Non-contact Modal Testing of Hard-Drive Suspensions Using Ultrasound Radiation Force International Modal Analysis Conference (IMAC XXIV) February 2, 2006

1
Non-contact Modal Testing of Hard-Drive
Suspensions Using Ultrasound Radiation
ForceInternational Modal Analysis Conference
(IMAC XXIV) February 2, 2006

• Thomas M. Huber
• Physics Department, Gustavus Adolphus College
• Dan Calhoun
• Advanced Product Development, Hutchinson
  Technology, Incorporated
• Mostafa Fatemi, Randy Kinnick, James Greenleaf
• Ultrasound Research Laboratory, Mayo Clinic and
  Foundation

2
Introduction

• Overview of hard drives and head-gimbal-assembly
  suspensions
• Non-contact, ultrasound stimulated excitation
• Overview
• Selective excitation by varying focus position
• Selective excitation of neighboring parts
• Selective excitation by phase shift
• In-Situ measurements of suspension vibration
• Conclusions

3
Hard Drive HGA Suspension

• HGA (Head Gimbal Assembly) suspension holds hard
  drive read/write heads
• Read/write head is attached to the flexure
• Flexure can gimbal around dimple
• Head flies over spinning disk surface
• Hinge and load beam provide downward force to
  balance lift from flying head
• Suspension length about 10-14.5 mm , thickness of
  25-100 µm
• Typical width about 4-6 mm

4
Hard Drive HGA Suspension

• Head slider flies about 10 nm above surface of
  the disk
• Scale to macroscopic size equivalent to 747
  flying about 1mm above ground
• For reference a human hair 50µm (or 50,000 nm)
  in diameter
• In operating hard drive, vibration of head
  (pitch, roll, or sway) may cause loss of data
  or head crash
• Instead of damping vibrations, suspensions are
  engineered to have specific vibrational
  frequencies

5
Leading Manufacturer Hutchinson Technology

• Headquarters in Hutchinson, MN (about 50 miles
  west of Minneapolis)
• Manufacturing plants in Hutchinson and Plymouth,
  MN, Sioux Falls, SD, Eau Claire, WI.
• Typical production rate of 14 million suspensions
  per week
• Worldwide market leader of suspension assemblies
  
• Virtually all shipped to other countries for
  integration into hard drives
• Production monitoring involves resonance testing
  of small fraction of suspensions
• Suspension mounted on mechanical shaker for
  excitation (1-20 kHz)
• Laser doppler vibrometer used for non-contact
  measurement
• RD measurements of resonance frequency and
  deflection shapes

6
Weaknesses in current shaker/vibrometer test
protocol

• Smaller hard drives require smaller suspensions
• Requires modal testing between 1 kHz to about 50
  kHz
• Existing mechanical shakers not useful above 20
  kHz
• Fixture modes vibrations of support assembly
  unrelated to suspension
• Use of shaker assembly eliminates possibility of
  in-situ testing of operating hard drive
• Can ultrasound radiation force be used for
  non-contact excitation?

7
Ultrasound Stimulated Radiation Force Excitation

• Vibro-AcoustographyDeveloped in 1998 at Mayo
  Clinic Ultrasound Research Lab by Fatemi
  Greenleaf
• Difference frequency between two ultrasound
  sources causes excitation of object. Detection
  by acoustic re-emission
• Technique has been used for imaging in water and
  tissue
• Recently, we have also used the ultrasound
  radiation force for modal testing of organ reeds
  and MEMS devices in air

8
Ultrasound Stimulated Vibrometry for Suspensions

• Pair of ultrasound frequencies directed at
  suspension
• One ultrasound frequency differs from the other
  by frequency ?f that may be in the audio range
  or higher frequency
• Difference frequency ?f between ultrasound beams
  produces radiation force that causes vibration of
  object
• Vibrations were detected using a Polytec laser
  Doppler vibrometer
• In some experiments, comparison of ultrasound
  excitation and mechanical shaker

9
Experiment Details Dual Element Confocal
Transducer

• 600 kHz broadband (gt100 kHz bandwidth)
• 70 mm focal length 1 mm focus spot size
• Confocal (concentric elements with different
  frequencies)
• Inner disk fixed at f1550 kHz
• Outer ring swept sine
• f2551570 kHz
• Difference frequency of ?f 1 kHz 20 kHz
  Caused excitation of suspension
• Dual beams mean essentially silent operation
  since frequencies only combine at small spot on
  suspension

10
Experiment Details Amplitude Modulated
Excitation
Instead of two transducer elements producing the
two frequencies, an alternate method is an
amplitude-modulated signal to cause excitation

• Dual sideband, carrier suppressed amplitude
  modulated signal centered, for example, at 550
  kHz
• Difference frequency of ?f 1 kHz 20 kHz
  between the two frequency components caused
  excitation
• Better for excitation since entire transducer
  producing the same signal (no need for mixing
  near surface).
• Unfortunately, small fraction of both frequencies
  are combined in transducer, so some audio emitted
  

11
Ultrasound excitation of HGA Suspension

• Goal To determine whether vibrationalresonances
  of suspension can be excitedusing ultrasound
  radiation force
• To simulate an operational disk, end of
  suspension clamped the gimbal head was simply
  supported on flat surface
• Confocal ultrasound transducer usedto excite
  modes from 1 kHz to 50 kHz
• Vibrometer measured resonance frequencies and
  deflection shapesat several ultrasound focus
  positions
• Brüel Kjær mechanical shakerused for
  comparison

