Title: A%20selection%20of%20slides%20for%20Jim%20Volk%20to%20consider%20for%20his%20talk%20to%20MAC%20(May%2002),%20covering%20recent%20work%20at%20SLAC%20and%20LBNL.%20People%20to%20be%20credited:%20Jose%20Alonso,%20Jin-Young%20Jung,%20Seung%20Rhee,%20Cherrill%20Spencer,%20Jim%20Spencer,%20SLAC%20Magnetic%20Measurements%20Group.
1A selection of slides for Jim Volk to consider
for his talk to MAC (May 02), covering recent
work at SLAC and LBNL.People to be credited
Jose Alonso, Jin-Young Jung, Seung Rhee,
Cherrill Spencer, Jim Spencer, SLAC Magnetic
Measurements Group.
NLC - The Next Linear Collider Project
2(No Transcript)
3Recent Work at SLAC on measuring radiation fields
how they effect pm materials.
- In order to establish if pm magnets can be used
in the NLC we must first predict how much
radiation of what types will be produced in
various NLC beamlines in the vicinity of the
magnets. Only when these predictions are
believeable can we proceed to testing candidate
pm materials with appropriate types and levels of
radiation. - We have looked at old measurements of radiation
levels in the SLC linac and damping rings and
tried to relate this data to supposed lost beam
power. Could not make the so-called data from the
2 beamlines match in terms of rads produced per
watt of beam lost. - Further analysis of data showed them to to be
generated from a few measurements with many
artificial assumptions and calculations
sufficient to lead to large enough errors in the
data to be of little use to extrapolating to NLC
conditions. - Furthermore the radiation measurements were made
with detectors that did not distinguish between
gammas, electrons, neutrons and protons.
4Why it is important that we can accurately
predict how much radiation will occur in NLC
- We know that permanent magnet materials will lose
some of their magnetization if exposed to
some radiation types at some level we need
to quantify all these somes in order to make
the technology decision for the NLC magnets
electromagnets or permanent magnets (pms). - The result of this decision may be different in
different areas of the NLC because the expected
radiation levels vary by orders of magnitude from
beamline to beamline and within beamlines,
according to our current (but faulty) data.
Plus, the types of radiation produced vary within
beamlines. - Accurately predicting radiation levels and types
is only half the story published data on the
effect of various types of radiation on various
types of pms show sufficient variation in loss of
magnetization for apparent similar radiation
conditions that we are not confident in the
applicability of many of the published
experiments to the ( roughly predicted) NLC
conditions. - Virtually all materials that will be used in the
NLC tunnels will be affected by radiation at some
level- important that we can accurately estimate
it.
5What we do know about radiation effects on some
up to date pm materials.
- Different radiation types have vastly different
effects- summarizing data from best papers - To cause a given effect, (e.g. 20 loss of Br
for certain type of NdFeB) requires - Ö 1 kGray protons (200 MeV Ito et al)
- Ö 10 kGray neutrons (fast Cost, Brown)
- Ö 5 MGray electrons (17 MeV Okuda et al)
- gtgt 10 MGray photons (60 Co several authors)
- 1 Gray 100 rads
- So simple measurement of DOSE is INADEQUATE to
assess likely radiation-induced effects. But that
is all we have for radiation in our best role
model for the NLC the SLAC Linear Collider
(SLC).
6More of what we do know about radiation effects
on pm materials.
- Different magnetic materials have very different
responses to radiation - Measurements show SmCo to be significantly more
resistant than NdFeB in exposures to several
different radiation types - - Factors of 103 are typical (similar effects
for 103 times more dose on SmCo) - BUT (compared to NdFeB)
- SmCo is significantly more expensive (Ö x 10)
- SmCo has long-lived activation products (worst
offender 71-day 58 Co) -
- Measurements in NdFeB show significant
variations depending on - Metalurgical properties Higher coercivity can
have Ö x 10 effect - Macroscopic environment Optimizing
permeance can have Ö x 10 effect - - Brick geometry (Thick blocks better than
thin disks) - Remagnetization restores full Br, even if
material suffers gt80 radiation-induced
demagnetization
7Radiation-Induced Demagnetization(Japanese
experience with 200 MeV protons)
- Material Type has large impact
- Red N48 High Br (1.4T) Low Hc (1.15 MA/m)
- Blue N32Z Lower Br (1.14 T) Higher Hc (2.5
MA/m) - High coercivity material almost 30 times more
resistant - Material Shape has large impact
- (All samples are discs 10 mm dia)
- Circle thickness 2 mm (Pc 0.5)
- Triangle thickness 4 mm (Pc 1.0)
- Square thickness 7 mm (Pc 2.0)
- - Higher Permeance coefficient
- increases resistance (x 10)
- SmCo is much more resistant than NdFeB
Would not put a Neo magnet anywhere close to 200
MeV protons, but will the electrons escaping from
any NLC beampipe produce any 200 MeV protons
? This is the kind of detail we need to establish
in order to choose between electromagnets and
permanent magnets.
8At SLAC/LNBL we have started a 3 pronged
approach to answer the questions about radiation
levels and their effects on permanent magnet
bricks
- I. Field characterization Careful systematic
measurements with appropriate detectors - A. Identify dosimeters capable of
differentiating radiation types - Currently testing dual silicon-based devices
MOSFET (ionizing-radiation (x-ray) sensitive) and
Pin Diode (displacement-damage (neutron)
sensitive) - - Promising first results in SLC DR where we
have had 3 dosimeters taking data for last 6
months , but - - Encountering calibration and normalization
difficulties and the DR is a mixed source and
levels vary dramatically from place to place - - Using a Californium source to calibrate
dosimeters and irradiate samples - Continue searching for other dosimeters
(talking to experts at LBNL for e.g.) - - Inexpensive, for simultaneous measurements
at many sites around DR - - Capable of differentiating different
radiation-types - - Good lifetimes in very high fields
-
9Second prong of SLAC/LBNL effort to understand
radiation effects on permanent magnets II A
Subject pm samples to known radiation sources.
