Abstract

Mechanical Engineering Issues

Sushil Sharma, NSLS-II Project
The NSLS-II storage ring requires multipoles and dipoles of unprecedented field quality in order to achieve its
performance goals of high brightness and beam stability. Many mechanical engineering issues were encountered in
producing the magnets of such high field quality within the constraints physical envelops and interface requirements.
These mechanical engineering issues are described as they were encountered during the various phases of the project,
namely, lattice design, conceptual designs, specifications, interface requirements, magnet production and assembly,
and magnet transportation. Some technical solutions and recommendations that were developed in resolving these
issues are also discussed.

*Work performed under auspices of the United States Department of Energy, under contract DE-AC02-98CH10886

BROOKHAVEN SCIENCE ASSOCIATES

Mechanical Engineering Issues

Sushil Sharma
ME Group Leader ASD/NSLS-II
April 11-12, 2012
2

BROOKHAVEN SCIENCE ASSOCIATES

Outline
Mechanical design, engineering and production issues related to:

Lattice and magnet types

Conceptual designs
Specifications
Interfaces: girders, utility, vacuum chambers
Magnet Production: stacking and bonding
Magnet Reassembly
Magnet Transportation

BROOKHAVEN SCIENCE ASSOCIATES

Lattice and Magnet Types

NSLS-II Lattice
The Lattice should be finalized

early. Sufficient spaces should be

allocated between the magnets.
The magnet types should be
optimized. Multiple magnet types
impact on:

Design effort
Contractual effort
Suppliers support
Girder assemblies
Schedule remediation plans
Spares
4

BROOKHAVEN SCIENCE ASSOCIATES

Conceptual Designs - 1

Segmented-Pole Construction

NSLS-II Segmented-Pole Designs, December 2006

Segmented-pole , or half-yoke constructions

Half-yoke construction for sextupoles requires
removable mid pole.
Meeting field quality requirements with removable mid
pole has been challenging.
The magnets must have sufficient iron to avoid
magnetic saturation and to make them structurally
robust.

NSLS-II Removable Midpole Design, June 2008

BROOKHAVEN SCIENCE ASSOCIATES

Conceptual Designs - 2

Secondary machining (EDM, milling or profiling) should be a

requirement even if the magnets are procured as per
specifications.
Soleil Magnets were precision stamped but mating surfaces for
some magnets had to machined precisely (grinding).
The inclusion of nose pieces (solid or laminated) and end plates
need to be clearly specified and taken into account in terms of
magnetic length, field quality, field matching and mechanical
interferences.

Quadrupole Nose Pieces

35-mm Dipole
6

BROOKHAVEN SCIENCE ASSOCIATES

Magnet Specifications Issues

A lack of clarity on magnet specifications affects design, production,
testing, acceptance, and leads to unnecessary paperwork (DVRs, DRs,
Waivers, etc.)

Missing tolerances on specifications (max, min, range).

Missing or imprecise specifications :
Reproducibility (definition, first-articles/production).
Magnet-to-magnet variation of yoke lengths.
Current (I) range for field harmonics.

Specification on derived rather than primary quantity (integrated field vs

magnetic length).

Over-constrained specifications (pressure drop, flow, temperature rise)

Stringent specifications (thermal switch resistance, natural frequencies).
Specifications not enforced (weight density of the yokes).
Avoid using or equivalent in the specifications.

BROOKHAVEN SCIENCE ASSOCIATES

Vibration Specification - 1

Horizontal Motion

Magnets stability specification 25 nm (rms, 4-50 Hz)

Girder-assembly natural frequency should be > 30 Hz.
The girder natural frequency is affected by the magnets
natural frequencies.

BROOKHAVEN SCIENCE ASSOCIATES

Vibration Specification -2

The girder-assembly resonant frequencies do not change

significantly when n (magnet) > 2 n(assembly).

BROOKHAVEN SCIENCE ASSOCIATES

Water Flow Specification

(6-10 oC) temperature-rise specification with an inlet temperature of

29.4 oC insures heat removal by cooling towers.
Only one supplier provided a flow restrictor to meet the (6-10 oC)
temperature-rise specification.
The flow restrictor started clogging within a few hours. It was decided
to use thin tube flow restrictors between the magnets and the magnet
headers. The temperature rise specification was waved.
The hole in the copper conductor could have been smaller

Flow
Restrictors
Flow Restrictor Location
10

BROOKHAVEN SCIENCE ASSOCIATES

Girder Interface

Base Plates and Mounting Hardware

Pre-assembly of a Girder

The interface to the girder (magnets baseplate, size and

locations of the mounting holes, slots for water
connections) need to be precisely defined in the
interface control drawings. Frequent interference checks
are necessary.
Base plates and mounting hardware should take into
account alignment, stability and transportation
requirements ( 4 oversize rods, flat nuts, small gap
between the nuts).
11

Base Plate, Corrector

Magnet
BROOKHAVEN SCIENCE ASSOCIATES

Utility Interface
Manifold connections
for dis-assembling a
quadrupole.

Recommendation of Accelerator Advisory

Committee: No forces should be applied to
the magnets (such as in making water
connections) after the magnets are aligned
by vibrating wire.
Girders are equipped with magnet headers.
The manifold material should have been
uniquely identified (PEEK).
APS Water Connections
12

BROOKHAVEN SCIENCE ASSOCIATES

Magnet Vacuum Chamber Interface - 1

~ 280 mm

Beam Aperture
25mm x 76 mm

~ 3m
Exit Port

Machined Profile

Ante Chamber

Machining of the chambers grooves:

2-mm gap between the pole faces
and the chamber.
4-mm gap between the coils and
the chamber.
10-mm extra length on each side.
Can vacuum chambers be designed
with ante chamber on the in-board side?

