© 2005 ASM International. All Rights Reserved. www.asminternational.

org
Medical Device Materials II (#05107G)

High Strain Accelerated Fatigue Failure Testing of NiTi Implantable Devices

B. Shuman, C. Finger
Spiration Inc, Redmond, WA
Ken Perry
EchoBio, Bainbridge Island WA

Abstract Introduction

Accelerated testing, using frequencies higher than clinically
observed, has been used to reduce time and cost of fatigue
testing. Unfortunately, changes in a device design or
manufacturing process may require repeating fatigue testing to
show that the change does not adversely affect the specified
device lifetime.
OBJECTIVE: There were three main objectives of this study:
To test implantable Nitinol (NiTi) devices in an accelerated
fatigue failure test at frequencies and strains higher than seen
clinically; To investigate the use of this testing to decrease
fatigue testing time; To investigate the use of higher strains as
a screening tool for design or processes that may affect fatigue
lifetime.
METHODS: Different groups of NiTi devices were built with
different processing parameters. Quality of electropolish,
radius of curvature and pre-test strain were varied and tested
against a control group to obtain fatigue lifetime or Strain-
Number of cycles to failure (S-N) curves. The resulting S-N
curves were generated for each group of data. The S-N curves
were overlaid to determine the relative difference in fatigue
life between a test group and its control group.
RESULTS: The divergence of the two fatigue lifetime (S-N)
curves reveals which processes or designs produce devices
with longer fatigue life. For certain testing groups, optimal
fatigue testing strain values can highlight the differences
between groups better than other fatigue testing strain values.
Accelerated high strain testing produced results in 1 day of
testing while clinically equivalent strain testing took 28 days.
CONCLUSIONS: High strain fatigue testing can produce
relevant fatigue data much faster than testing at clinically
equivalent strains. Optimal strains for determining the
difference in fatigue life between processes or designs can be
determined by overlaying the S-N curves for two or more
groups of samples. Future design and process changes or
improvements can be tested using an optimal alternating strain
based on the divergence of the S-N curves. The results of the
testing could be used to determine if the new design or process
has compromised the fatigue lifetime of the device.
Medical device implants experience cyclic strains resulting
from the in-vivo conditions in which they are implanted.
Accelerated fatigue testing is commonly used to determine the
fatigue resistance of a medical device design and, by
definition, utilizes frequencies higher than clinically observed.
These tests usually subject the test samples to clinically
equivalent strains and take weeks to complete. The results
typically show if a design “passes” the specified fatigue life
rather than showing what the fatigue life of the design actually
is or how it compares to fatigue life of other designs. The
objective of this study is to investigate the use of accelerated
fatigue testing in combination with strains higher than
clinically observed. The study includes the investigation of
optimal mean and alternating fatigue strains for use in
generating fatigue data used to discern the difference between
different designs or processes. Fatigue life curves are
generated for samples from the different designs and processes
and compared to determine if the groups have different fatigue
lifetimes.
Methods:
Testing was performed on NiTi anchors of the Spiration®
Intra-Bronchial Valve (IBV™) Device shown in Fig. 1
(www.spirationinc.com). These devices are intended to be
used in human airways to block the flow of air to diseased
areas of the lung tissue.
The samples tested were processed with different properties.
Quality of electropolish (EP), radius of curvature and pre-test
strain were varied to obtain different testing groups. Each
group was fatigue tested along with a control group to obtain a
fatigue lifetime curve.
© 2005 ASM International. All Rights Reserved. www.asminternational.org
Medical Device Materials II (#05107G)

Fatigue Testing the IBV devices:

Testing was performed using a high frequency linear motor

and custom data acquisition software. The samples were
fixtured such that the anchor tip aligned with and was parallel
to the axis of movement. Each anchor rested against one of
ten actuator pads that were fixed to the motor. Electrical
current was run through the anchor and actively monitored to
determine when the sample had failed. Figure 3 shows how
the devices were fixtured against the actuator.

