NuMI Horn Testing

1
NuMI Prototype Horn Final Power Supply Testing

• Magnetic Field Measurement
• Vibration Measurement
• Pulse-to-pulse stability
• Fatigue Lifetime Measurement
• DeBugs
• (Description of Horn for reference)

2
Magnetic HornGeneral Design Features
Outer Conductor
Stripline
B
Inner Conductor
p
I
Spray Nozzle
Focus p toward detector

• Large toroidal magnetic field
• Requires large current, 200 kAmp
• Thin inner conductor, to minimize p absorption
• Water spray cooling on inner conductor
• Most challenging devices in beam design
• Prototype test 1999-2000 to check design

Insulating Ring
Drain
3
Horn Fabrication Precision Welding
4
Prototype Horn at Test Standupstream bell endcap

Features Anodize outer conductor
(corrosion, insulation) Nickel plate inner
conductor (corrosion, fatigue)
5
Horn Power SupplyCapacitor Bank SCRs
Moves as single 11 ton unit Other side
looks identical Supplies 205 kA to two horns
in series Variable 2 ms to 5 ms
half-sin-wave pulse width 1.87 second
repetition rate 120 7.5mF caps.
6
Horn Power SupplySCRs
SCRs (3 of 12)
Stripline
7
Horn Power SupplySCRs
Charging supply and controls/interlock
rack Second charging supply added in series
for some cap-bank and pulse-width
combos Now using one supply at 860 Volts max.
output Problem during operation on circuit
with diodes to protect supply from cap. bank
back-feed in case of fault had to use
higher rating component
8
Horn Field Measurement Main horn field between
conductors
Field between conductors a 1/R as expected
very symmetrical agrees within meas. error with
current
9
Horn Field Measurement in field-free region
through center of horn
2 F/N criterion on flux in M.E. beam
(approximately scaled to 0.85 ms test pulse
from 2.6 ms operational pulse)
Measurement with probe moving along horn axis
Error or fringe field in field-free region down
center of horn is so small that no correction
should need to be put into the Monte Carlo.
10
NuMI Horn 1Vibration Measurements
Why -Confirm ANSYS modeling so believable horn
lifetime estimation. -Horn resonances driven by
transmission line, ground vibrations,
etc.? -Archival data useful if have a
stress/fatigue failure   How -Use laser
non-contact absolute position sensor
eclipseometer. Bandwidth 20kHz. Sensitivity
0.1 nanometer.   Measure at upstream inner
cond. bell, 205 kA, 850 ?sec half-cycle, 1.9
sec rep rate.   Resonances described by a linear
model x ?Ai sin(?it)
exp(-t/?i) Resonance Relative
Decay Source Freq. (Hz)
Amplitude Const. (sec) 1287
-1.0 0.01 Horn Innerconductor 1175
-1.0 0.01 Horn Innerconductor 188
0.05 1.0 Transmission Line 168 -
0.16 0.6 Transmission Line 163
0.16 0.6 Transmission Line -0.30
0.2 Horn Thermal
11
NuMI Horn 1Vibration Measurement on Horn Bell
Endcap
(ANSYS gives 71 mm)
DATA
6 mm
DATA
55 mm
1.17 kHz (ANSYS 1.19 kHz)
Linear Model
Linear Model
2 sec
0.03 sec
12
Pulse-to-pulse Stabilitypeak current and field
monitor
Bdot coil field monitor (sensitive to water
temp) I1 Current in Stripline 1 I2 I3 I4
1
1
Initial warm up about 14 hours of data
13
6 million pulses2/3 of a NuMI Year on horn
prototype
14
Metal Flakes in Horn
Specks of metal collected on ceramic insulator
at bottom of horn - appears to be Nickel Parts
of horn visible through ports, both anodized
O.C. and nickel plated I.C, look like new
would have to disassemble horn to locate
source of flaking No operational problem noted
yet, but indicates water filter important to
prevent possible clogged spray nozzles In
horn, under water Some flakes removed
15
Lesson from PrototypeWater line
A water line overstressed a feedthrough,
which cracked and allowed a drip to
develop (The slow leak would not have been
fatal to a horn in the NuMI
beamline) Have modified the design Old water
tube New strain-relieved water tube
16
Horn and Stripline
Afterthought Unistrut brace between
striplines was added to reduce
vibration. Brace cracked, has been
removed. More noise without brace, but
vibration of horn only marginally
affected
(Temporary temperature probe)
17
Summary of Test
Minor troubles encountered Redesign water
line Beef up charging supply snubber
circuit Very successful test Horn has taken 6
million pulses without breaking Magnetic field
looks great Vibration matches expectations Power
Supply stability is good
18
General Horn System Parameters

• Parameter Horn 1 Horn 2
• Neck Radius (cm) 0.9 3.9
• Wall Thickness, Neck (mm) 4.5 5.0
• Outer Conductor Radius to i.d. (cm)
  14.9 32.3
• Inductance (nH) 685-690
  457
• Resistance (µ?) 208 (meas.)
  lt112
• Average Power from Current Pulse (kW) 17.0
  lt7.5
• Power Flux at Neck (W/cm2) 14.5
  lt4.7
• Temperature Rise at Neck (oC) 22.8
  lt7.1

Note Above heat load numbers are from original
design pulse width of 5.2 msec
19
Mechanical Loading and Analysis

• Mechanical Loading of Horn is the Result of
• - Current pulse thermal expansion from resistive
  heating (peak at the
• end of the pulse)
• - Magnetic forces (peak at the mid-pulse)
• - Beam heating from particle interaction in
  material
• Horn 1 Horn 2
• Inner conductor resistive heating 17 kW
  lt7.5 kW
• Inner conductor beam energy deposition
  kW 0.4 kW
• Outer conductor beam energy deposition 14.5 kW
  5.4 kW
• (1 thick) (1 thick)

