0803134592

STP 1438
Stainless Steels for Medical and

Surgical Applications
Gary L. Winters and Michael J. Nutt, editors
ASTM Stock Number: STP1438
INTERNATIONAL
ASTM International
100 Barr Harbor Drive
PO Box C700
West Conshohocken, PA 19428-2959
Printed in the U.S.A.
Library of Congress Cataloging-in-Publication Data

ISBN:0-8031-3459-2
Symposium on Stainless Steels for Medical and Surgical Applications (2002 : Pittsburgh, Pa.)
Stainless steels for medical and surgical applications / Gary L. Winters and Michael J.
Nutt [editors].
p. cm.--(STP; 1438)
"ASTM Stock Number: STP1438."
Includes bibliographical references and indexes.
Contains papers presented at the Symposium on Stainless Steels for Medical and Surgical
Applications held in Pittsburgh, Pa., on May 6 and 7, 2002.
ISBN 0-8031-3459-2
1. Medical instruments and apparatus---Congresses. 2. Medical instruments and
apparatus--Design and construction--Handbooks, manuals, etc. 3. Medical instruments and
apparatus industry--Technological innovations--Congresses. 4. Steel,
Stainless--Congresses. 5. Corrosion resistant alloys--Congresses. I. Winters, Gary L. I1.
Nutt, Michael J. II1. Title. IV. ASTM special technical publication ; 1438.
R856.A2S96 2002
681'.761---dc21
2003040381
Copyright 9 2003 ASTM International, West Conshohocken, PA. All rights reserved. This material
may not be reproduced or copied, in whole or in part, in any printed, mechanical, electronic, film, or
other distribution and storage media, without the written consent of the publisher.
Photocopy Rights
Authorization to photocopy items for internal, personal, or educational classroom use,
or the internal, personal, or educational classroom use of specific clients, is granted by ASTM
International (ASTM) provided that the appropriate fee is paid to the Copyright Clearance
Center, 222 Rosewood Drive, Danvers, MA 01923; Tel: 978-750-8400; online:
http://www.copyright.com/.
Peer Review Policy

Each paper published in this volume was evaluated by two peer reviewers and at least one editor.
The authors addressed all of the reviewers' comments to the satisfaction of both the technical
editor(s) and the ASTM Intemational Committee on Publications.
To make technical information available as quickly as possible, the peer-reviewed papers in this
publication were prepared "camera-ready" as submitted by the authors.
The quality of the papers in this publication reflects not only the obvious efforts of the authors and
the technical editor(s), but also the work of the peer reviewers. In keeping with long-standing
publication practices, ASTM International maintains the anonymity of the peer reviewers. The ASTM
International Committee on Publications acknowledges with appreciation their dedication and
contribution of time and effort on behalf of ASTM International.
Printed in (to come)

(to come)
Foreword
This publication, Stainless Steels for Medical and Surgical Applications, contains papers presented
at the symposium of the same name, held in Pittsburgh, Pennsylvania, on May 6 and 7, 2002. The
symposium was sponsored by ASTM Committee F04 on Medical and Surgical Materials and
Devices. Gary L. Winters of Cedar Creek, Texas, and Michael J. Nutt of Spinal Innovations, in
Bartlett, Tennessee, presided as symposium chairmen and are editors of the resulting publication.
The editors would like to express their appreciation for the help provided by Markus Windier from
Sulzer Orthopaedics, Winterthur, Switzerland, in promoting initially the need for such a symposium
and in encouraging participation by several European research groups.
We, of course, want to thank all the authors who contributed to the research presented at the symposium and, in particular, we would like to thank the presenters, some of whom made the effort and
sacrifice to travel to the United States from Brazil, France, Germany, Sweden, Switzerland, and the
United Kingdom in order to attend the symposium.
We would also like to express our thanks to the ASTM staff that helped make the symposium and
publication possible: most notably Dorothy Fitzpatrick for her help with the symposium planning and
Maria Langiewicz for the handling of manuscript submission and review. We are also indebted to all
the reviewers who volunteered their time and expertise for their careful consideration and critique of
the manuscripts.
Gary L. Winters
Cedar Creek, Texas, USA
Michael J. Nutt
Spinal Innovations,
Bartlett, Tennessee,USA
Contents
ix
OVERVIEW
ALLOY DESIGN, MECHANICALPROPERTIES
Characterization of Sandvik Bioline High-N--A Comparison of Standard Grades

F1314 and F1586----CHRISTINA HARALDSSONAND STEPHEN COWEN
X 15 T.N: T M A New Martensitic Stainless Steel for Surgical Instruments-NiCOLASPEROT, JEAN-YVES MORAUX, JEAN-PAUL DICHTEL, AND BRUNO BOUCHER
13
Development of a Platinum-Enhanced Radiopaque Stainless Steel (PERSS| -CHARLESH. CRAIG, HERBERT R. RAD1SCH, JR., THOMAS A. TROZERA, PAUL C. TURNER,
R. DALE GOVIER, EDWARD J. VESELY, JR., NEV A. GOKCEN, CLIFFORD M. FRIEND,
AND MICHAELR. EDWARDS
28
Mechanical and Corrosion Properties of Forged Hip Stems Made of High-Nitrogen

S t a i n l e s s Steel--MARKUS WINDLERAND RAINER STEGER
Metallurgical and Mechanical Evaluation of 316L Stainless Steel Orthopaedic Cable-JOHNA. DISEG! AND LYLE D. ZARDIACKAS
39
50
PROCESSING, QUALITY ASSURANCE CONCERNS, AND M R I TESTING
Processing Platinum Enhanced Radiopaque Stainless Steel (PERSS | for Use as

Balloon-Expandable Coronary Stents---J. ZACHARYDENNIS, CHARLES H. CRAIG,
HERBERT R. RADISCH, JR., EDWARDJ. PANNEK, JR., PAUL C. TURNER, ALBERTG. HICKS,
MATI'HEW JENUSAITIS, NEV A. GOKCEN, CLIFFORD M. FRIEND, AND MICHAELR. EDWARDS
61
Quality Aspects of High-Nitrogen Stainless Steel for Surgical Implants-MARKUS WINDLER, RAINER STEGER, AND GARY L. WINTERS
MRI Safety and Compatibility of Implants and Medical Devices---TERRYo. WOODS
72
82
CORROSION
Fatigue of Small Bone Fragment Fixation Plates Made from Low-Nickel Steel-LUKAS ESCHBACH, GIANNI BIGOLIN, WERNER HIRSIGER, AND BEAT GASSER
93
Comparison of Anodic Polarization and Galvanic Corrosion of a Low-Nickel Stainless

Steel to 316LS and 22Cr-13Ni-5Mn Stainless Steels---LYLED. ZARDtACKAS,
SCOTT WILLIAMSON, MICHAEL ROACH, AND JAY-ANTHONY BOGAN
107
Fatigue Behavior and In-Vitro Biocompatibility of the Ni-Free Austenitie

High-Nitrogen Steel X13CrMnMoN18-14-3--ILIA TIK~qOVSr&HOLGERBRAUER,
MARTINA MOLDERS, MARTIN WIEMANN, DIETER BINGMANN, AND ALFONS FISCHER
119
Influence of Maerophage Cells on 316L Stainless Steel Corrosion--SUZANNErl. PARKER,

HSIN-YI LIN, LYLE D. ZARDIACKAS, AND JOEL D. BUMGARDNER
137
Comparison of Notch Sensitivity and Stress Corrosion Cracking of a Low-Nickel

Stainless Steel to 316LS and 22Cr-13Ni-5Mn Stainless Steels--LYLE D. ZARDIACKAS, MICHAEL ROACH, SCOTT WILLIAMSON, AND JAY-ANTHONY BOGAN
154
Comparative Electrochemical Studies of F 1586-95 and F 138-92 Stainless Steels in

Sodium Chloride, pH = 4.0 Medium--RLrrH F. V. VmLAMm,
ARNALDO H. P. DE ANDRADE, CELSO A. BARBOSA, ALEXANDRE SOKOLOWSKI,
AND SILVIA M. L. AGOSTINHO
168
Corrosion Behavior of Platinum-Enhanced Radiopaque Stainless Steel (PERSS| for

Dilation-Balloon Expandable Coronary Stents---BERNARDS. COVINO,JR.,
CHARLES H. CRAIG, STEPHEN D. CRAMER, SOPHIE J. BULLARD,
MARGARET ZIOMEK-MOROZ, PAUL D. JABLONSKI, PAUL C. TURNER,
HERBERT R. RADISCH, JR., NEV A. GOKCEN, CLIFFORD M. FRIEND,
AND MICHAEL R. EDWARDS
176
Comparison of Corrosion Fatigue of BioDur| 108 to 316L S.S.