Figure 1: The Spiration® Intra-Bronchial Valve (IBV™). The Figure 3: Test setup showing 10 different devices being tested
arrow points to one anchor
Preliminary testing was performed to determine the optimal
Range of deflection and strain test frequency, mean strain and alternating strain. Tests were
The mean fatigue strains were chosen to mimic in-vivo mean run at different frequencies while contact between the part and
strain as closely as possible. Figure 2 below shows a section actuator was monitored. The contact between the sample and
of the force-deflection curve of the IBV Device that was the actuators was verified in two ways. First the voltage
chosen for the operating range of fatigue testing. across the part was monitored. If the voltage was erratic, then
the sample was jumping off of the actuator pad. Second, a
strobe light was used to confirm that the samples were actually
moving at the designated frequency. If the anchor was not
Force-Displacement Curve contacting the actuator, the frequency was faster than the
response frequency of the sample, and was reduced.
0.0006 Fatigue testing was then performed by cycling samples at
various mean and alternating deflections in 37 ºC air at 40 Hz.
0.0005
Relative Force

1 Optimal mean and alternating strains were determined based

0.0004 on the differences between the S-N curve of the test groups.
2
0.0003 Using the optimal testing parameters, tests were performed to
3 determine the sensitivity of the test method. Test sensitivity
0.0002
4 was determined by testing groups that were hypothesized to
0.0001 have different fatigue lifetimes, but had no observable
0 difference in surface characteristics under SEM.
0 0.5 1 1.5 Finally, the relative difference in fatigue life between the
sample groups and their control groups was tested.
Displacement (mm)
Results and Discussion
Figure 2 Force deflection curve for the fatigue test, N=4.
Actual force values were removed for proprietary reasons Samples that were EP and samples that were not EP processed
had significant differences in surface characteristics under
An FEA model was used to relate the fatigue testing SEM. However, samples with different pre-test strain could
conditions to the estimated material strains. Based on the FEA not be differentiated under SEM.
and material tensile tests the fatigue test deflections resulted in
strains spanning between the austenite and martensite phase of
the material. Therefore, a non-linear FEA solver, ABAQUS,
was used.
© 2005 ASM International. All Rights Reserved.
Medical Device Materials II (#05107G)
Figure 4. SEM image representing all Electropolished
sample, both pre-test strained and non-strained samples
Figure 5. SEM image of a sample that was electropolished ½
the normal time
Figure 6. SEM image of sample without electropolish
www.asminternational.org
The samples from the pre-test strain group were strained to
7% for 5 cycles prior to fatigue testing. The control group was
strained to 7% once prior to fatigue testing. When the pre-test
strained samples were fatigue tested, they lasted longer than
the control group. The differences are more detectable at
higher alternating strains and less detectable at lower
alternating strains as shown in Fig. 7.
2.0%
1.8%
1/2 Amp. Alt. Strain
1.6%
1.4%
1.2%
1.0%
0.8%
0.6%
0.4% Control
0.2% Pre-Test Strained
0.0%
1E+3 1E+4 1E+5 1E+6 1E+7 1E+8
Cycles to failure
Figure 7. S-N curves for Pre-test strained (strained 5 times)
samples versus control samples (strained once). Open
markers indicate samples that did not fail.
Statistically, the groups were indistinguishable at lower
strains, however, at high ½ amplitude alternating strains
(1.7%), the T-test p-value was 0.03. Thus, the results of the
test sensitivity testing showed that the test was sensitive
enough to show a difference in fatigue life when SEM
observation could not. This conclusion was based only on the
more conservative, low cycle fatigue data. The apparent bi-
modal behavior of this group of data was not investigated.
Testing the Effect of Bend Radius
Testing was performed on samples that had a slightly different
geometry. The geometrical difference was related to the bend
radius of the part at location of the highest strain. It was
hypothesized that the geometry could be correlated to fatigue
resistance due to differences in strain concentrations. The
data, which was acquired at an effective mean strain of 1.4%
and an effective ½ amplitude alternating strain of 1.4%, is
displayed in Fig. 8.
© 2005 ASM International. All Rights Reserved. www.asminternational.org
Medical Device Materials II (#05107G)

