presentation ppt

Report
Superconducting RF
Cavity/Cryomodule Development
at Fermilab
(Industrialization)
C.M. Ginsburg (FNAL)
Proton Accelerators for Science and Innovation
2nd Annual Meeting
Rutherford Appleton Laboratory, UK
3-5.April 2013
Overview
Fermilab
 SRF activity at FNAL/ANL is in support of Project X, ILC, or other future
SRF projects
 Explicitly includes industrial development, and associated R&D for improved
performance and reliability, and reduced cost
 Infrastructure availability and personnel development permit the
development of industrial partners for SRF cavities and cryomodules
 FNAL philosophy: the laboratory does not duplicate activity or compete
with industrial capability
 Industry can provide most materials and services more quickly at lower cost
 Exceptions involve substantial infrastructure, e.g., cryogenic systems
 Most of the industrialization focus has been for the ILC, so this talk will
be weighted toward ILC (electron) technology
 Most conclusions are broadly applicable to other SRF projects
 ...Proton accelerators in particular
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
2
Outline




Fermilab
Cavity vendor development
Cavity processing vendor development
Cavity and cryomodule value engineering
Cryomodule assembly
 Not yet industrialized in the US; XFEL example will be instructive
 Existing industrialization workshops (ILC)
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
3
Bare cavity sequence
Fermilab
Each inspection, processing and test step is recorded in an electronic traveler
For large projects like XFEL, it may make sense to dress cavities before vertical test
5.Apr.
2013
4
Ginsburg
Ginsburg
(FNAL)
PASI Workshop
- NGLS Review
Feb
5-8,2nd
2013
4
Cavity dressing sequence
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
Fermilab
5
CM assembly sequence (part 1)
Receive dressed
cavities at CAFMP9
Fermilab
Receive peripheral parts
Assemble
dressed Cavities
to form a String in
the Cavity String
Assembly Area
(Clean Room)
Install String Assembly to
Cold Mass in the Cold
Mass Assembly Area
Transport the Cold
Mass to CAF-ICB
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
6
CM assembly sequence (part 2)
Fermilab
Install the Cold Mass back to the
Cold Mass Assembly Fixture in
Cold Mass Assembly Area
Align Cavity String
to the Cold Mass
Support
Install the String assembly with
the cold mass into the Vacuum
vessel in the Vacuum Vessel
Assembly area
Ship Completed
Cryomodule to ILCTANML for testing
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
7
ILC cavity: international effort
Fermilab
• ILC cavity fabricators
–
–
–
–
–
–
–
–
Research Instruments (Germany)
Zanon (Italy)
Advanced Energy Systems (US)
Niowave and Roark (US)
PAVAC (Canada/US)
Mitsubishi Heavy Industries (Japan)
Toshiba (Japan)
Hitachi (Japan)
• ILC cavity processing facilities
–
–
–
–
–
DESY
Jefferson Lab
KEK
Fermilab/Argonne joint facility
(Industrial processing facilities: RI, AES, Zanon)
• Results from past 3 years have been collected in worldwide
database, as a means to further track progress and provide input to
ILC machine design
5.Apr.
2013
8
Ginsburg
Ginsburg
(FNAL)
PASI Workshop
- NGLS Review
Feb
5-8,2nd
2013
8
“Up-to-second-pass” ILC Production
Yield Plot - Method
•
Cavity from vendors who have manufactured a cavity that has
surpassed 35MV/m in vertical test:
–
•
•
•
•
•
ACCEL or ZANON or (AES SN>=5) or (MHI SN>=12)
Fine-grain cavity
Use the first successful (= no system problem) test
Standard EP processing: no BCP, no experimental processes
(Ignore test limitation)
Second pass
–
if (Eacc(1st successful test)<35 MV/m) then
•
if (2nd successful test exists) then
– plot 2nd test gradient
•
else
– plot nothing [assume 2nd test didn’t happen yet]
•
–
else
•
–
5.Apr. 2013
endif
plot 1st successful test gradient
endif
Ginsburg (FNAL) 2nd PASI Workshop
9
International cavities from established vendors using established processes
2nd pass yield for >35 MV/m for integrated sample is (57 +- 8)%
for 2010-2012 alone is (69 +- 13)%
5.Apr. 2013
Ginsburg
(FNAL)
2nd PASI
Ginsburg
(FNAL)
2ndWorkshop
PASI Workshop
C.M. GInsburg et al., KILC12, Daegu, S. Korea
http://ilcagenda.linearcollider.org/contributionDisplay.py?contribId=85&sessionId=36&confId=5414
2nd pass
1st pass
ILC Cavity Performance Benchmark
Fermilab
10
Gradient Summary
•
•
•
Good progress worldwide in cavity production, processing, and test
AES has been qualified as an ILC cavity vendor during this activity
Progress is a partnership between industry and laboratories, results are
dependent on both performing well
–
–
•
Utility of XFEL test data for ILC will be limited by XFEL requirements, but huge data set
Efforts to exceed ILC gradient spec will continue
–
•
constant vigilance required afterwards to stay there
Yield statistics to the ILC specification show improvement with time
–
•
Scars, pits, stains, dirt and residue introduced at different steps
Early defects are not typically overcome by the standard processing steps
The typical learning curve at each company requires a ‘few’ cavities
–
•
Fermilab
Field emission prevention at all gradients remains important
Laboratory processing and test facilities are coming up to speed, recent
throughput at Fermilab for instance is very good
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
11
New Vendor Development: PAVAC
Fermilab
 Canadian company with new facility in
Batavia
 1-cells: 6 fine-grain cavities fabricated
– Half use “smart-”bells TE1PAV001-3
• First weld together half-cells, then add
beamtubes