and 22Cr-13Ni-5Mn S.S.--LVLE D. ZARDIACr.AS,MICHAELROACH,
SCOTT WILLIAMSON, AND JAY-ANTHONY BOGAN
194
WEAR AND CORROSION-RELATED WEAR
Investigation into Wear-Induced Corrosion of Orthopaedic Implant Materials-MARKUS WINDLER, JULIE E. MACDOUGALL, AND ROLF SCHENK
211
Results of In-Vitro Studies about the Mechanism of Wear in the Stem-Cement

I n t e r f a c e o f T H R - - D m T E R WIRZ, BRUNO ZURFLUH, BEAT GOPFERT, FENG LI,
WILLI FRICK, AND ERWIN W. MORSCHER
Fretting and Anodic Current of the Taper Interface of Stainless Steel Hip
Stems/Cobalt-Chromium Femoral Heads---MENGKEZHUANDMARKUSWINDLER
222
235
CLINICAL EXPERIENCE
The Cemented MS-30 Stem in Total Hip Replacement, Matte versus Polished Surface:
Minimum of Five Years of Clinical and Radiographic Results of a Prospective
Study--BERNHARD BERLI, REINHARDELI(E, ERWIN W. MORSCHER
249
Corrosion Products Generated from Mechanically Assisted Crevice Corrosion of

Stainless Steel Orthopaedic Implants---ROBERTM. URBAN,JOSHUAJ. JACOBS,
JEREMY L. GILBERT, ANASTASIAK. SKIPOR, NADIM J. HALLAB, KATALIN MIKECZ,
TIBOR T. GLANT, J. LAWRENCEMARSH, AND JORGE O. GALANTE
262
Author Index
273
Subject Index
275
Overview
The Symposium on Stainless Steels for Medical and Surgical Applications undertook to bring together, at one time, current research and development, recent clinical experience, and the substantial
history of stainless steels in medical applications. This was done with the expressed purpose of bringing the attendees and participants of the symposium up to date on the current status of the latest technologies and to provide the most current information. Ultimately, this new information may stimulate further research and technical developments in the application of stainless steels for implantable
medical devices and surgical instruments.
This Special Technical Publication (STP) furthers the aims of the symposium by documenting the
entirety of the body of research that was presented in 20-minute synopses during the symposium. This
information, which has not been previously published, is to be made available not only to the attendees and participants of the symposium, but also to the general public. In this manner, the information in the symposium can be distributed more widely and be more helpful in furthering research and
technical developments in the application of stainless steels for implantable medical devices and surgical instruments.
This publication will be helpful to individuals engaged in the development of stainless steels for
implant applications, or instrument applications in particular. It will also be of interest to other individuals, who are developing, making, or using implantable medical devices and/or surgical instruments in general. This publication may make both of these groups aware of alloys, testing, or applications that they had not previously considered.
This new information can stimulate new research or technical innovations. This information can
also be used by individuals to contact researchers or producers with similar interests. Finally, this information may become part of the justifications for device approval by regulatory bodies, and for creation of material specifications by standards writing bodies, which can benefit producers, users, and
regulatory bodies.
This publication addresses some of the issues in the medical uses of stainless steels, such as corrosion, wear, biological response, radiopacity, and the high cost of medical products. New alloys are
discussed as solutions to some of these issues by offering more biocompatible, higher quality, radiopaque, or low cost alternatives for orthopaedic implants and stents. Several corrosion papers address this key concern for all medical devices. Some of these papers offer creative options for this
field. Finally, there are other papers that deal with such timely issues as injuries due to the interaction
of implants and MRI examinations.
This symposium covered a wide range of topics on stainless steels, with most of the presentations
dealing with narrow segments of a specific topic. Therefore, a single theme of the presentations may
be that work on stainless steels for medical uses continues and that stainless steels may be part of the
answers for some of the issues facing the surgical community today, such as biological response, corrosion resistance, mechanical performance, quality, and cost.
Certainly, iron-base alloy systems offer excellent opportunities as implant and instrumentation alloys in the medical field to satisfy special needs such as radiopacity, wear resistance, superior me-
ix
STAINLESS STEELS FOR MEDICAL AND SURGICAL APPLICATIONS
chanical properties, or freedom from allergic responses. As such, work should continue on increasng the corrosion resistance and mechanical properties of stainless steels, as well as the product forms
and applications, while decreasing cost and any negative biological responses.
Gary L. Winters
Cedar Creek, Texas
Michael J. Nutt
Spinal Innovations
Bartlett, Tennessee
Alloy Design, Mechanical Properties
Christina Haraldsson 1 and Stephen Cowen 2
Characterization of Sandvik Bioline High-N - A Comparison of Standard Grades
F1314 and F1586
Reference: Haraldsson, C. and Cowen, S. "Characterization of Sandvik Bioline HighN - A Comparison of Standard Grades F1314 and F1586," Stainless Steelsfor
Medical and SurgicalApplications, ASTMSTP 1438, G. L. Winters and M. J. Nutt, Eds.,

ASTM International, West Conshohocken, PA, 2003.
Abstract: ASTM standard grades for surgical implants F1314 and F1586 are both N-
strengthened austenitic stainless steels with high resistance to pitting corrosion. The latter
is generally used in Europe while the former is preferred in the United States. In the
present study, Sandvik grade Bioline High-N, conforming to ASTM F 1586, has been
characterized with respect to microstructure, mechanical properties, corrosion resistance
and inclusion content and compared with ASTM F 1314-material.
In the present comparison it has been shown that Bioline High-N 3 compared to ASTM
F1314 material has similar or slightly superior mechanical properties, pitting corrosion
resistance and nonmetallic inclusion content. All these properties are of critical
importance for medical implant material.
Keywords: Stainless steel, implant materials, nonmetallic inclusions, pitting corrosion

resistance, mechanical properties, microstructure
1 Research Engineer, Product Development, Metallurgical Research, AB Sandvik Steel,

SE-811 81 Sandviken, Sweden.
2 General Manager, Sandvik Metinox, Long Acre Way, Holbrook Sheffield $20 3FS,
U.K.
3 Bioline High-N is a registered trademark ofAB Sandvik Steel, Sandviken, Sweden.
Copyright*2003 by ASTM International
www.astm.org
STAINLESSSTEELS FOR MEDICAL AND SURGICAL APPLICATIONS
Introduction
Stainless steel is a widely used material for orthopedic implants due to its competitive
price level, wide availability of product forms and relative ease of manufacture. The most
common grade of stainless steel used for implants conforms to ASTM specification for
Wrought 18 Chromium-14 Nickel-2.5 Molybdenum Stainless Steel Bar and Wire for
Surgical Implants (UNS $31673), (F 138).
Whilst ASTM F138 has excellent benefits in a wide range of medical implant
applications, it does have certain limitations with regard to ultimate tensile strength and
corrosion resistance. In specific permanent load bearing orthopedic devices, such as hips,
knees and shoulders, and in products for trauma fixation such as intermedullary nails and
screws, it is often the case that additional strength, enhanced fatigue properties and better
corrosion resistance are required for optimum prosthetic performance and patient benefit.
Materials conforming to both ASTM Specification for Wrought Nitrogen
Strengthened-22 Chromium-12.5 Nickel-5 Manganese-2.5 Molybdenum Stainless Steel
Bar and Wire for Surgical Implants (F1314) and ASTM Specification for Wrought
Nitrogen Strengthened-21 Chromium-10 Nickel-3 Manganese-2.5 Molybdenum Stainless
Steel Bar for Surgical Implants (F1586) have additional benefits that make their usage in
certain applications more favorable than material conforming to ASTM F 138. The base
for Bioline High-N material (conforming to ASTM F1586) is a fully austenitic 316
material. Alloying with higher levels of Nitrogen (0.25 - 0.50%) in combination with
increased levels of Chromium (19.5 - 22.0%) and small additions of Niobium (0.25 0.80%) facilitates solid solution strengthening within the material which in turn delivers
higher mechanical strength and enhanced resistance to pitting corrosion. This is achieved
without the need for additional Nickel that would increase the cost of the material [1].
In the present study Sandvik Bioline High-N, conforming to ASTM F1586, has been
characterized with respect to microstructure, mechanical properties, corrosion resistance
and nonmetallic inclusion content and compared with ASTM F1314-material. In cases
where experimental results are missing for the ASTM F 1314-material the standard
requirements have been substituted.
Materials
Nitrogen-strengthened austenitic stainless steels conforming to ASTM F1314 and