Fatigue life of different radius IBV anchors S-N Curve for EP vs 1/2 Time EP and no
vs. Bend Radius EP
2.0%
2.53

1/2 Amp. Alt. Strain

2.03 1.5%
Radius (mm)

1.53
1.0%
1.03
0.53 0.5% EP
0.03 1/2 EP
No EP
1000 10000 100000 0.0% Series3
Cycles to failure 1E+3 1E+4 1E+5 1E+6 1E+7 1E+8
Cycles to Failure (n)
Figure 8. Regression analysis of bend radius and cycles to
failure, R2<0.01 Figure 9. S-N curve for EP, ½ time EP and No EP devices.
Open markers indicate samples that did not fail.
The results of this data show no correlation between bend
radius and cycles to failure. Prior work1 2 3 4 has shown Electropolishing has been shown to reduce micro cracking
martensitic phase transformation to increase fatigue resistance thus it was expected that the fatigue life would increase
under certain circumstances. It is possible, however not significantly. Accordingly, Fig. 9 shows the difference in
investigated here, that the decreased bend radii increased fatigue resistance between devices that were electropolished
fatigue resistance by inducing local martensitic phase and devices that were not. The difference in fatigue life is
transformation at the point of highest strain. Strain induced shown at different alternating strains. The difference between
martensite may play a more significant role in increasing the samples that have been electropolished for ½ the amount of
fatigue life of samples with a smaller radius. Samples with a time as the EP group can also be differentiated, however, the
larger radius do not have the martensite advantage, however, results are more ambiguous at higher strains.
they have less strain. Martensite strain transformation could
be “leveling the playing field”. The S-N data shown in Fig. 9 was used to test other groups of
non-EP samples at the optimal ½ amplitude alternating strain
Testing the Affect of Electropolish of 1.1%. Figure 10 shows the relative difference in fatigue life
Preliminary testing was performed to determine the optimal between three groups of samples using an effective ½
mean and alternating strain that would best discern the amplitude alternating strain of 1.1% and a mean strain of
differences between electropolished and non-electropolished 2.7%.
sample groups. High alternating strains and low mean strains
resulted in larger differences between the groups. This is
likely due to the geometry of the IBV device. Therefore, an
effective mean strain of 2.7% was selected for the fatigue 2E+4
Cycles to failure (n)

testing. p=0.004
p=0.006
The next set of testing was run at the selected mean strain 1E+4
while the alternating strain was varied for three groups: EP, ½
time EP and no EP samples. Figure 9 shows the resulting S-N
curve for the three groups of data. 5E+3

0E+0
Non-EP EP Non-EP 2

Figure 10: Summary of fatigue life for electropolished

samples versus two different non-electropolished sample
groups. Error bars represent 1 standard deviation. N=6 each
group
© 2005 ASM International. All Rights Reserved. www.asminternational.org
Medical Device Materials II (#05107G)

Fig. 9 contains more information about the different sample References:

groups than Fig. 10, however, selecting one alternating strain
1
allows testing to be more concentrated and statistically Harrison WJ, Lin Z.C.: The Study of Nitinol Bending
significant. Thus, Fig. 10 shows with statistical significance Fatigue, SMST 2000, 391
that the fatigue life of the IBV device is superior after an EP
2
process. This data took 1 day to acquire while a clinically Pelton A., Gong X, Duerig T.: Fatigue Testing of Diamond-
equivalent strain test would have taken 28 days. Furthermore, Shaped Specimens, SMST 2003, 293
the clinically equivalent strain test would have likely resulted
3
in no failures, making it even more difficult to understand the Morgan N.B., Painter J., Moffat A.: Mean Strain Effects and
differences in fatigue lifetime between the groups. Microstructural Observations During In-vitro Fatigue Testing
of NiTi, SMST 2003, 303
4
Conclusions Tabanli R, Simha H, Berg B,: Mean stress effects on fatigue
of NiTi, Material Science and Engineering A273 (1999) 644
Sensitivity evaluations shows that this testing can more
accurately discern the difference between samples that have a
reduced fatigue life and samples that do not when compared to
SEM surface analysis.
The results of different radius of curvature testing shows that
different radius of curvatures do not correlate to fatigue life
for the range of radii tested.
These results show that samples not electropolished or with
poor electropolish finish have a significantly reduced fatigue
life.
Finally, this work has shown that high strain fatigue testing
can be used as a screening tool to screen designs quickly.
Different geometry or radius of curvature, different amounts
of electropolish or different amounts of pre-test strain are
examples of designs or processes that can be quickly evaluated
for their affect on fatigue life.

