Transcript R&D Progress and Plans Tom Himel For the R&D Board 24 May 07 Cost Review.

R&D Progress and Plans
Tom Himel
For the R&D Board
24 May 07 Cost Review
1
Preface
• There is a large amount of R&D going on.
• There was a 1.5 day MAC review last month
devoted solely to R&D.
• I will now squeeze all that into a 50 minute
talk.
• 500 slides in 50 minutes is 6 seconds a slide.
• In case you don’t catch it all, the MAC slides
are at
http://ilcagenda.linearcollider.org/conferenceD
isplay.py?confId=1388
24 May 07 Cost Review
2
Preface (2 of 2)
• I’ll actually pick and
choose, concentrating
on the large risk*cost
items and highlighting
a few others.
• Otherwise listening to
my talk would be like
trying to drink from a
fire hydrant
• By necessity, some
things will be left out.
24 May 07 Cost Review
3
Contents
• Process
• Goals of the R&D planning
• Major risk mitigation R&D efforts:
–
–
–
–
–
Cavity gradient
Cryomodule at full gradient
Linac string test
E cloud
BDS system test
• A sampling of other R&D
– High power RF
– High availability
• Long term view
24 May 07 Cost Review
4
Process (1 of 3)
• There was ILC R&D going on long before the
GDE was formed.
• The GDE formed an R&D board in early 2006
to coordinate the R&D.
• This board first made a long “ideal R&D list”
and prioritized the items by the needs of the
baseline design and risk involved and the
level of effort needed to execute the task.
• This list was made public and can be found at
http://www.linearcollider.org/wiki/lib/exe/fetch.php?id=rdb%3Ardb_exter
nal%3Ardb_external_home&cache=cache&media=rdb:rdb_external:rd_
master20060407.xls
24 May 07 Cost Review
5
Process (2 of 3)
• The prioritized list was made available to
institutions doing R&D and to funding
agencies.
• Our R&D board was asked to review and
advise several funding agencies (DOE, KEK,
and PPARC) on which R&D proposals they
should fund.
• Our advice was used in addition to regional
concerns to determine the actual budgets.
• We have had a significant influence on R&D
planning.
24 May 07 Cost Review
6
Process (3 of 3)
• More recently we have established a number of
“task forces” to plan and coordinate R&D in a
number of major areas such as cavity gradient,
Damping Rings, BDS, Sources…
• Each has 1 or more R&D board members and
some of the leaders from the relevant system.
• Each operates in the way deemed best for the
problem at hand, varying from detailed
coordination of R&D at labs around the world, to
simply stating what needs to be done.
• The planning is very much a work in progress.
24 May 07 Cost Review
7
Goals of the R&D
• Reduce technical risk of baseline or reduce cost.
• R&D plan and priorities were mostly made before
formal risk analysis that Ewan presented. They still
agree well as same brains made up both.
• R&D mostly concentrates on baseline, some on
alternatives.
• Try to have R&D that has major effects (civil footprint)
in hand 1 year before EDR finish. But funds are often
tight so schedules stretching to near end of EDR.
• There will still be some residual risk when the EDR is
finished. OK, as still time to do more R&D and
engineering as we await funding approval and dig
tunnels.
24 May 07 Cost Review
8
Cavity Gradient – Goal
• Current status: Many 9 cell cavities have been produced
with gradients > 35 MeV/m. Not reproducible and needs
several attempts at final processing
• Goal: After a viable cavity process has been determined
through a series of preparations and vertical tests on a
significant number of cavities, achieve 35 MV/m at Q0 = 1010
in a sufficiently large final sample (greater than 30) of ninecell cavities in the low power vertical dewar testing in a
production-like operation e.g. all cavities get the same
treatment.
– The yield for the number of successful cavities of the final
production batch should be larger than 80% in the first test.
After re-processing the 20 % underperforming cavities the yield
should go up to 95%. This is consistent with the assumption in
the RDR costing exercise.
24 May 07 Cost Review
9
Cavity Gradient – Plan
•
There are three main activities which are closely coupled and partially
progressing in parallel
– This is needed to separate cavity preparation and production issues
•
•
1. Single-cell R&D
– Establishing more reliable final preparation parameters.
– Focus on the final rinse after EP before HPR.