F1586 are very similar in composition (Table 1). Historically the latter has generally been
used in Europe while the former has been more predominant in the United States.
Regarding chemical composition the most significant differences between F1314 and
F 1586 is the Vanadium (V) addition in the F1314 material, its lower levels of Niobium
(Nb) and also the higher Nickel (Ni) level (Table 1), F1586 allows up to 0.08% carbon,
which is well above the maximum level for ASTM F1314 (Table 1). But, to avoid
carbide precipitation the carbon content in F1586 is generally kept well below the
maximum level that reduces the difference in carbon content between the two materials.
Vanadium and Niobium both have a stronger tendency to form carbides compared to
HARALDSSON AND COWEN ON SANDVIK BIOLINE HIGH-N
Chromium (Cr). This prevents the precipitation o f the detrimental chromium carbides
within the material. The affinity to carbon, nitrogen, oxygen and sulfur for different
elements are schematically shown (Figure 1)[2]. This must be seen as a general ranking
since neither the temperature nor the elements' differing solubility in different phases are
considered.
Table 1 - Chemical compositions (weight-%)

Element
A S T M F1586
Biolin e High-N 1 A S T M F1314
Carbon
0.08 max
0.06 max
0.030 max
Manganese
2.00 - 4.25
4.0
4.00 - 6.00
Phosphorus
0.025 max
0.025 max
0.025 max
Sulfur
0.01 max
0.003 max
0.010 max
Silicon
0.75 max
0.60 max
0.75 max
Chromium
19.5 -22.0
20.5
20.50 - 23.50
Nickel
9.0 - 11.0
9.5
11.50 - 13.50
Molybdenum
2.0 - 3.0
2.4
2.00 - 3.00
Nitrogen
0.25 - 0.5
0.4
0.20 - 0.40
Niobium
0.25 - 0.8
0.3
0.10 - 0.30
Vanadium
0.10 - 0.30
Copper
0.25 max
0.50 max
Iron
Balance
balance
Nominal composition
C:
N:
O:
S:
Fe Mn Cr W M o V
TaNbTiZr
Fe MnWMoCrNb SiVCaCeTa A1
Ti
Zr
Fe
CrNb MnVC Si Ta Ti AIZrMgCeCa
Fe
Cr
Nb Ta Mn Ti AI
Zr Mg
Ca Ce
Ib
Increasing affinity
Figure 1 - Ranking o f the alloying elements affinity to (7, N, O a n d S
V is a rather strong ferrite former. Patti and Schiller [3] have suggested a factor o f 5 in
the Cr-equivalent that should be compared with their suggestion o f a factor 1.75 for Nb,
formula (2). The reason for the somewhat higher Ni level in the F 1314 material compared
to the F1586 might hence be to keep the material fully austenitic, formula (1)
Nieq, = Ni + Co + 0 . 5 M n + 0 . 3 C u + 30(7 + 2 5 N
Crequ = Cr + 2 S i + 1 . 5 M o + 5 V + 5.5"A1 + 1.75Nb + 1.5"Ti + O. 75"W
(1)
(2)
STAINLESSSTEELS FOR MEDICALAND SURGICAL APPLICATIONS
Experimental
Microstructure
Microstructural investigations were performed in a Leica DMR light optical

microscope (LOM) using specimens electrolytically etched in 40% NaOH at 3V for
approximately 10s.
Nonmetallic Inclusions
The inclusion content has been determined on longitudinal samples. Both the ASTM
Test Methods for Determining the Inclusion Content of Steel (E45) Method A and the
Swedish Standard Steel-Method for assessment of the content of nonmetallic inclusions
(SS 11 11 16) have been used.
The inclusion assessment using method ASTM E45-A is described as the rating
number of the largest detected inclusion according to a comparative chart. Several
inclusions of the same largest size, or just one single inclusion, generate the same results
i.e. the inclusion frequency is not considered here. The standard specification
requirements for ASTM F 1314 and F 1586 respectively are given in (Table 2), It can be
noted that F1314 allows a somewhat higher inclusion rating for both BT and CT,
otherwise the requirements are identical.
Table 2 - Maximum inclusion ratings according to ASTM E45-A

AT
AH
BT
BH
CT
CH
DT
DH
ASTM F1586
1.5
1.5
2.0
1.5
2.0
1.5
2.5
1.5
ASTM F1314
1.5
1.5
2.5
1.5
2.5
1.5
2.5
1.5
A comparative measurement has been performed using the Swedish standard

SS 11 11 16. The difference between this and the former measurement is that the Swedish
standard takes into account every inclusion larger than a specific size and not just the
largest one. This results in an inclusion frequency expressed as the number of
inclusions/cm 2 in different size classes. The minimum inclusion size included is 3 ~-n
and the following terminology is used.
Inclusion type
B=
C=
D=
Brittle inclusions
Brittle-ductile inclusions
Undeformed inclusions
Inclusion width
T = thin
3-6 microns
M = medium 6-11 microns
II = heavy
11-22 microns
Corrosion Resistance
The pitting corrosion resistance has b ~ tested in accordance with Method A in

ASTM Test Methods for Pitting and Crevice Corrosion Resistance of Stainless Steels and
Related Alloys by Use of Ferric Chloride Solution (G48).
A comparison of the pitting resistance equivalent, PRE, has been completed using
formula (3)[4]
PRE = Cr + 3 . 3 M o +16" N
(3)
At the Royal Institute of Technology an electrochemical study of the resistance to

localized corrosion has been performed on ASTM F138 and F1586 (Bioline High-N) [5].
The investigations were performed on annealed material at 25~ with a scanning rate
10mV/min in a phosphate buffered saline (PBS) solution and in a very aggressive
simulated crevice solution designed to cause crevice corrosion by a high concentration of
chlorides and a low pH value. The composition of the two solutions is given below
(Table 3).
Table 3 - Comp,.xs'itions o f the PBS and simulated crevice solutions
NaC1
Na2I-[PO4
KH2PO4
PH
PBS solution
8.77 g/L
1.42 g/L
2.72 g/L
7
Simulated crevice solution

146.25 g/L
1.42 g/L
2.72 g/L
0.85
Results
Microstructure
The microstructure in both F1314 and F1586 material is austenitic without any traces
of ferrite when examined in x100 magnification. Oblong primary precipitates rich in
niobium and nitrogen can be seen rather frequently in both materials. In the F1314
material the precipitates generally seem to have sharper comers (Figures 2 and 3). In
material of type F1586 the precipitate has been identified as Z-phase [6, 7] with the
nominal chemical formula CrzNb2N2 [8]. Trials to dissolve the Z-phase in Bioline HighN by heat treatment have been performed in the range 900-1200~ without success.
Figure 2
High-N
Microstructure of Bioline
Figure 3 - Microstructure of F1314material
Nonmetallw Inclusions
The ASTM E45 results (Table 4) show that both materials fully conform to the
standard specifications. From the results it can be seen that neither of the materials has
any A-type inclusions due to very low S-content. In the investigated F1314 material, only
B-type inclusions were found, whereas Bioline High-N has inclusions of both type B and
C, although the C-type dominates. This is due to differences in the method of
deoxidization used. Both materials have inclusions of type D - the investigated F1314
material with somewhat higher ratings compared to Bioline High-N. For F 1586 the
ASTM ratings presented are average values for 17 heats, approximate heat weight 60
tonnes, investigated as billets of dimension 120xl20mm square. The F1314 material was
investigated as bar samples of diameters 8 and 11 mm. It should be noted that the heavier
deformation of the F1314 material might have caused the D-type inclusions to crush and
form inclusions of type B.
Table 4 - Inclusion rating according to ASTM E45 method A, plate IlL

AT
AH
BT
BH
CT
CH
DT
Bioline High-N ~ 0,0
0,0
0,1
0,0
1,0
0,9
0,9
F 1314-material2 0,0
0,0
0,9
0,0
0,0
0,0
1,3
1Average of 17 heats
2 Average of 2 heats
DH
0,1
0, !
The inclusion frequency assessment (Table 5) clearly shows that Bioline High-N has a
much lower total number of inclusions in the investigated size range, 3-22 ~tm, compared
H A R A L D S S O N A N D C O W E N ON S A N D V I K BIOLINE HIGH-N
to the investigated F1314-material. The calculated area-% of inclusions is an average of