13

Installed Vacuum Chamber

BROOKHAVEN SCIENCE ASSOCIATES

Magnet Vacuum Chamber Interface - 2

Exit Port

Girder No. 6

Girder No. 4
C-Type magnet spacers are not
optimum for magnet re-assembly.
A better spacer design is possible by
optimizing the design of the exit port:
Reduce the diameter of the exitport tube.
Reduce the gap between the exitport tube and the spacer blocks.
Simplify exit port of chamber no. 6.

Exit-Port Tube

14

LBL - New Sextupole

BROOKHAVEN SCIENCE ASSOCIATES

Stacking Fixture and Bonding Issues - 1

Ref: C. Spencer, Magnet Lecture 2, Guanajuato, 2012

FE Analyses Pressure in psi

Separately Moving
End Plate

Stacking fixture should be able to apply

uniform pressure (3 6 MPa) on the stack of
laminations. 3 MPa is the recommended
pressure, but a higher pressure results in
better pressure uniformity.
The stacking fixture without tie rods allows
more flexibility in distributing the pressure.
The stack should be compressed after each
bundle of ~ 200 mm of laminations.
15

Equal Tension on 9 Tie Rods ( )

Tension in Central Rods = 3 x Outer Rods

BROOKHAVEN SCIENCE ASSOCIATES

Stacking Fixture and Bonding Issues - 2

Stack shrinkage (~ 5 mm/m)

in oven due to compression
and oozing out of the epoxy
coating (~ 7 m thick). The
Belleville washers stacks
Belleville Washers - Stacking
should be designed to take
this shrinkage into account .
Uniform pressure is difficult
Load-Deflection Curves
to achieve in a curved dipole.
For epoxy curing,
Dipole Stack Thermal Analysis
ThyssenKrupp recommends
a soak time of 2 hours at
160C, or 1 hour at 180C. A
thermal analysis is
recommended to establish
the expected temperature
profile in the yoke.
FE Analyses: by V. Ravindranath

16

BROOKHAVEN SCIENCE ASSOCIATES

Magnet Reassembly - 1

NSLS-II Wide Quadrupole

NSLS-II Wide Sextupole

Mating Surfaces

Reassembly specifications should be clearly defined

and verifiable.
Avoid bending deformations and frictional sliding. The
spacer blocks should be under compression.
Avoid multiple mating surfaces.
Mating surfaces should be machined precisely.
Preferred Spacer Block
17

BROOKHAVEN SCIENCE ASSOCIATES

Magnet Reassembly - 2
Bolting Sequence and Specifications:

NSLS-II magnets experience

higher deformations under a
horizontal force than a vertical
force.
Spacer blocks will prevent
deformations under a vertical
force.
Applying vertical force first will
increase the friction between
horizontal mating surface.
By keeping horizontal force onehalf the vertical force, the
probability of horizontal sliding is
reduced.

FE Analyses: by
V. Ravindranath

Lifting from Top Lugs

Recommended Bolt Torques (NSLS-II):

Vertical: 50-60 N.m
Horizontal: 25-30 N.m
Anti-seize lubricant such as Kema RG-100

Design the interface without

horizontal bolts !
18

BROOKHAVEN SCIENCE ASSOCIATES

Magnet Reassembly -3
NSLS-II Wide Sextupole

NSLS-II Normal Sextupole

NSLS-II Quadrupole (BINP)

Magnets should be structurally robust to insure

reproducibility of the magnetic field.
C-clamps were removed from NSLS-II normal
sextupoles they adversely affected the
repeatability.
FE analyses showed that the spacer bars can
significantly stiffen the magnets, thus improving
reproducibility.
FE Analyses: by V. Ravindranath

19

BROOKHAVEN SCIENCE ASSOCIATES

Magnet Transportation - 1
Smax = 0.36 MPa

Gravity Stresses

Smax = 0.34 MPa

FE Analyses: by V. Ravindranath

Additional Plates 27 mm Thick

Micro-fissures were discovered in the yokes of the 35

mm dipoles.
Under its self-weight the maximum normal tensile
stress (Smax ) = 0.36 Mpa.
Adding extra top and bottom plates were not
effective.
The magnet feet are removed during transportation.
The maximum change in the pole gap, ~ 10 m.

20

Gravity Deformations

Gap

BROOKHAVEN SCIENCE ASSOCIATES

Magnet Transportation - 2
Smax = 0.46 MPa

Smax = 0.14 MPa

FE Analyses: by
V. Ravindranath

Lifting from Top Lugs

Lifting from Bottom Lugs

Lifting the magnets from top lugs produced less stresses than
the less-optimally placed bottom lugs.
Bottom lugs were subsequently removed.

21

BROOKHAVEN SCIENCE ASSOCIATES

Team Work

BINP Directors Office

Roller Machine at BINP

BINP Workshop No. 8

Dipole Lamination Peeling-Off

22

BROOKHAVEN SCIENCE ASSOCIATES

Acknowledgement

Thanks to all magnet teams.

23

BROOKHAVEN SCIENCE ASSOCIATES