12
Photos of Setup
13
Comparison of Shaker and Ultrasound Excitation

• Ultrasound excitation 501 520 kHz swept sine
  and 500 kHz fixed tone (red curve)
• Brüel Kjær mechanical shaker (blue curve)
• Ultrasound excitation reproduces the resonances
  measured using mechanical shaker
• Ultrasound excitation produces a cleaner spectrum
  than shaker
• Shaker has fixture modes (resonances of supports
  or shaker) 2 kHz to 4 kHz
• Ultrasound focused only on suspension, so little
  excitation of supports

14
High Frequency Ultrasound Excitation

• Current resonance testing of suspensions to 20
  kHz
• Limited by 20 kHz upper limit of mechanical
  shakers used
• As suspensions get smaller, desire resonance
  testing up to 50 kHz
• Ultrasound excitation Amplitude modulated swept
  sine with 550 kHz central frequency
• Resonances clearly seen up to 50 kHz
• Should be possible to measure resonances to over
  100 kHz with this transducer

15
Selective excitation Changing ultrasound focus
position

• Ultrasound focus (ellipse of about 1mm by 1.5 mm)
  centered on suspension (red curve) and
  towards edge of suspension (blue curve)

Selective Excitation For ultrasound focus
towards the edge (blue curve), large increase in
amplitude of torsional modes at 6, 10, 13 and 15
kHz relative to the transverse modes at 2, 7, and
16 kHz.
16
Mode shapes determined using ultrasound excitation
2.0 kHz
7.2 kHz
6.0 kHz
10.8 kHz
17
Pair of Unsupported Hard Drive Suspensions

• Suspensions clamped at one end and free at other
  7.25 mm separation
• Transducer mounted perpendicular and behind
  suspensions
• Resonances up to 50 kHz
• 1mm focus leads to little cross excitation
  (focused ultrasound allows selective excitation
  of single suspension)
• Technique may be useful for analyzing suspensions
  before they are separated during manufacturing
  process

406 Hz
4.7 kHz
6.1 kHz
25 kHz
18
Selective Excitation using Phase-Shifted Pair of
Transducers

• Uses a pair of ultrasound transducers to allow
  selective excitation of transverse or torsional
  modes
• Radiation force from two transducers has variable
  phase
• If driving forces are in phase, selectively
  excites transverse modes while suppressing
  torsional modes
• If driving forces are out of phase, selectively
  excites torsional modes while suppressing
  transverse modes

19
Photos of apparatus used for phase-shift
excitation
20
Phase-shifted selective excitation

• Two transducers, each with dual sideband
  suppressed carrier AM waveform
• Trials to date low-cost 40 kHz diverging
  transducers
• Modulation frequency swept from 100 5000 Hz
• Difference frequency Df leads to excitation from
  200 Hz 10 kHz

• Variable phase shift between modulation signal
  applied to transducers
• A 90 degree phase shift in signal results in 180
  degree phase shift of radiation (driving) force

21
Phase-shifted selective excitation

• Adjust amplitudes of two transducers to give
  roughly equal response
• The pair of 40 kHz transducers not exactly
  matched (note different amplitudes near 5 kHz)
• When both transducers turned on simultaneously
  with same modulation phase
• Enhanced Transverse Mode
• Suppressed Torsional Mode

22
Phase-Shifted Selective Excitation of Suspension

• Driving in-phase excites transverse but
  suppresses torsional mode (blue curve)
• Driving out-of-phase excites torsional while
  suppressing transverse mode (red curve)

23
Selective Excitation of Torsional/Transverse Modes

• The maximum amplitude for the transverse modes is
  at angles near 0 degrees, with a minimum near 90
  degrees
• The maximum amplitude for torsional mode is at
  angles near 90 degrees, with minimum near 0
  degrees.
• By shifting the phase by 90 degrees, the ratio of
  the lowest transverse divided by torsional mode
  can change from above 201 to smaller than 13.
• Selective excitation via phase shifted ultrasound
  has been demonstrated for several other types of
  devices, including rectangular cantilevers and a
  MEMS mirrors

24
In Situ Testing For Rotating Disk

• Ultrasound excitation is non-contact and no
  fixture
• Allows for in-situ testing
• May be useful for diagnosing integrated system
  problems
• Red curve Ultrasound off
• Vibration due to windage of flying head
• Blue curve Ultrasound on
• Vibration in excess of windage

25
Conclusions

• Ultrasound allows excitation of resonances and
  deflection shapes
• Completely non-contact for both excitation and
  measurement
• Produces same resonances of suspension as
  mechanical shaker
• Does not excite fixture modes
• Useful for frequencies up to 50 kHz or more
• Selective excitation
• Localized excitation can excite part without
  exciting neighboring parts
• Select transverse/torsional modes by moving
  ultrasound focus point
• Select transverse/torsional modes using phase
  shift between two transducers
• Ultrasound excitation can be used for in-situ
  testing in a hard drive

Ultrasound excitation shown to be feasible for
resonance testing of hard drive suspensions
26
Acknowledgements

• This project includes support from the
• The Gustavus Presidential Research Award program
• Student assistant John Purdham (GAC 06)
• This material is based upon work supported by the
  National Science Foundation under Grant No.
  0509993
• Any opinions, findings and conclusions or
  recomendations expressed in this material are
  those of the author(s) and do not necessarily
  reflect the views of the National Science
  Foundation (NSF)

Thank You