0.0080% for Bioline High-N, compared to 0.0140% for the investigated F1314 material.
Table 5 - Inclusion content according to SS 11 11 16 as No. of inclusions.lOree
BT
BM
BH
CT
CM CH
DT
DM DH
Bioline High-N t
0
0
0
6
4
2
129
10
0
F 1314-material 2
95
0
0
0
0
0
575
90
7
-7 Average of 9 heats
2 Average of 2 heats
Corrosion
The ASTM G48 test was performed on double samples of both the F1586 and F1314
material. The samples were prepared in the same way and tested at the same time under
similar conditions. The results (Table 6) show that the materials have about the same
resistance to localized corrosion, F 1586 being slightly superior.
Table 6 - Calculated PRE-values and CPT-values from ASTM G48 test
PRE
Specimen I Specimen II Average
Bioline High-N 34,8
40 ~
50~
45~
ASTM F1314
33,0
40 ~
40~
40~
Polarization curves obtained from the four cyclic potentiodynamic measurement runs
of Bioline High-N in PBS solution (Figure 4) show rather small deviations between the
separate runs [9].
1.5
1.0
O
~,
o.~
B.I
L)
I11
-0.5 ~
-1.0
104
10"7
10~s
10-s
104
10-3
I (Amps/cm 2)
Figure 4 - Cyclic polarization curves for Bioline High-N obtained in PBS solution, four
separate runs.. Scan rate l OmV/mm
10
The pitting potential and the protection potential in PBS solution were determined to
1.10 V/SCE and 1.04 WSCE respectively to be compared with 0.36 V/SCE respectively 0.16 V/SCE for F138 material (Table 7) [10]. The results suggest that Bioline High-N has
a very low tendency to pitting initiation due to a high pitting potential as well as a strong
ability to re-passivate if the passive film is damaged due to its high protection potential.
Table 7 - Results from cyclic potemiodynamic polarization measurements with regard to
Bioline High-N
Pitting potential
Protection potential
PBS-solution
1.10 V/SCE
1.04 WSCE
Simulated crevice solution

0.9 WSCE
0.8 WSCE
The results from the cyclic potentiodynamic polarization in the simulated crevice
solution resulted, as expected, is somewhat lower values for both the pitting and the
protection potential (Table 7) [10]. Despite the much higher aggressiveness of the
simulated crevice solution only a slight decrease in both the pitting and protection
potential was achieved, indicating excellent resistance to localized corrosion.
Mechanical Properties
For the F1314 material standard data has been used due to no actual test result being
available.
Table 8 - Mechanical requirements according to material ,specifications

ASTM F1314
ASTM F1586
Bioline
High-N z
Annealed
Yield strength Rp 0,2/
380
430
525
MPa
UTS Rm / MPa
690
740
862
Elongation A / %
35
35
39
Cold-worked
Medium hard Hard
Yield strength Rp 0,2/
862
700
1000
MPa
1100
2
UTS Rm / MPa
1035
1000
10
2
Elongation A / %
12
20
Typical values
2 See Figure 5
11
When comparing the minimum material requirements ASTM F 1586 have higher
overall mechanical properties in the fully annealed condition when compared to the
F 1314 material (Table 8). When the two materials' minimum requirements are compared
in the cold worked or hardened condition both the UTS and 0.2% proof stress figures for
the F1586 material exhibit higher properties combined with higher values of elongation
(for the same UTS level) than for the F1314 material (Table 8).
Bioline High-N therefore, exhibits high strength with good ductility, making it an
obvious choice for medical implant devices. Further study has also revealed that High-N
material performs in a very linear manner under cold working and a graph of tensile
strength vs. reduction of area during cold work can be seen in Figure 5.
17oo 1-. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . T 3s
16oo
14oo ~
1200
- - - - -
~..--~:~~"ii
_ ..=
9
,
-=r,-
~ " i ' ' m
...............+i'-'+~
~ " ~
'
+ - - -
+,,oo+ .-::-
-,oo+
600
500 ~400 |
300
-_---
/
loo !
5,00
. . . . . . .
~-
~,
.
.
.
.
10,00
.
.
.
.
15,00
20,00
.=',.,.- 9
9 ~,,
,+'++'
25,00
9
9
. . . . .
*~
__
.
.
+,
200
--t
.
30,00
1,00%
0,20%
El% (SO)
- 10
--
5
'"
.
35,00
UTS
- ~
40,00
o
45,00
"
-I.htear
io,2O%1
RofA%
Figure 5 - Tensile strength vs. reduction of area during cold worlang of Bioline High-N.
Conclusion
Bioline HighoN has been characterized and compared to both ASTM FI586 and
F1314 standards. Chemical composition, microstructure, nonmetallic inclusion,
corrosion resistance and mechanical properties have been compared.
The microstructure in both materials is austenitic with precipitates of a niobium rich
nitride. According to minimum requirements for mechanical strength in ASTM
specifications F1314 and F1586, the latter is slightly superior.
Nonmetallic inclusion evaluation has been performed both according to ASTM E45
method A and as a frequency analysis on both material grades. Bioline High-N has
consistently been shown to be well within the demanding requirements ofF1586,
demands which are somewhat harder to fulfil when compared to F1314.
The pitting corrosion resistance has been found to be very good due to high Cr, Mo
and N-levels. The critical pitting potential (CPT), determined by ASTM G48-test has
12
shown a somewhat better result for Bioline High-N compared to the investigated F1314
material.
It has been found that Bioline High-N conforming to ASTM F1586 exhibits lower
nonmetallic inclusion levels, somewhat superior pitting corrosion resistance with higher
levels of mechanical strength when compared to material meeting the requirements of the
F1314 specification.
References
[1] Leclerc, M.F., "Corrosion; 3~dedition"; Eds Shrier,L.L; Jarman, R.A. and Burnstein,
G.T.; pp. 164-180; Butterworth-Heinemann Ltd, Oxford, England, (1994).
[2] Hillert, M., Compendium in "Phase Transformation, basic course part lI"page Fe20
("Fasomvandlingar, grundkurs del H") Stockholm 1986, Royal Institute of
Materials Science and Technology.
[3] Patti,G. and Schiller,P., Journal of Nuclear Material, Vol. 141/ 143, 1986, pp.417426.
[4] Lorentz, K. and Medawar, G., "Uber das korrosionsverhalten austenitisscher ChromNickel-molybdan-Stahle mit und ohne Stickstoffzusatz unter besonderer
BerOcksichtigtmg ihrer Beanspruchbarkeit in Chloridhaltigen LOsungen." Thyssen
Forschung, 1, 1969, pp.97-108.
[5] Pan, J., Karlrn, C., and Ulfvin, C., "Electrochemical Study of Resistance to
Localized Corrosion of Stainless Steels for Biomaterial Applications," Journal of
The Electrochemical Society, Vol. 147 (3), 2000, pp. 1021-1025.
[6] Robinson, P.W., and Jack, D.I-L, "Precipitation of Z-phase in a high-nitrogen stainless
steel," New Developments in Stainless Steel Technology, ASM, Metals Park OH
1985, pp. 71-76.
[7] Ornhagen, C., Nilsson, J.-O., and Vannevik, H., "Characterization of a nitrogen-rich
austenitic stainless steel used for osteosynthesis devises," Journal of Biomedical
Materials Research, Vol.31,1996, pp. 97-103.
[8] Andrews, K.W., Dyson, D.J., and Keown, S.R., "The interpretation of Electron
Diffraction patterm'," Adan Hilger, London, England, 1971.
[9 ] Hjalmarsson, C., "Comparative study of local corrosion resistance for implant grade
stainless steels using electrochemical methods," MSc Thesis at the Royal Institute
of Technology, Division of Corrosion Science, 1997.
Nicolas Perot, t Jean-Yves Moraux, 1Jean-Paul Dichtel, 2 and Bruno Boucher 3