- Have subjected Sm Co5 and NdFeB bricks from
Vacuumschmelze to successive irradiations with a
pure, well-calibrated Co 60 ? source, in between
have measured their magnetization in 3 axes in a
SLAC Helmholtz coil set-up no deterioration
after 277 KGray. - Placed same 2 bricks, attached to a piece of
steel, in the SLC DR near the extraction Y where
we understood the radiation levels would be
high, left them for 70 days, then took them out
and remeasured them no deterioration, but have
not yet characterized the radiation. - Then placed same 2 bricks within a few inches of
a Californium source emitting just 2.2MeV
neutrons also measuring radiation with the Pin
Diode detector. Will take them out after 1
month has passed. - Will continue to expose bricks to radiation in
pure sources and in the SLC DR, and to try to
understand the calibration of the new detectors.
10More on second prong of SLAC effort to understand
radiation effects on pms
- II continued
- B. Develop techniques for measuring very small
changes in magnetization - Available sources of pure radiation not
strong enough to simulate many years exposure,
will most likely have to extrapolate from very
small effects - C Network with research community engaged in
this field - Jyvaskyla Finland (alphas, protons)
- Wakasa-Bay Japan (protons) ( first visit
already made) - Osaka University (electrons) (first visit
already made) - Oak Ridge (neutrons)
- MSU (neutrons, gammas, deuterons)
- D Develop understanding of mechanism for
radiation-induced demagnetization - Select most radiation-resistant material
- Develop magnet designs with greatest potential
for demagnetization resistance
11Third prong in SLAC/LBNL effort to understand
radiation effects in the NLC
- III Model with a well-used computer modelling
program, FLUKA, the shower of particles and gamma
rays initiated by a single electron of known
energy leaving a bunch at a known angle and
passing through a specified vacuum pipe ( in the
first case conforming to an already designed
copper vacuum pipe in the wiggler of the main
DR), through a specified length of air and then
passing through representative pieces of magnet
material steel and Neo. Validate program by
modelling an exisiting beampipe in SLACs damping
rings and comparing to new measurements. - Using predictions of NLC beam loss percentages,
extrapolate above FLUKA results to predict
radiation levels and types in various NLC
beamlines. - Judge if commercial pm materials will withstand
the predicted NLC radiation types at predicted
levels.
12New data on the electromagnetic quad with
improved magnetic measuring set-up.
13(No Transcript)
14Developing Methodologies to Estimate Overall
Magnet Costs- including repair costs
15(No Transcript)
16Magnet Failures at SLAC Jan. 97 Dec. 2001
SLAC Downtime (Avg)
30
25
20
Downtime (hr)
15
10
5
0
Small
Medium
Large
Magnet Size
SLAC Downtime
200
180
160
140
120
Hours
100
80
60
40
20
0
Small
Medium
Large
Magnet Size
17SLAC Downtime by Failure
180
40
35
30
25
Downtime (hr)
20
15
10
5
0
Insulation
Water Leak
Water
Human Error
Connector
Other
Blockage
Failure Type
18Water Cooled Magnets (Med Large) Failures
19(No Transcript)
20(No Transcript)
21LBNL SLAC work on designing magnets (pms and
ems) for the damping rings.
- RECALL Main Damping Ring lattices have been
published with detailed requirements on all
magnets, including field quality and distances
between adjacent magnets. - Prior to last MAC meeting we designed permanent
magnet versions of DR quadrupoles and transport
line dipoles using 2-D PANDIRA code. The NdFeB
(Neo) style magnets were of reasonable size so
since that meeting we have investigated the Neo
quads, with rotating rods to generate the /-10
adjustability, in more detail to see if they
could meet all the requirements. - LBNL employee Jin-Young Jung learnt how to model
magnets in 3-D using the well-known TOSCA code so
we could estimate the loss of integrated field
strength through any end effects of these magnets.
22PANDIRA 2-D model prediction of the torque needed
to turn one tuning rod in a DR quad.
Very high torques were predicted, 200 lb-in.
Modelling process was validated by modelling the
main linac wedge quad prototype and comparing
predicted torques with measured ones good
agreement. These much higher torques needed in
the larger aperture DR quads, plus the concern
about radiation damage to the pm material led us
to, temporarily, abandon the pm magnets for the
DR and revert to designing electromagnets.
23(No Transcript)
24(No Transcript)
25NLC DR quad as an electromagnetTOSCA model
26NLC DR Gradient Dipole POISSON design
27NLC DR Gradient Dipole TOSCA design. Studying
end effects and fringe fields.
28Latest data on the wedge-style pm quadrupole from
the SLAC rotating coil measurement set-up
29Photo of the wedge01-6 on the SLAC measuring
set-up.
Wedge magnet, secured to V block
Rotating coil read-out
Cooling fan for readout stand
Aluminium V block, secured to granite table
Tuning rods rotation mechanism. One rod only
connected. Smaller gear wheel turned with wrench.
Heidenhain 0.5µm indicator
30(No Transcript)
31(No Transcript)