.
Vacuum pumping scheme and other vacuum properties of the chamber are described in reference [7].

OPERATION EXPERIENCE AND FIRST RESULTS

Along with PM wiggler replacement with CCUs, 5 CHESS beam lines, A1, A2, G1 and G2/3, had to be upgrade
to accommodate undulator beams. Some upgrade challenges as well as achievements are listed below.

020003-3
 The high intensity of the undulator radiation raised heat load engineering challenges on some of the beamline x-
ray optics and front-end hardware. Many beamline elements (primary beam stops, video position monitors, apertures
and x-ray optics) had to be reengineered and replaced. In the course of the upgrade, novel video beam position
monitors were designed and installed to record the horizontal and vertical positions of the x-ray beams and provide
position stability of 2-5 microns [5].
To handle the elevated heat loading A1 monochromator crystal had to be changed from silicon to diamond. As a
result, the fixed-wavelength side-bounce macromolecular crystallography station changed wavelength from
approximately 12 to 19 keV.
A2 beamline front-end and optics were replaced as well. Transitioning away from silicon to high-heat-load
diamond optics has allowed A2 to access the full undulator spectrum without any need to filter. The beamline can
now deliver incident photons at any energy betweeen 5 keV and 70 keV. The narrow band-pass from the double-
bounce diamond <111> monochromator combined with the collimated beams from the CCU provide very high
resolution beams for scattering, imaging, and spectroscopy experiments. The additional brightness of the CCU
source enables this higher resolution operation without any decline in flux-on-sample as compared to the pre-
upgrade A2 [6].
Because the new undulators are positioned further away from the G-line x-ray optics than in the past, the existing
wide-band-pass multilayer optics designed for wiggler beams were left in place. Presently G1 operates with flux 3.4
times higher than prior to upgrade. At G3, flux of over 1014 photons/sec was measured in a 1 mm2 at 11.2 keV at
standard CERS operation conditions. This is approximately 2.5 times the prior G3 record. In both cases, the
undulator upgrade represents substantial gains for G-line users.

SUMMARY
In summer 2014 we replaced one of the permanent magnet wigglers feeding 5 CHESS beam lines on West side
with two canted compact undulators and upgraded several beam lines optical elements to accommodate undulator
beams. The upgrade noticeably increased experimental capability of A1, A2, G1 and G2/3 CHESS beam lines.
In 2016 we plan to replace East CHESS wiggler with two canted compact undulators to feed 4 new experimental
stations.

ACKNOWLEDGMENTS
Authors would like to thank CESR operating group for establishing storage ring operation with undulator
magnets and CESR vacuum group for work on undulator vacuum chamber,
We also express great appreciation to KYMA engineering and technical staff for efficiency and creativity in the
CCU technology adopting and for “in-time” undulator delivery.
This work was supported by NSF award DMR-1332208.

REFERENCES
1. A. Temnykh et al., Compact Undulator for Cornell High Energy Synchrotron Source, IEEE Transactions on
Applied Superconductivity, Vol. 22, No. 3, June 2012.
A Temnykh et al., 2013 J. Phys.: Conf. Ser. 425 032004.
2. R. Carr, Adjustable Phase Insertion devices as X-ray sources (NIMA, vol. 306, p. 391, 1991).
3. A. B. Temnykh et al., Construction of CHESS compact undulator magnets at Kyma, In Proc. SPIE 9512,
Advances in X-ray Free-Electron Lasers Instrumentation III, 951205 (May 12, 2015), doi:10.1117/12.2186301.
4. B. Podobedov and I. Zagorodnov, Impedance Minimization by Nonlinear Tapering, In Proc. PAC07.
5. P. Revesz, A. Lyndaker, A. Pauling and A. Temnykh, Novel video beam position monitor for canted
undulators at CHESS, http://news.chess.cornell.edu/articles/2015/Revesz150227.html
6. J. Ruff, A2: Out with the wigglers; in with the undulators,
http://news.chess.cornell.edu/articles/2014/Ruff141103.html
7. Y. Li et al, Vacuum Performances of 5-mm Undulator Chamber for Cornell High-Energy Synchrotron Source,
Presentation on AVS 61st International Symposium & Exhibition, Nov 9-14 2014, Baltimore, USA

020003-4