X 15 T . N TM : A new Martensitic Stainless Steel f o r Surgical Instruments
Reference: Perot, N., Moraux, J.Y., Dichtel, J.P., and Boucher, B., " X 15 T.NTM: A new
Martensitic Stainless Steel for Surgical Instruments," Stainless Steels for Medical and
SurgicalApplications, ASTMSTP 1438, G. L. Winters, and M. J. Nutt, Eds., ASTM
International, West Conshohocken, PA, 2003.
Abstract: The edge retention and wear resistance of steels are dependant upon hardness
and structure. Until now, two main principles have been adopted to increase the hardness
ofmartensitic stainless steels:
1.Either through the introduction of a structural hardening mechanism into low carbon
martensitic grades such as, AISI 630 and XM 16,
2.Or by increasing the carbon content of the martensite and subsequently strengthening it
through the presence of primary carbides.
The first option has not allowed to date, the mass production of steels with hardness
levels greater than 52/54 HRC. Furthermore, the absence of carbides deleteriously affects
cutting quality. The second option has two distinct disadvantages directly linked to the
presence of large carbides, which render the steel brittle and reduce corrosion resistance,
through chromium depletion of the matrix in the immediate vicinity of the carbides.
In X 15 T.N TM,(UNS S 42025) nitrogen is used as a partial substitute for carbon, causing
hardening by interstitial insertion as well as by the formation of fine nitrides and
carbonitrides. In this way, a hardness level (58/60 HRC) is obtained equivalent to a high
carbon martensitic steel without chromium depletion, thereby giving a substantial
improvement in corrosion resistance, i.e. equivalent or superior to AISI 630 and XM 16.
Keywords: medical instruments, hardness, corrosion resistance, cutting, wear resistance,

nitrogen alloyed steel, martensitic steel
1Market Manager and Chief Metallurgist, respectively, Aubert & Duval Holding,
33 av du Maine, 75755 Paris cedex 15 (France).
2Engineer, Aubert & Duval Alliages BP63, 92233 Gennevilliers eedex (France).
3President, Forecreu America, 37-35 W. Belmont Av, Chicago, IL, 60618 (USA).
13
www.astm.org
14
Foreword
Medical instruments are used in a vast field of applications, and properties required
from stainless steel grades are widely diversified. For many applications, the selected
grade has to simultaneously satisfy the following basic characteristics
1.High hardness for wear resistance.
2.Good shape keeping edge retention for cutting tools, profile keeping for screw drivers.
3.Satisfactory corrosion resistance against cleaning and disinfecting products.
In addition, grades have to posses good fatigue and tenacity properties to avoid premature
rupture resulting from high usage. Included in this category are cutting tools in general
and ancillary instruments, such as drills, rasps, screwdrivers
Some Examples of Existing Grades (Table 1)
Steel Type AIS1420B: 13% chromium and O.3% carbon
A 52 HRc hardness combined with chromium carbide in the structure gives this grade
very good wear resistance as well as good edge retention. The main disadvantage is poor
corrosion resistance. The nominal 13% chromium content is close to the minimum
content to insure passive film formation. The matrix chromium content is lowered by the
precipitation offr23C6 type carbides in the hardened and tempered condition.
Steel Type AIS1431:17% chromium, 2% nickel and O.16% carbon
A higher chromium content ensures a better corrosion resistance, but a lower hardness
of 43 HRc (47HRc can be reached on small parts).
Steel Type AIS1440C: 17% chromium, 1% carbon and addition of molybdenum
Due to its high carbon content, 60 HRC hardness can be reached, its structure being
made mostly of Cr23C6 type carbides. These characteristics allow extremely good edge
retention, but, unfortunately, these carbides weaken the Chromium matrix and give this
grade very poor corrosion resistance and fracture toughness.
PEROT ET AL. ON X 15 T.N
15
Precipitation Hardening Martensitic Steels such as AIS1630, XM 16
These grades offer better corrosion resistance because nickel is added to the base
composition which contains high chromium and low carbon content. Hardening is
obtained through fine precipitation of intermetallic compounds : Precipitation Hardening
(PH steels). The main advantage is that the 17% of chromium is solely dedicated to
corrosion resistance, optimizing it. However, hardness is limited to 47HRC, and the
absence of carbides negatively affects the tools' edge retention.
Table 1- Typical Composition of some Martensitic Stainless steels (%)

C
Cr
Mn
Si
Ni
AISI XM 16
<
0.03
12
<
0.5
<
0.5
<
0.015
<
0.015
8.5
AIS1420 B
0.3
13
"42010" (T.R)
0.2
14.5
AIS1431
0.16
16.5
<
AIS1630
AIS1440 C
X 15 T.N
0.07
17
1.0
17
0.40
15.5
<
<
<
<
1.0
1.0
0.030
0.040
<
1.0
<
1.0
<
0,030
<
0.040
<
<
<
<:
1.0
<
1.0
<
0.030
<
0.040
<
1.0
<
.0
<
0,030
<
0.040
<:
1.0
1.0
0.030
0.040
<
<
<
<
0.6
0.6
0.005
0.020
Mo
OTHER
Hardness
HRC
Max.
+Cu+Ti+
Cb
52
52
0.7
0.8
51
43-47
+Cu+Cb
0.5
2.0
44
60
0.3
N: 0.2%
59
Existing steels available offer a dilemma, either a high carbon level offering good
mechanical properties and poor corrosion resistance, or low carbon level improving
corrosion but lowering mechanical properties. X 15 T.N was designed to address this
dilemma and offer a good compromise between hardness, corrosion resistance and
mechanical behavior.
Composition
Typical Chemical Composition is as follows: 0.40% carbon, 15.5% chromium, 2.00%

molybdenum, 0.30% vanadium, and 0.20% nitrogen.
Numerical simulation, using the most recent thermodynamic data, was used to
optimize chemical composition (Figure I). This was necessary to obtain a stabilized
martensitic composition that contained a high nitrogen content. For reasons of cost and
16
manufacturing capabilities reasons, we chose to use only conventional melting processes.

For these reasons, powder metallurgy technology or high pressure remelting such as
Pressurized Electro Slag Remelting (PESR) were not taken into consideration. The final
composition is based on the achievable properties and structure following a easy heattreatment.
~
L + ct
1400
"
26OO
a+y
VN+C
800
1400
Ct + M 23Cs + VN + Cr2N
6oo
I0~
40O
O.2
'
01,
",
01.8
t.0
% CARBONE ( weight / polfls )
Figure 1 - Phase Diagram

It is possible to summarize the optimized balance as follows:
Carbon: This level, combined with le nitrogen content, is adequate for obtaining a high
hardness without an excessive number of carbides after hardening and tempering.
Chromium: This level allows, in relation to the carbon content, the formation of a passive
layer to ensure high corrosion resistance.
Molybdenum: The addition of molybdenum provides a more stable passive layer for
improved corrosion resistance, especially in a chloride environment.
Vanadium: Added to allow the formation of vanadium nitride/carbonitride for improved
cutting edge capabilities. Moreover, it allows secondary hardening that can be used with
some special heat treatments for parts designed to receive a Physical Vapor Deposition
(PVD) type coating.
Nitrogen: Added to replace part of the carbon, to insure proper hardening response and
improved corrosion resistance.
Nickel: None added to X 15 T.N.
Micrographs of AISI 440C and X 15 T.N, shown in figure 2, clearly illustrate the
improvements brought by X 15 T.N as regards morphology, size and carbide density.
17
Figure 2 - Comparison of Structures
Melting Route
The melting and refining route (Figure 3) starts with an arc furnace primary melting
operation followed by Argon Oxygen Decarburization (AOD) refining. At that stage, the
nominal composition of the steel is established. At a later stage, an Electro Slag
Remelting (ESR) operation is performed with a slag that is formulated for this grade.
ESR lowers the sulfur content for increased corrosion resistance, uniformly distributes
the nitrogen content and provides microstructural homogeneity. Finally, the nonmetallic
inclusion content is lowered thereby increasing toughness and fatigue characteristics.
Afterwards, X 15 T.N is converted by forging, hot-rolling and sometime cold-working, to
manufacture bars, rods and sheets.
I AIR(ArcMELTING
]
Furnace)
A.O.D
(Argon Oxygene Decarbarization)
+
I (Electro
E.S.R
SlagRemelt) 1
Figure 3 - Melting Route
18
Heat T r e a t m e n t a n d R e s u l t a n t Mechanical Characteristics
Continuous Transformation Temperature (CCT) diagram (Figure 4) and tempering

curve (Figure 5), illustrate perfectly the heat treatment possibilities for the grade. For
medical instrument applications, best performances are obtained through two different
quenching and tempering treatments to which a useful annealing treatment can be added
if the instrument is made from a forging.
9c
Aust~nitisation / Austenitization : 1075~
I 1967~
- 30 mn / min
000
90O
8O0
70O
600
50O
40O
300
2OO
100
10
100
Sec.
1000
10000
1(10000
15ma/ mia
! mn/min
2 real rain
lhlhr
4blhr
2 It/hr
24 h/hr
8h/hr
Figure 4 - Continuous Transformation Temperature (CTT) diagram
"F
2100
400
600
1180
t00
200
300
400
l~
'1200
"1:
HRC
80
600
MO
"r
Figure 5 - Tempering Curve (austenization between