– Half use dumb-bells – TE1PAV004-006
• First weld each half-cell to a beamtube,
then weld together
• “smart bell” cavities exhibited
multipacting ~18-22 MV/m possibly due
to unusual shape
 9-cells: 10 fine-grain cavities were ordered;
order later changed to 650 MHz
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
TE1PAV001
12
New Vendor Development: Niowave-Roark
Fermilab
 Six 1-cells tested extensively from 2008
– BCP/VT @Cornell, some had add’l prep/tests
– Useful information learned, e.g., defect on die
– Primarily being used for commissioning and
materials studies now
 Six 9-cells received
 QC shows fabrication is not yet as stable as other vendors
 Performance is moderate
 Tumbling R&D
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
13
Vendor Surface Processing
Fermilab
• NR flash BCP’d the six 9-cells – insufficient data to comment
• AES flash BCP’d the latest batch of six 9-cell cavities – all
show some pitting but performance is typically good
– Does BCP cause the pitting?
• Process not well controlled, e.g., acid flow too fast
• Pitting worse on lower surface than upper
– Does material cause the pitting?
• Pits re-emerge after tumbling
• R&D on sheet corners anticipated
• RI did bulk EP on half of the latest batch (six of twelve)
– Performance more likely to improve after heavier “light” EP
• So far, no performance advantage, but potential advantage
justifies a controlled promotion of industrial processing
– AES has a new EP machine
– Process was qualified on a 1-cell cavity
– AES to bulk EP six cavities this year
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
14
Industrial Surface Processing
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
Fermilab
15
Cavities Flash BCP’d at Vendor
Fermilab
TB9AES013: Pits observed in all three images, but generally enhanced by EP. Pits are not restricted
to just the equator weld or the heat affected zone.
1) Optical inspection of equator weld before EP
2) Photo before electropolishing.
3) Photo after electropolishing ( ~ 120 microns removed)
1
2
3
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
16
Cavities Bulk-EP at Vendor
Fermilab
RI bulk-EP removal amount (um)
133 153 138 130 152 140
*
*KEK grinding repair
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
17
Value Engineering: FNAL Dressed Cavity
Cavity
Costs
6-13%
Fermilab
Dressed Elliptical SRF
Cavity Fully Burdened
Cost Breakdown
*Fermilab Costs
11-13%
63-71%
11-12%
Second Pass
HPR Reprocess
• Processing is ~ 13% of the cost of a dressed cavity
5.Apr.
18 2013
Ginsburg
Ginsburg
(FNAL)
PASI Workshop
- NGLS Review
Feb
5-8,2nd
2013
18
Value Engineering: FNAL/ANL Cavity Processing
Fermilab
Fully Burdened Cost
Breakdown
*Fermilab Costs
• 26%*13% =4% of the cost of a
dressed cavity is material
removal
• Processing costs dominated
by touch labor
Estimated FNAL Cost Breakdown
5.Apr.
19 2013
Ginsburg
Ginsburg
(FNAL)
PASI Workshop
- NGLS Review
Feb
5-8,2nd
2013
19
Value Engineering: FNAL ILC Cryomodule
Fermilab
15%
43%
26%
ILC Type-3 Cryomodule M&S actual cost
5.Apr.
20 2013
Ginsburg
Ginsburg
(FNAL)
PASI Workshop
- NGLS Review
Feb
5-8,2nd
2013
20
FNAL CM Assembly Throughput
Fermilab
14 days
13 days
9 days
14 days
8 days
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
21
FNAL Cryomodule Assembly
Fermilab
• CAF infrastructure is fully functional for the 1.3 GHz pulsed
cryomodule assembly.
• We have assembled two 1.3 GHz and one 3.9 GHz cryomodules
at CAF. Our experience is still too limited to fully assess each
step of the assembly and make optimization.
• New assembly tooling will be needed to assemble the 325 and
650 MHz cryomodules but the main infrastructure of the CAF
looks adequate to assemble these cryomodules.
• Cavity dressing/qualification and assembly components
preparation for cryomodule assembly will probably require some
automation in order to increase the throughput for future projects.
• Cryomodule assembly throughput requirements will dictate hiring
and training technicians. Training required for CM assembly is
lengthy, especially for cleanroom work.
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
22
ILC Industrialization Workshops
Fermilab
• Two ILC industrialization workshops took place,
with substantial industrial and lab participation
– PAC10 Kyoto satellite meeting
• http://ilcagenda.linearcollider.org/conferenceDisplay.py?confId=4530
– SRF2011 Chicago satellite meeting
• http://ilcagenda.linearcollider.org/conferenceDisplay.py?confId=5182
• Discussion topics: niobium material, cavity
fabrication, industry regional differences, CM
fabrication
• Webpages provide a useful resource
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
23
Fermilab
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
24
Summary
Fermilab
 ILC has provided a great opportunity for US SRF industrial development
 Cavity vendor development
 Cavity processing vendor development
 Cavity and cryomodule value engineering exercises are ongoing for
future projects
 Existing industrialization workshops (ILC) provide a resource for
understanding cost reduction targets
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
25
Acknowledgements
Fermilab
• Many thanks to our Fermilab, national, and international
collaborators for their hard work and excellent contributions to the
cavity and cryomodule development presented here
• Material for this presentation was provided by T. Arkan, J. Kerby*,
A. Rowe (FNAL).
*now at Argonne National Laboratory
5.Apr. 2013
Ginsburg (FNAL) 2nd PASI Workshop
26

similar documents