Title: Non-contact Modal Testing of Hard-Drive Suspensions Using Ultrasound Radiation Force International Modal Analysis Conference (IMAC XXIV) February 2, 2006
1Non-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
2Introduction
- 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
3Hard 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
4Hard 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
5Leading 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
6Weaknesses 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?
7Ultrasound 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
8Ultrasound 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
9Experiment 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
10Experiment 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
11Ultrasound 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
12Photos of Setup
13Comparison 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
14High 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
15Selective 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.
16Mode shapes determined using ultrasound excitation
2.0 kHz
7.2 kHz
6.0 kHz
10.8 kHz
17Pair 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
18Selective 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
19Photos of apparatus used for phase-shift
excitation
20Phase-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
21Phase-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
22Phase-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)
23Selective 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
24In 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
25Conclusions
- 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
26Acknowledgements
- 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