980~176
and 1077~176
Annealing
The annealing treatment state is given by the following :
1. Heat up to 840~176
2. Slowly cool in furnace or under ashes.
A hardness of approximately 207 HB will be obtained.
Heat Treatmentfor HRc _~58 :Dedicatedfor Maximum Hardness (Figure 6)

Typical mechanical properties:
UTS
: 336KSI / 2320MPa
Charpy V-Notch Impact Strenght: 7.5 ft.lb/10J
YSo2%
: 265KSI / 1825MPa
K I C : 12.7KSI~/in/14MPa~/m
Elongation: 4%
Fatigue limit (rotating bending) 107cycles : 135KSI / 928MPa
19
20
9~
I~
1050~C I tD O'F
25rain~in
3~mln
600"C
rich : GaS
Or ON
~ ) 2 houm r r ~ x ~ j m
-80~
ro
r
2
hr
Figure 6 -Thermal cycle for HRc _>58

Heat Treatment for HRC _>55 : Dedicated for Easy Heat Treatment (Figure 7)
Typical mechanical properties:
: 313KSI / 2160MPa
Charpy V-Notch Impact Strenght: 7.5 ft.lb/10J
YS0.2~ : 233KSI / 1610MPa
K]C : 18.2KSI,fin / 20MPax/m
Elongation : 4%
Fatigue limit (rotating bending) 107cycles : 125KSI / 865Mpa
UTS
"G / "T:
1010"G 1 1 8 S r ~
25 m/n/h~
30
m~ r n l n ' l ~
(~) indicative ~ l t ~
~) 2 ~
Time
maximum
Figure 7 - Thermal cycle for HRC_>55
21
A close look at these properties illustrates :

1. Low nonmetallic inclusion content and a fine carbide distribution provides high
resistance to crack initiation and outstanding fatigue behavior.
2. The caracteristics of nitrogenized martensitic steel is a low toughness. This is the
result of the intensity of internal stress created by the very high supersaturation of the
martensite of these grades in nitrogen and carbon.
This property, seen only in highly nitrogen enriched martensitic steel, brings to the
designer the assurance of having a durable instrument as long as he avoids local stress
concentration resulting from inadequate design (sharp angle) and/or poor surface finish
(machining grooves). A good surface f'mish improves tenacity and at the same time
increases corrosion resistance.
Corrosion Resistance
Salt Spray Test

The closest "real life" environment test to grade corrosion performance is also the
most simple: the grade either corrodes or it does not corrode after a given exposure time.
Charts of comparative tests between 440C AISI, XM 16, AISI 630 and X 15 TN are
shown in Figure 8. They illustrate the drastically superior behavior of X 15 TN.
Figure 8 - Salt Spray Test

X 15 T.N performs at least as well as precipitation hardening grades, such as AISI 630 and
XM 16, in spite of a much higher hardness, respectively 15 and 7 HRC higher.
For a similar hardness, (59/60 HRc) results on the X 15 TN are far better than AISI 440C.
22
After prolonged exposure, observed rusting o f X 15 TN is superficial and can be removed

with a light rubbing with sand paper whereas, 440C has deep corrosion pits (Figure 9)
Figure 9 - Comparison after soft polishing

These tests were performed according to NFX 41-002 standard on polished samples,
Tests conducted on rough machined surfaces show a slower rusting acceleration on AISI
630 and 440C than on X 15 TN and AISI XM 16. Tests run according to ASTM B 117
and DIN 50021 reach the same conclusion.
Potentio-Kinetic Tests
Potentiokinetic tests allow the evaluation o f passive layer behavior and the corrosion
risks in various grade-environment combinations.
CORROSK)N
CURRENT
I
t '
PAa~E ~GE
/
LL~ ~
II
SLOPEOFTHE
TRANSPASSIVE
BRANCH
,i
CORROSIONPOTENTIAL
CROSS OVER
9 I
BREAK
THROUGH
POTENTIAL
D
POTENTIAL (mV)
S/NGPOTEN/~IL
oL,c ~ p , ~ POTB~r~L
Figure 10 - Typical Potentio-Kinetic Curve
PEROT ETAL. ON X 15 T.N
23
The typical feature of potential-current diagrams, and the terms generally used, are shown
in Figure 10.
Tests in Sulfuric Acid Solution

Tests were made at a temperature o f 20~176
in a 1% sulfuric acid solution, argon
saturated, typical environment to test general corrosion. Two potential closed loops were
made (Figure 11 and Table 2). Current levels at the corrosion peak allow us to evaluate
clearly the behavior of the three grades tested.
1% H2804 De-oxygenated Solution

, r
.m
a(~*a~ t
L_i
i,i -. . . .
AIS! 440 C - 60 HRC
'
.....
X 15 T.N - 59 HRC
Figure 11 - Current Density/Potential Curves (Second Cycle)

Current levels at the corrosion peak allow us to evaluate clearly the behavior of the three
grades tested.
Table 2 - Corrosion Current at the Peak
Cycle
First
Current Density (laA/cm 2)

AISI 420B
AISI 440C
X 15 T.N
52HRC
60HRC
59HRC
Increasing potential
2400
1100
400
Decreasing potential
10
500
<6*
Increasing potential
2700
2700
<6*
Decreasing potential
20
700
Second
<6*
*limite
measurement.
Corrosion currents observed on X 15 TN are much lower by a large margin than those o f
AISI 440C. They are also better than AISI 420 that, furthermore, has a lower hardness.
24
Tests in O.9%NaCl Solution, (37~
6~ Aerated Condition)
Tests made on X 15 T.N at a 59 HRC hardness show a good behavior in this chloride
environment and confirm salt spray test results (Figure 12).
lO
Io
1o-
lO1o-
E
10"
10"
10'
[
I
-1000
,L,,
-500
U (SCE) [mV]
Figure 12 - Current Density - Potential Curves in O.9 % NaCl
Oxygenated Solution (37~ / 98. 6~
Drill Penetration and Cutting Tests
Drilling Test in bone
Conditions
Drills
Depth of holes
Speed
Cutting Fluid
Feed
Tool life End Point
: 3.2 mm gray drills

: 15ram
: 500 rpm
: Dry
: 25.5 mm/min.
: 0.030" or 300 hole limit
Test Result
X15TN (59HRc)
220 holes - 0.032" wear
AISI 630 (H900)
10 holes - 0.032"wear
AISI XM16 (H900) 35 holes - 0.030"wear
25
Cutting Test on CA. T.R.A. Machine

Most tests to qualify cutting behavior o f instruments are made with prototype parts put
on test bed which simulate the conditions o f use. Then it is difficult to match the data of
different laboratories. In an over hand the knife industry is using the C.A.T.R.A. machine
from British Cutlery Association. During this test, the maximum thickness which a
cutting device is able to cut is measured. All test parameters are fixed: paper quality,
speed and load.
The decreasing o f the maximum thickness reached is recorded test after test without edge
sharpening. A second way is to record the total thickness o f paper cut after a fixed
number o f tests.
In the Figure 13 we compare data for steels including AISI 440Amod. (0.55%C,
15%Cr), AISI 420HC (0.42%C, 13%Cr) and X 15TN for 50 mm/sec cutting speed and
45 N load.
lUDI
TOTALDEPTHCUT(ram)
AFTER60 CYCLES
i,.. \
I\
"~
1o.. ~. k
AIS1420HC
AIS1440A mod.
188.6
210.0
X15TN
412.2
', \
,oF \.
] ~u,.q.,~
,/
0
. . . . ".,,,, : , , . . .
,
~10
"",
20
30
I
rIO
I
U
NL~,519E~OF CW~r.J~8
Figure 13 : C.A.T.R.A Test
Workability
Hot Working is easy in the temperature range o f 2140/1830~ (1170/1000~