– E.g. Ultrasound, Short EP (or HF rinse), H2O2
2. Tight-loop
– International multi-cell cavity exchange
– 1st round
• Comparison of regional differences in preparation and testing
– 2nd round
• Use single-cell results and implement on 9-cell cavities.
•
3. Production-like effort
– Monitor ongoing productions
• Esp. XFEL preparation
• Use qualified and new vendors
– Use improved preparation process for an ultimate batch of cavities
•
A lot of data will be (is already) available by the time of the EDR writing
24 May 07 Cost Review
10
Cavity Gradient – Cost/Benefit
• Optimistic scenario with final batch + tightloop
– Costs 36 MILCU for the R&D
– Gives highest confidence about the gradient
distribution
• This needs to be compared to:
– A reduction of the average gradient for the ILC
from design of 31.5 to 28
– ~ 600 MILCU
24 May 07 Cost Review
11
Cavity Gradient – Results
2005
2007
24 May 07 Cost Review
2006
KEK single cell results:
2005 – just learning
2006 – standard recipe
2007 – add final 3 μm fresh acid EP
Note: multi-cells are harder than singles
12
Module Test – Goal
• Intermediate goal
– Achieve 31.5 MV/m average operational accelerating gradient in
a single cryomodule as a proof-of-principle. In case of cavities
performing below the average, this could be achieved by
tweaking the RF distribution accordingly.
– Auxiliary systems like fast tuners should all work.
• Final goal
– Achieve > 31.5 MeV/m operational gradient in 3 cryomodules.
– The cavities accepted in the low power test should achieve 35
MV/m at Q0 = 1010 with a yield as described above (80% after first
test, 95% after re-preparation).
– It does not need to be the final cryomodule design
24 May 07 Cost Review
13
Module Test – Plan
• Enough good cavities for the cryomodules are expected from
the cavity gradient program. Module assembly plans:
• DESY
– 2007: M7: Being tested now. See next slide
– 2007: M8: Probably no slow-down to select best cavities
– 2008: M10 – could select best cavities from several regions
• US
– 2007 – Assemble a kit of parts from DESY to get first
assembly experience at FNAL
– 2008 – assemble 2 cryomodules from US produced
parts. Second may be made by selecting the
best available cavities.
– 2009 – build 2 more cryomodules
• Japan
– 2009-10: Build, test, 3 cryomodules
24 May 07 Cost Review
14
Module Test – Results
24 May 07 Cost Review
DESY
15
Module Test – Results
DESY
24 May 07 Cost Review
16
String Test – Goal
• Build 1 RF unit (3 cryomodules + 1 Klystron) to fully
check:
– What gradient spread can be handled by LLRF system.
This test should be done with and without beam loading.
– For heating due to high frequency HOMs.
– Amplitude and phase stability.
– Static and dynamic heat loads.
• To partially check:
– Reliability
– Dark current
– for degradation or other weaknesses
• The ILC cryomodule is enough different than that of
the TTF that a new system test is warranted.
24 May 07 Cost Review
17
String Test – Plan
• Use cryomodules built for module tests and for
industrialization.
• Build 1 RF unit at KEK and 1 at Fermilab.
• Do this in a phased manner, starting with smaller
tests with modules that don’t meet specs.
• Full – to spec – RF unit should work before 1%
of the final industrial production of ILC
cryomodules is complete.
• There will be a larger second phase string test to
verify quality of the modules going into the ILC.
24 May 07 Cost Review
18
String test – Cost/Benefit
• The risk if we don’t do the string test is that
we will build ~1.5 BILCU of cryomodules and
then discover a design flaw.
• Fixing them all could take years and easily
cost more than 20% of the original cost.
• If there is a medium risk (25%) of this type of
error then the risk*cost ~75 MILCU plus the
loss of a few years in schedule.
– Note the risk would be high 50-100% if not for
the TTF.
• The planned string tests will cost over 50
MILCU.
24 May 07 Cost Review
19
E cloud – Goal
• Ensure the e- cloud won’t blow up the e+
beam emittance.
– Do simulations (cheap)
– Test vacuum pipe coatings, grooved
chambers, and clearing electrodes effect on ecloud buildup
– Do above in ILC style wigglers with low
emittance beam to minimize the extrapolation
to the ILC.