Note Above numbers from original design pulse
width of 5.2 msec
20
Mechanical Loading and AnalysisAreas of Highest
Mechanical LoadingValues for 5.2 msec Pulse Width

• US end cap minimum stress before pulse is -1030
  psi maximum stress at mid-pulse is -9020 psi
  mean stress is -5025 psi with an alternating
  stress of 3995 psi Stress ratio R0.11
• Under the above calculated stress levels,
  allowable maximum stress for 107 cycles at endcap
  is 26.5 ksi resulting in fatigue safety factor of
  2.9
• Neck of horn stress at mid-pulse is 4351 psi
  stress at end of pulse is
• -3742 psi mean stress is 304 psi with
  alternating stress 4047 psi Stress ratio R
  -0.86 (Note Negative value of R results in lower
  value of fatigue stress limit)
• Under the above calculated stress levels,
  allowable maximum stress for 107 cycles at neck
  is 15.3 ksi resulting in fatigue safety factor of
  3.5
• Stress in conductor weldment regions is very low
  (ltlt4 kpsi)
• Fatigue data from Aerospace
  Structural Metals Handbook

21
Corrosion ConsiderationsFactors Affecting
Fatigue Life

• Moisture reduces fatigue strength
• For R -1, smooth specimens, ambient
  temperature
• N108 cycles in river water, smax 6 ksi
• N107 cycles in sea water, smax 6 ksi
• Hard to interpret this data point
• N5107 cycles in air, smax 17 ksi
• The above data is motivation for utilizing
  corrosion/encapsulating barrier layer over
  aluminum substrate

22
HornCorrosion Barrier Layer

• I.C. Electroless nickel reasonable corrosion
  barrier properties, non-dielectric, more
  expensive, limited vendor base with large tank
  capacity
• Conducted fatigue test of nickel coated aluminum
  samples at the 107 fatigue limit and compared
  results with equivalent non-coated aluminum
  specimens coated samples survived 1.7x 107
  cycles, non-coated samples failed at 0.6x 107
  cycles
• Use high phosphorus electroless nickel (0.0005 -
  0.0007 thick) on inner conductor and conductor
  supports
• O.C. Anodizing best solution for lower stress
  thick cross-section areas Type III (hard coat
  sulfuric acid, 0.0023), Rc 60-65, dielectric
  strength of 800 V/mil
• Type III hardcoat anodize is selected for outer
  conductor and thick lead in portion of inner
  conductor not suitable for thinner/higher
  stress areas of inner conductor due to
  approximate 60 reduction of fatigue strength
• Provides extra protection against I.C. to O.C.
  short circuit

23
Horn Fabrication Precision Welding

• Single pass, full penetration CNC weld is
  required to minimizing conductor distortion,
  assure repeatability, and control internal weld
  porosity
• Proper cleaning, handling, fixtures, and weld
  parameters are crucial to minimize
• conductor distortion and internal weld porosity
• NuMI approached welding solution via parallel
  paths
• 1) Identify vendor base to subcontract critical
  horn conductor welding
• - Vendor base for CNC TIG welding extremely
  limited and expensive less
• flexible fabrication path than in-house
• - Prototype horn 1 fabricated in this manner
  using Sciaky as prime contractor,
• ANL as subcontractor
• 2) Investigate the development of welding
  capability in-house
• - Have specified, benchmarked, purchased, and
  commissioned a Jetline fully
• automated TIG welding system for producing
  controlled conductor weldments
• - System installed at MI-8 horn facility
• - Long term solution for welding 4 initial horns
  (production and spare horn 1
• and horn 2)

24
Technical Progress Prototype Horn 1 Design Summary

• Conductor Fabrication
• Inner conductor fabricated from 6061-T6 billet
  per QQA 200/8
• Relatively good strength (UTS 45 ksi, YS 40
  ksi, R-1 FS 16 ksi)
• Available in variety of sizes and shapes
• Welds readily
• Relatively good corrosion resistance
• All prototype horn inner conductor parts CNC
  machined by Medco to tolerances better than
  0.002
• Inner conductor welding complete - CNC TIG -
  Overall tolerances held to 0.010 over 133.375
  length (straightness and radial deviation from
  ideal)
• Outer conductor overall tolerances better than
  0.010
• Outer conductor anodized, inner conductor uses
  electroless Ni coating
• Stripline contact surfaces use 0.0005 silver
  brush plating

25
Prototype Horn 1 Design Summary

• Water Seals
• - Total of 64 water seals in horn
• - Utilize EVAC aluminum delta seals on KF style
  flange
• Bolted Connections
• - Utilize TimeSert threaded inserts, pullout
  exceeds 9600 lb. on 3/8 insert
• - As a reference, maximum end wall reaction is
  approximately 4270 lb.
• Current Contact Surfaces
• - Current surfaces have 32 µin finish,
  0.0003-0.0005 silver plate finish
• - Interface clamping pressure exceeds 1400 psi
• - As reference, lithium lens secondary contact
  lead is 5.01 in2 for 6285 Arms
• Prototype horn 1 contact area is 9.2 in2 for
  7250 Arms.
• Corrosion/Erosion Control
• - Outer conductor and thick lead in section of
  inner conductor employs 0.0023
• thick Type III hard coat anodize followed by
  mid-temp nickel seal
• - Inner conductor utilizes 0.0007 thick high
  phosphorus electroless nickel
• Inner Conductor Spider Support Columns
• - Design has been experimentally tested to 36
  million cycles at defections of 0.031
• with 80 lbs. axial preload with no failures