Cold Forming characteristics, in operations such as rolling, extrusion, drawing .... are
similar to a number o f other martensitic stainless steels o f 400 series.
Machinability, in annealed condition, is close to AISI 420 or 431.
26
For Cleaning/Passivation, the HNO3-Na2Cr2OT,2H20 solution, generally used with

martensitic steels is appropriate.
Welding is possible, either without filler metal or if needed with AISI 430 filler welding
wire.
Applications
As of today X 15 TN has found application as ( Figure 14 )
Figure 14
1. Disposable drills for trauma and femoral reamers used to implement hip prothesis.
2. Screwdrivers generally used to fix trauma devices and spine systems
3. Cutting guides used during knee operations in order to guide the saw which permit to
remorve the appropriate part of bone
4. Dental instruments such as microdrills
5. Saw blade used together with cutting guides during knee operations
6. Acetabular reamers used to remorve the cartilage before fixing the acetabular cup
linked to the head of the hip prothesis.
27
Conclusions
The composition choices made when developing X15T.N have allowed us to respond
in a balanced way to various and sometimes conflicting property requirements. XI 5TN
has a similar hardness to that o f AISI 440C, and due to its free and homogeneous
structure it exhibits good edge retension. Its corrosion resistance is greatly enhanced.
Moreover, X 15T.N offers a better toughness/fatigue than AISI 440C.Compared to
XM 16 or 630, heat treatment is not as easy for X 15 T.N and toughness is at a small
disadvantage, but cutting properties and wear resistance are greatly superior, while
corrosion characteristics are comparable. Machining is comparable to other martensitic
stainless steel e.g. AISI 420-431 and 440C.X 15 T.N is registered as UNS S 42025 and an
introductory paper was presented at the revision o f A S T M F 899, 1995 edition.
Charles H. Craig, 1 Herbert R. Radisch, Jr., 1 Thomas A. Trozera, 2 Paul C. Turner, 3

R. Dale Govier,3 Edward J. Vesely, Jr.,4 Nev A. Gokcen, 5 Clifford M. Friend, 6 and
Michael R. Edwards 6
Development of a Platinum-Enhanced Radiopaque Stainless Steel (PERSS| 7
Reference: Craig, C. H., Radisch, H. R., Jr., Trozera, T. A., Turner, P. C., Govier, R. D.,
Vesely, E. J., Jr., Gokcen, N. A., Friend, C. M., and Edwards, M. R., "Development of a
Platinum Enhanced Radiopaque Stainless Steel (PERSS|
Stainless Steelsfor
Abstract: Balloon-expandable coronary stents may be made o f stainless steel
conforming to ASTM or ISO specifications, referred to by A S T M as UNS $31673 alloys.
A need exists for an alloy with enhanced radiopacity to make stents more visible
radiographically and more effective clinically. A research program was initiated with the
objectives o f enhancing fluoroscopic radiopacity while maintaining adherence to the
A S T M and ISO specifications. These objectives were ultimately achieved by adding a
noble metal, platinum, to UNS $31673 by vacuum induction melting a commerciallyavailable alloy. Freedom o f the resulting microstructure from formation o f harmful
topologically close packed phases was ensured by use o f phase computation methodology
(New PHACOMP), and confirmed by X-ray diffraction and transmission electron
microscopy. Platinum was chosen since it is over twice as dense as nickel and, with
approximately half its effect as an austenitizer, allows nickel content to be reduced to a
minimum level.
Keywords: stainless steel, platinum, coronary stent, radiopacity, New PHACOMP
Introduction
Balloon-expandable coronary stents may be made o f a low-carbon, austenitic, nonmagnetic, implant grade stainless steel that is similar to AISI/ANSI 316L. Bar and wire
are described in ASTM Specification for Wrought 18 Chromium-14 Nickel- 2.5
Research & Development (R&D) Engineer and Vice President for R&D, respectively, Boston Scientific
Corporation/InterventionalTechnologies, 3574 Ruffin Rd., San Diego, CA, 92123.
z Consultant, 437 Pine Needles Dr., Del Mar, CA, 92014.
3 Associate Director for Thermal Treatment Technologies and Chemist respectively, Albany Research
Center, U.S. Department of Energy, 1450 Queen Ave., SW, Albany, OR, 97321.
4 Consultant, 387 River Bend Dr., Huntsville, AL, 35824.
s Consultant, 385 Palos Verdes Dr. W, Palos Verdes Estates, CA, 90274.
6 Professor and Head, and Senior Lecturer, respectively, Department of Materials and Medical Sciences,
Cranfield Postgraduate Medical School, Shrivenham, Swindon SN6 8LA, United Kingdom.
7 Registered trademark: Boston Scientific Corporation/InterventionalTechnologies, San Diego, CA.
28
www.astm.org
CRAIG ET AL. ON PERSS
29
Molybdenum Stainless Steel Bar and Wire for Surgical Implants (UNS S31673), F138,
while sheet and strip are described in ASTM Specification for Wrought 18 Chrominm14 Nickel- 2.5 Molybdenum Stainless Steel Sheet and Strip for Surgical Implants (UNS
$31673), F139. In composition and properties, F138 and F139 alloys parallel
Composition D of Part 1 (Wrought stainless steel) of International Standard ISO' 5832-1
(Implants for surgery - metallic materials). Low carbon content is stipulated to assure
freedom from intergranular corrosion susceptibility; F138 and F139 also require these
alloys to meet corrosion susceptibility tests. Pitting resistance is ensured by controlling
chromium and molybdenum weight percent, w, by the formula
Composition (C) = w Cr + 3.3 w M o > 26.
(1)
Concerns for both corrosion and fatigue resistance are addressed by maintaining a finegrained structure that is free from ferrite and low in nonmetallic inclusions.
Tubular, mesh-type coronary stents made from these alloys are placed through noninvasive procedures inside diseased arteries that have been identified as areas where
blood flow is restricted. The stent is mounted on a balloon catheter that may have gold or
platinum marker bands at each end, which will increase radiographic visibility by
increasing overall fluorescent radiopacity of the stent and deployment system, enabling
precise placement of the stent within the artery. Once coronary stents are deployed and
their deployment systems are removed, including marker bands, stents with poor
radiopacity can no longer be visualized radiographically. This represents a problem if it
is found that the stent needs further dilatation.
The design of the coronary stent is such that it is at its minimum diameter, in a
compressed state on the balloon catheter, to facilitate winding its way through arteries to
the point where it is to be deployed by pressurization of the balloon catheter. During the
expansion from a compressed to an extended state, the struts of the mesh-type stent may
be partially reduced in cross-sectional area due to plastic strain that results from having
surpassed the elastic limit of the alloy. After deployment, because of reduced thickness,
the struts may exhibit even greater indications of less-than-ideal fluoroscopic radiopacity.
To enhance the fluorescent radiopacity of these stainless steel stents, strut thicknesses can
be increased, but this will be at the expense of overall flexibility and expandability.
Alternatives to use of UNS $31673 alloys have been sought due to the less-than-ideal
radiopacities exhibited by deployed coronary stents made of these alloys. In their stent
handbook, Serruys and Rensing [ 1] compare most coronary stents that are available today
in terms of their mechanism for deployment and other features such as radiopacity and
flexibility. They noted that the following approaches have been used: tantalum, platinum
wire (as a core inside a cobalt alloy tube), and gold electroplating. Other means of
applying radiopaque coatings are a topic of current research interest [2]; these include
coatings of iridium, platinum and gold. These alternatives to UNS $31673 alloys have
used external means to enhance radiopacity or replaced it with more radiopaque material.
Background and Initial Approach