Recommendations

These methods were not validated to clinically relevant

conditions, therefore it is recommended that validation of
these methods be completed prior to extrapolating results to
clinically relevant conditions.
It is also recommended that future work include measurement
or control of internal sample heating to understand and
account for this affect on fatigue life or strain. However, the
sample groups in this study were tested at the same boundary
conditions, thus, it is not likely that the high testing frequency
affected the conclusions.
This work shows that applied fatigue testing can be performed
on medical devices without lengthy testing. Using this
method of determining the sensitivity of a test and the
resulting fatigue characteristics of different designs and
processes, the design time can be reduced by ruling out
processes and designs that reduce fatigue life. This method of
testing can also be used as a quality tool to periodically check
the fatigue resistance of manufactured product.

Acknowledgments:
Thanks to the Spiration team including Tim Zilinsky, Seung
Yi, Lyle Jackson, Randy Walls, and Mitch Reynolds for their
support
 ASM International is the society for materials engineers and scientists,
a worldwide network dedicated to advancing industry, technology, and
applications of metals and materials.

ASM International, Materials Park, Ohio, USA

www.asminternational.org

This publication is copyright © ASM International®. All rights reserved.

Publication title Product code

Medical Device Materials II 05107G

To order products from ASM International:

Online Visit www.asminternational.org/bookstore

Telephone 1-800-336-5152 (US) or 1-440-338-5151 (Outside US)

Fax 1-440-338-4634

Customer Service, ASM International

Mail
9639 Kinsman Rd, Materials Park, Ohio 44073, USA

Email CustomerService@asminternational.org

American Technical Publishers Ltd.

27-29 Knowl Piece, Wilbury Way, Hitchin Hertfordshire SG4 0SX, United
In Europe Kingdom
Telephone: 01462 437933 (account holders), 01462 431525 (credit card)
www.ameritech.co.uk
Neutrino Inc.
In Japan Takahashi Bldg., 44-3 Fuda 1-chome, Chofu-Shi, Tokyo 182 Japan
Telephone: 81 (0) 424 84 5550

Terms of Use. This publication is being made available in PDF format as a benefit to members and customers of ASM
International. You may download and print a copy of this publication for your personal use only. Other use and distribution is
prohibited without the express written permission of ASM International.

No warranties, express or implied, including, without limitation, warranties of merchantability or fitness for a particular purpose,
are given in connection with this publication. Although this information is believed to be accurate by ASM, ASM cannot
guarantee that favorable results will be obtained from the use of this publication alone. This publication is intended for use by
persons having technical skill, at their sole discretion and risk. Since the conditions of product or material use are outside of
ASM's control, ASM assumes no liability or obligation in connection with any use of this information. As with any material,
evaluation of the material under end-use conditions prior to specification is essential. Therefore, specific testing under actual
conditions is recommended.

Nothing contained in this publication shall be construed as a grant of any right of manufacture, sale, use, or reproduction, in
connection with any method, process, apparatus, product, composition, or system, whether or not covered by letters patent,
copyright, or trademark, and nothing contained in this publication shall be construed as a defense against any alleged
infringement of letters patent, copyright, or trademark, or as a defense against liability for such infringement.