TQC Design Details and Plans R' Bossert

1
TQC Design Details and Plans R. Bossert
US LHC Accelerator Research Program
bnl - fnal- lbnl - slac
DOE Review November 1-4, 2005
2
TQC objectives

• TQC magnets are technological quadrupole models
  based on the collar-yoke-skin mechanical
  structure
• Design goals
• Achieve Gmaxgt200 T/m.
• Fabricate, test and evaluate 2-layer shell-type
  coil design without internal interlayer splices
• Fabricate, test and evaluate mechanical
  structures based on collar-yoke-skin support.
• Develop and evaluate coil fabrication and magnet
  assembly technologies
• Performance study
• magnet quench performance training, re-training,
  SSL
• field quality geometrical harmonics, coil
  magnetization, iron saturation, alignment, field
  quality correction
• quench protection conductor parameters, quench
  heaters
• Compare TQC and TQS designs, technology and
  performance parameters

3
TQC01 coil

• Coil
• 2-layer shell-type
• Inner-layer wedges
• Inner-layer pole glued into the coil
• Cable
• Strand MJR, 0.7 mm
• Number of strands 27
• Keystone angle 1 deg
• Width 10.077 mm
• Thickness 1.26 mm
• Insulation 0.125 mm S2-glass sleeve
• Identical to TQS coils.

4
TQC01 parameters
TQC01 Specifications
5
TQC design approach

• TQC design is based on the MQXB mechanical
  structure (collar, yoke, skin, end plate, etc.).
• TQC uses available coil winding and curing
  tooling (winding tables, mandrels, presses,
  etc.).
• Mechanical structure, tooling and infrastructure
  exist for 1-2 m long and up to 6 m long magnets.

MQXB cross-section.
6
TQC mechanical structure

• Modified MQXB collar blocks with outer-layer
  poles for coil alignment. Inner-layer poles are
  glued into the coil.
• Radial yoke cut per quadrant to provide
  symmetrical load.
• Control spacers for collared coil alignment and
  yoke motion control.
• Preload shim at each midplane to control
  coil-yoke interference.
• 12 mm thick stainless steel skin.
• Mechanical structure and coil pre-stress are
  being studied and optimized using short (40 cm)
  mechanical model.

7
TQC Assembly

• Impregnated coils are assembled and surrounded by
  layers of Kapton ground wrap.
• Assembly is hung vertically over collaring press,
  and collar packs are placed over coils.
• Collars are incrementally keyed, in 3 inch
  longitudinal sections, applying azimuthal preload
  to the coils of 70MPa after keying is complete.
• Control spacers, preload shims, yoke packs and
  skin are assembled in press.
• Hydraulic pressure is applied and skin is welded,
  applying the fully assembled preload of 140MPa to
  the coil through the preload shims.
• Preload to coils from yoke/skin is limited by the
  control spacers at room temperature.
• During cooldown, parts shrink, allowing preload
  on coils to increase to 150MPa.

8
TQC Mechanical Analysis
Stresses in coil and components derived from
Finite Element Analysis. All stresses are well
below yield strength of material.
Stresses in TQC01 Coils and Components (MPa)
9
TQC Forces when Powered
10
TQC Mechanical Model
Work is underway on a mechanical model to test
these assumptions. A dry run with an
instrumented aluminum tube has been completed.
Strain in the aluminum tube was measured while
the collaring keys were inserted, incrementally,
in 3mm steps until they were fully inserted.
Azimuthal stress in the aluminum tube increased
by approximately 10 MPa per mm of key depth.
Since key depth can be controlled during the
keying operation to about 1mm, the incremental
stress between keyed sections can be controlled
to within 10 MPa.
11
TQC analysis Mechanical Model
Mechanical model assembly with instrumented
practice coils is currently taking place.
12
TQC End Loading
End load is applied by a combination of radial
force through the collars by the skin, and end
force applied by four preload screws, or
bullets through 50mm thick stainless steel end
plates. A total force of 14000N (3000 lbs.) is
applied to each end.
This system is identical to that used for Nb3Sn
dipoles at Fermilab. HFDA06, the most recent
dipole model, was tested with this system and
remained preloaded during all phase of operation.
13
TQ Coil Fabrication Experience

• The TQ coil manufacturing has been very
  successful to date. LBNL and FNAL have
  successfully collaborated on their completion,
  with technicians from both labs present
  throughout the fabrication process.

• 4 practice coils have been manufactured.

• 3 coils made of MJR strand for TQS01 have been
  wound and cured. The 4th is currently being
  wound.