24 May 07 Cost Review
20
E cloud – Plan
• Grooved chambers and special coatings are
being tested in PEP-II and KEK-B straight
sections.
• Lots of simulations have been done and
bench-marked against existing accelerators.
– Still, the long extrapolation leaves us nervous.
• Plans are being developed to test special
chambers in wigglers in CESR and KEK-B.
– The funding is not yet assured for these most
definitive tests.
24 May 07 Cost Review
21
E Cloud – Cost/Benefit
• If we don’t do the R&D, there is a high (50%)
risk that we have to build a second e+ DR at
a cost of 200 MILCU. Cost*risk = 100 MILCU
• The first 2 types of R&D cost only a few
million.
• The costs of the KEK-B and CESR tests are
difficult to evaluate as they involve the
dedicated use of the whole ring and it is
unclear which costs should be accounted to
the ILC. The scale is 10 MILCU.
24 May 07 Cost Review
22
E Cloud – Results
SLAC
24 May 07 Cost Review
23
BDS System Test – Goal
• Build ATF2, a scaled BDS prototype at KEK
to test:
– Optics design including never before done
local chromatic correction
– Keeping a beam small (35 nm) and stable to a
few nm for days at a time
– Laser wires
– Intra-train feedback
– BPM’s
– High availability power supplies
– Tuning algorithms
24 May 07 Cost Review
24
BDS System Test – Plan
• ATF2 is already under construction by a multiregional collaboration.
• Will be commissioned in 2009 with optics
tests done in 2010
24 May 07 Cost Review
25
BDS System Test – Cost/Benefit
• Cost is ~5 MILCU
• Ameliorates a medium (25%) risk of having to
do a major BDS redesign that could lengthen
the BDS and cost 200 MILCU extra
• Would be a bargain at twice the price
24 May 07 Cost Review
26
High Power RF – Goals
• The baseline HPRF design is mature and has
very little risk.
• The R&D concentrates on cost reduction
– A Marx modulator to replace the bouncer
modulator
– Modified RF distribution system
– Sheet-beam klystron to replace multi-beam
klystron
• If all are used, the HPRF cost is cut in half.
24 May 07 Cost Review
27
Marx Modulator – Results
SLAC
24 May 07 Cost Review
28
Marx Modulator – Results
100kV Output –1400 μsec, Leveled
SLAC
Long term test planned in coming year.
24 May 07 Cost Review
29
Modified RF Distribution System – Plans
SLAC
24 May 07 Cost Review
30
Sheet Beam Klystron – Plan
• Build beam tester and klystron by Summer
2008
• The beam tester will validate 3-D beam
transport simulations and allow a more rapid
turnaround for electron gun changes
• The klystron will be developed in parallel with
little feedback from the beam tester. A rebuild
of the klystron can incorporate design
changes motivated by the beam tester
24 May 07 Cost Review
31
Sheet Beam Klystron – Plan
24 May 07 Cost Review
32
High Availability – Goals
• Optimize overall ILC design to maximize the
availability while minimizing cost. Do this by
trading off redundancy vs. quick access to
components vs. energy overhead …
• Design and test problematic components to
get a long MTBF and a short MTTR
• Components that need new designs anyway
should get extra attention to making them
reliable.
24 May 07 Cost Review
33
High Availability – Plans
• DR kicker pulser design is modular so they
keep running when a part breaks.
• The Marx modulator has extra modules for
same reason.
• Power supplies for ATF2 are prototypes for
ILC and need only 4 of their 5 regulator cards
to function.
• Controls is testing commercial ATCA
hardware and prototyping ILC specific boards
for this standard high availability standard.
24 May 07 Cost Review
34
Long Term View
• Linear Collider R&D has been going on for 30+
years
• A lot of progress has been made, turning a concept
most thought was impossible into one where we are
fine tuning the design to minimize the risk and cost.
• We are reaching the point of diminishing returns
and need to use a larger fraction of our resources
on development and engineering.
• This is an important but difficult step as we must
change ingrained habits.
24 May 07 Cost Review
35
Summary
• R&D is being coordinated internationally.
• The largest R&D projects are addressing the
largest risks.
• Some R&D is also being done on nonbaseline designs which can reduce costs.
• This talk has left out far more R&D projects
than it has included. Hopefully you were able
to drink up what I did include.
24 May 07 Cost Review
36