Cryogenic Permanent Magnet Undulators

Report
Cryogenic Permanent Magnet Undulators
Finn O’Shea
March 27, 2013
HBEB 2013, Puerto Rico
Outline
1)
Motivation for shorter period technology

2)
Prototype undulator has been tested

3)
As usual, it is money
Results are ‘unsuprising’ and that is a good thing!
Improving performance of cryogenic undulators through the
use of rare-earth element poles – the DPU project

Cryogenic compatible magnetic materials lead to improved
performance

Can we do the same with the pole material?
2
Motivation
o Researchers from many branches
of science are using modern light
sources to do a lot of cutting edge
research
o LCLS accepts ¼ of proposals for
beam time
o These machines are big (km) and
expensive ($1/2 billion)
o Linacs are the really expensive part
o NGLS is being built through a
mountain – real estate is getting
more expensive
3
Where do short periods fit in?
 “Moore’s Law” of radiation brightness is doubling every 10 months
since the 1960s – this has come from bigger, more expensive
facilities
 Increased brightness does not lead to increased access unless the
facilities become more common
 The path to cheaper access is to increase the amount of radiation
produced by each electron
 Lots of beam lines on synchrotrons
 Multiple beam lines at FELs: LCLS-II, NGLS, SwissFEL
 Reduce the energy of the electrons required to produce the desired
wavelengths
4
To make shorter periods…
 The magnetic material needs to be very radiation resistant
 In vacuum undulators have higher exposure because of the smaller
gap and no vacuum chamber wall for protection
 Cryogenically cooled magnets show a modest increase in remanent
field and a massive increase in coercivity
 They also show an increase in resistance to radiation induced
demagnetization
 Originally attributed to coercivity increase,
more likely due to increase in heat
capacity decreasing the effects of local heating
 Rad damage is not well understood:
 Radiation damage is reversible with
remagnetization  no structure change
5
Strategy to increase survivability
1)
Use material that has maximum remanent field at room
temperature
 Limitation is ability to assemble the undulator
2)
Cool as much as possible to get the highest possible
magnetic field and largest coercivity/heat capacity
 The clear choice is PrFeB
 No SRT (NdFeB)
 High remanent field (SmCo)
6
Benabderrahmane, NIM A 669, 1 (2011).
Example of performance change
LCLS Normal
LCLS Low Charge/CPMU
 Energy = 13.6 GeV
 Energy = 4.5 GeV
 Charge = 250 pC
 Charge = 250 fC
 Norm Emittance = 0.4 μm
 Norm Emittance = 33 nm
 Saturation Length: 60 m
 Saturation Length: 15 m
 Pulse Energy = 1.5 mJ
 Pulse Energy = 2.8 μJ
 Pulse Length = 100 fs
 Pulse Length = 0.5 fs
 B = 2 x 1033 ph/(s mm2 mrad2
 B = 1.3 x 1036 ph/(s mm2 mrad2
0.1%)
7
0.1%)
PRSTAB 070702 (2010)
CPMU9
Testing of Cryogenic Permanent Magnet Undulator – 9 mm at the Next Linear
Collider Test Accelerator
Cryogenic Permanent Magnet Undulator
– 9 mm period
• 9 mm period length
• 20 period prototype
• Compensated 1st integral
• Working temp down to 11K
• NLCTA experiment run at 43K
9
CPMU - II
 Design process was iterative: use FEM and BIM codes to
determine magnetically safe assembly and operating
conditions as material is characterized.
 Results in a 2D geometry with pieces that are strategically
chamfered to reduce reverse fields
10
Measuring the field
 Field is measured at
cryogenic temps on a
specially constructed
measurement bench at
HZ-Berlin
 Bpeak=1.15 T (K=0.97)
11
Next Linear Collider
Test Accelerator
Facilities at NLCTA made it an
excellent place to test the undulator
using a scaled experiment at optical
frequencies.
• Radiation bandwidth measurement
•Energy modulation measurement
• Confirmation of microbunching
12
Undulator Radiation
Bandwidth
 at NLCTA the bandwidth of
the radiation should be
dominated by the 5%
bandwidth of the single
electron radiation process
13
Laser seeding
 800 nm laser is used to seed
the FEL mechanism
 Laser is shorter than the
electron beam, they are
about the same transverse
width
 Leads to very 3D process
 Energy modulation strength
is function of distance along
beam and radius
 Genesis predicts:
K = 0.97
14
K = 0.94
Observation of microbunching
 Microbunching of electron
beam causes coherent
emission of transition
radiation
Laser off
 CTR signal shows up as near
field structure when laser is
turned on
 Null tests showed that this is
likely forward CTR from OTR3
rather than backward CTR
from OTR4, which is the
screen that is imaged
15
Laser on
Summary of Results
 Bandwidth of the undulator radiation is dominated by the 5%
interference bandwidth – consistent with expectations
 Energy modulation is consistent with expected value from
iFEL interaction
 CTR appears when modulation is turned on – microbunching
is occuring although scattering in OTR3 is spoiling the
measurement
16
Rare-earth poles
Dysprosium poles at RadiaBeam Technologies
Why replace CoFe?
 Vanadium Permendur (49% Fe, 49% Co, 2% V) is an excellent pole
material
 Saturates at low applied field:
μi~104 and Hsat<<0.1 T
 Bsat = 2.35 T
 The reason for replacing CoFe as the pole material of choice is to
get higher saturation induction
 Because we need to cool a CPMU anyway, what gains can be
realized?
18
Rare-earth elements
 Materials such as dysprosium,
gadolinium and holmium
show large sat. ind. at
cryogenic temperatures
 Dy -> 3.8 T (single crystal)
 Single crystals are hard to
grow and polycrystals are not
useful
 Secondary re-crystallization
can be used to develop
“texture”
19
Secondary Re-crystallization
 Rolling then annealing exploits an energy advantage that
results in the growth of the grains in-plane
20
21
Performance of textured Dy
 100 μm thick foils
 CoFe
• Competitive with CoFe if the applied field is
greater than ~0.10 T
• Thinner foils should work even better 
• Non-linear nature and mixing of 1120 and
1010 in different samples could be a problem
22
Short test undulator
 Most poles are CoFe
 One pole pair is replaced with Dy laminated poles
 The field is measured while the undulator is cooled
 BCoFe is compared to BDy
23
Results
~3%
24
Details
 Effect is reproduced in Radia with the magnetization curves
measured from one pole (destructive measurement via VSM)
 Dy Pole position exposed it to larger applied fields than a
typical pole
 working point is above
the Dy > CoFe point
 Shows promise
25
Summary
 Using shorter period undulators can:
 Decrease the electron beam energy required to reach a given
wavelength
 Extend the reach of existing facilities
 This can lead to cost savings and a potential increase in accessibility
 A prototype short period cryogenic undulator has been built and
tested in a successful scaled experiment at optical wavelengths
 Rare-earth poles have the potential to outperform CoFe poles in
undulators that can be cooled to the temperatures where they are
ferromagnetic
 Thank you!
26

similar documents