• A complete coil traveler is written and
  incorporated into the fabrication process.

14
TQ Coil Fabrication Experience - Winding
Some issues were identified and resolved during
the practice coil manufacturing process

• Gaps between turns and end parts appeared on
  practice coil 1 due to the necessity of grinding
  parts to place them onto the uncompressed coil
  during winding, a common practice when making
  Nb3Sn coils. This problem has been solved by
  cutting slots into certain end pats to make them
  more flexible.

• There were some instances of de-cabling during
  winding. They were controlled with winding
  techniques, primarily reducing winding tension at
  critical moments, adding a 360 degree twist
  between the cable tensioning device and the coil
  during winding, and changing the system which
  measures tension, allowing the cable to take a
  straight path directly from the tensioner to the
  coil, rather than passing through a series of
  rollers.

15
TQ Coil Fabrication Experience - Winding

• Ramp area between inner and outer coils
  deformed, during curing on early practice coils.
  Practice coil 2, which later exhibited tin
  leaks, showed significant damage. New tooling,
  incorporated in practice coil 4, alleviated
  stresses in this area and eliminated the
  deformations.

Practice Coil 2
Practice Coil 4
16
TQ Coil Fabrication Experience - Curing
After winding, cable insulation is injected with
ceramic binder, then coils are cured at 150C for
30 minutes in a closed cavity mold, subjected to
an azimuthal pressure of approximately 35 MPa.
Curing is done to set the coil size for reaction,
as well as allow the coils to be easily handled,
facilitating insertion into the reaction fixture
without damage.
Curing operation went well. All coils were
measured following curing before reaction.
17
TQ Coil Fabrication Experience - Measuring
After curing, azimuthal measurements of the coils
are taken. Measurements are taken in 3 inch
increments over the entire coil straight section
at pressures of 8, 10,15 and 20 MPa.
Measurements are defined as the arc length of
one side of a coil (one octant), compared to a
steel master of the nominal coil size.
Since the TQ coils have pole pieces potted into
them, it is not possible to measure each side
separately. Coils are measured as shown, and
averaged to obtain the single side measurement.
18
TQ Cured Coil Measurements
Coil measurements are done for several reasons

• To verify coil size and consistency.
• To determine the amount to shim the coil within
  the reaction fixture to ensure that an
  appropriate amount of pressure is being applied
  during the reaction process.
• To feed back information for adjustments in
  future coil sizes to obtain proper preload
  without shimming, optimizing field quality.
• To test for turn-to-turn shorts under load.

19
TQ Coil Fabrication Experience

• The practice coils 1 and 3 were reacted and
  impregnated at Fermilab. After reaction before
  impregnation and after impregnation coils were
  inspected and looked good. Techniques are
  documented and being incorporated into a traveler
  to be used for the real TQ coils.

20
TQ Coil Fabrication Experience

• Pictures of the cut surfaces of the impregnated
  coils have been taken with a microscope, showing
  impregnation to be complete

These pictures demonstrate that epoxy penetrated
inside the cable.
21
TQ Impregnated Coil Measurements

• After impregnation, Practice coils 1 and 3
  were measured and monitored for resistance at the
  same pressures done after curing.

22
Summary of Current Status

• 4 practice coils were successfully wound and
  cured at Fermilab, with the participation of
  technicians from both FNAL and LBL.

• Coils for TQS01 are currently being fabricated.
  The first three coils are complete, with the
  fourth beginning this week. Coils can be wound,
  cured and measured in 8 working days each, within
  the time originally estimated.

• 8 UL of cable for TQS01/TQC01 are insulated and
  on hand at FNAL. The remaining 2 UL have been
  fabricated by LBNL, and will be shipped when
  insulation material arrives.

• 2 practice coils have been reacted, impregnated
  and measured at Fermilab.

• 2 practice coils have been reacted and prepared
  for impregnation at LBNL.

23
Next

• Last two TQS01 coils will be completed and
  shipped to LBNL by the middle of November.
• TQC01 coils and two spare coils will be wound,
  reacted and potted. FY06 Q1/Q2
• TQC01 will be assembled and tested. FY06 Q2/Q3
• TQC02 coils will be wound, reacted and potted.
  FY06 Q2/Q3
• TQC02 will be assembled and tested. FY06 Q3/Q4
• TQC01 will be disassembled and the coils
  delivered to LBNL to use
• in TQE01. FY06 Q3/Q4
• TQC03, another identical 2-layer quad, will be
  built and tested.
• FY07 Q1-Q3