Boston Scientific Corporation/Interventional Technologies (BSC/IVT) developed a
30
STAINLESSSTEELS FOR MEDICAL AND SURGICALAPPLICATIONS
balloon-expandable coronary stent (LP Stent rMS) that is made from stainless steel foil.
The foil is formed into tubing and the stent is then fabricated by a photo-etch process.
To better understand the structural and functional embodiments of its LP Stent design,
IVT conducted an internal study [3] of available coronary stents at its corporate research
and development laboratories. The results of this study correlated well with IVT's initial
design study, which indicated UNS $31673 alloys represented the most universal solution
to the design and manufacturing features that were to be incorporated into the LP Stent.
However, all stainless steel stents that compared well to the flexibility ofBSC/IVT's LP
Stent were considered insufficiently radiopaque.
Since the photo-etch process chosen to fabricate the intricate LP Stent design was
specific to iron-base alloys, substituting tantalum would not only invalidate the work that
went into developing and refining this unique process and require an alternate process to
be developed, but it would potentially make the stent design too radiopaque. A platinum
and cobalt alloy wire design could not be applied to the LP Stent design or fabrication
process. With the electropolished finish, coatings would not adhere well to the LP Stent
and may be brittle, which could compromise the flexible design.
Having considered the above, it was decided to establish a research program to study
the addition of an additional element, tantalum, unspecified in F139, to the UNS $31673
alloy, thus maintaining adherence to this specification while increasing its radiopacity.
In the United States and elsewhere, government regulatory bodies monitor evolution
and implementation of all new stent designs, their materials of construction, and analyze
risks associated with in vivo and in vitro testing. Since radiopacity is a functional rather
than a structural requirement, F138 and F 139 alloys have not been subjected to regulation
in the past. ASTM Committee F04 (Medical and Surgical Materials and Devices) has
recently addressed this issue and a document is in draft. ASTM Test Methods for
Radiopacity of Plastics for Medical Use, F649m discusses radiopacity and provides
several comparative approaches based on use of film radiography. F640 refers to the
attenuation law (Beer-Lambert), citing ASTM Terminology Relating to Metallography,
E7, where transmissivity [4], T, the ratio of transmitted X-ray intensity It to incident
intensity/~ is
It / Ie = exp [- (kt) X ] = T
(2)
where ~t is the linear attenuation coefficient and X the thickness of the absorbing material.
The above applies for an idealized case [5] of a narrow-beam X-ray photon of incident
intensity Ie penetrating a layer of material of thickness X (where laX is dimensionless).
The remaining photons emerge from the attenuator with transmitted intensity It according
to the Beer-Lambert law, Eq. 2. Published data are often listed as the mass attenuation
coefficient, alP, where p in g/cm 3, is the density of the element. Values of densities and
of linear and mass absorption coefficients, at 80 keV and at 100 keV, are in Table 1 for
the elements Cr, Fe, Ni, Mo, Ta, and Pt. As shown in Eq. 3, the ratio &incident X-ray
intensities to transmitted intensities is the radiopacity [4], R, of the attenuating material
Ie / It = exp [(~) X ] = R
(3)
8Trademarkapplied for: BostonScientific Corporation/InterventionalTechnologies,San Diego, CA.
31
Table 1 - Linear and Mass Absorption Coefficients and Densities [4]
p
~p
80 keV
Ix
80 keV
Cr
7.180
Fe
7.874
Ni
8.902
Mo
10.220
Ta
16.650
Pt
21.450
0.491
0.595
0.731
1.962
7.587
8.731
3.522
4.687
6.504
20.052
126.324
187.28
0.317
0.372
0.444
1.096
4.302
4.993
2.273
2.927
3.952
11.201
71.628
107.10
Cp
100 keV
Ix
100 keV
Eq. 3 shows that as either the linear absorption coefficient or the thickness o f the material
is changed, so does radiopacity. For film radiography, we have these same relationships,
where the optical transmission density o f the radiographic film, D, is as shown below [5]
D = logI~/lt.
(4)
Here Ie is incident light from the optical densitometer on the radiographic film (absorber),
It is the light transmitted through the absorber, and Ie / It is the opacity o f the absorber.
From the above, with stent wall thickness X = 12.7 x 10 -3 cm, we calculate T and R
for Ta and Pt and for primary elements o f an alloy similar to UNS $31673, Table 2.
Table 2 - R a d i o p a c i t i e s & Transmissivities for Ta, Pt, and Primary Elements of 316L
Ta
4.974
0.201
2.484
0.403
Rso~v
T 8Okay
Rlook~V
Tl0o key
Pt
10.788
0.093
3.897
0.257
Cr
1.046
0.956
1.029
0.972
Fe
1.061
0.942
1.038
0.964
Ni
1.086
0.921
1.051
0.951
Mo
1.290
0.775
1.153
0.867
Table 3 shows the Table 2 values for Ta and Pt, while Table 4 shows calculated values,
using appropriate weight percentages, for primary elements of a 316L type stainless steel.
Low transmission o f both Ta and Pt contrasts with extremely high transmission o f 316L.
Table 3 - Calculated Radiopacities & Transmissivities for Tantalum and Platinum
Ta
100
-
W
W
Pt
Rs0kev
4.974
10.788
100
Tsokev
0.201
0.093
Rlookev
2.484
3.897
Tl~keV
0.403
0.257
Table 4 - Calculated Radiopacities & Transmissivities for Primary Elements of316L
18
65.5
14
2.5
1.066
0.938
1.040
0.961
32
As can be seen from Tables 5 and 6, when stainless steel with either tantalum or
platinum is compared to the native 316L-type alloy, a decrease in transmissivity is
obtained based on the enhanced radiopacity of the resulting alloy.
Table 5 - Calculated Transmissivities f o r a Stainless Steel Containing 5 w Ta or 5 w Pt
Cr
18
18
W
W
Fe
61.5
61.5
Ni
13
13
Mo
2.5
2.5
Ta
5
-
Pt
5
Ts0 kr
0.901
0.893
Tiook~v
0.940
0.934
Table 6 - Calculated Radiopacities for a Stainless Steel Containing 5 w Ta or 5 w Pt

Cr
18
18
W
W
Fe
61.5
61.5
Ni
13
13
Mo
2.5
2.5
Ta
5
-
Pt
5
Rso kev
1.110
1.120
R~ookeV
1.064
1.070
Initially, the argument was in favor o f adding tantalum since it is not a noble metal
and thus would be much more economical as an alloying element. It is a carbide former,
used in austenitic steels such as AISI 347 and AISI 348 in amounts up to 0.9 w to
decrease the alloy susceptibility to intergranular corrosion [6]. While determining the
solubility o f tantalum in UNS $31673 alloys is usually considered too difficult, a
quaternary stainless steel can be devised from binary data [7], comprised of 18 w Cr, 14
w Ni; added fourth (high-density) element, M, balance Fe. For such an alloy, it was
estimated that roughly 2.9 w Ta would be soluble.
An experiment was devised in which industrial radiographs were taken o f a stainless
steel base partially covered with a foil step-wedge made of tantalum; the thicknesses
were approximately 2.5 %, 4.0 %, and 5.0 % o f the stainless steel base. This allowed the
use o f an optical film densitometer to compare the resulting film densities. Based on
these results, it was determined that between 1 and 5 w tantalum would be required to
slightly enhance radiopacity o f coronary stents made o f the resulting alloy over the X-ray
energy range used in a catheterization laboratory (-65 keV to ~120 keV).
Tantalum was added to a commercial UNS $31673 alloy meeting F138 specifications,
BioDur 9 316LS and cast in 100 g ingots. Five ingots containing nominally 1.0 w to 5.0 w
Ta in the stainless steel matrix were cast. Compositions of the alloys are in Table 7.
Table 7 - Compositions o f Remelted BioDur 316LS containing Ta
Nominal w
Actual w
IVT # 18
1.00
0.84
IVT # 19
2.00
1.87
IVT # 20
3.00
2.58
IVT # 21
4.00
3.56
IVT # 22
5.00
4.40
The production of the alloys and their subsequent rolling was accomplished at the U.
S. Department o f Energy's Albany Research Center (ARC). These ingots were made in
an arc furnace, using a triple-melt procedure and one-third atmosphere argon; the ingot
9 Registered trademark of Carpenter Technology Corporation, Reading, PA.
33
(button) was contained on a water-cooled copper hearth and the components were melted
using a thoriated W tip to establish and control the arc, much as a welding rod might be
used. The buttons were heated slowly to 870~ then heated rapidly to 1150-1260"C;
they were then hot-rolled from an average thickness of 8.5 mm to 3.0 mm in 15 %
reductions, with five minutes for reheats between passes, annealed at 1040"C and waterquenched. The resulting shape was then milled to the thickness (2.54 mm) of the
stainless steel base that had been radiographed along with the tantalum step wedge so the
radiographic results could be compared. The film density measurements, D, compared
well with those from the plate-stepwedge.
X-ray diffraction results of the five cast and rolled specimens indicated the presence of
an eta phase, Fe2Ta, which is a Laves-type topologically close-packed (TCP) structure [6]
that began to appear with the 1.91 w Ta specimen. This indicated that the solubility of
tantalum in the BioDur 316LS alloy was less than initially predicted. The solubility for
tungsten can be similarly predicted to be about 13.6 w. Tungsten is attractive from the
point of view that, like tantalum, it is a dense element that would promote radiopacity if
added to the stainless steel matrix. The possibility of using tungsten was pursued as a
companion experiment to the use of tantalum; five 100 g buttons were cast as previously
described, containing from 1.0 to 5.0 w W. These were rolled into plate, machined, and
radiographed, confirming radiopacity expectations. However, at 4.0 w W and above, Xray diffraction clearly indicated the presence of ferrite.
Analysis
From the above experiments, it was obvious that sufficient amounts of tantalum or
tungsten would not be soluble in the 316LS matrix to achieve the objectives of enhancing
a functional requirement, radiopacity, while maintaining or improving upon structural
requirements. Analysis of these experiments entailed looking more closely at the
problem to be solved and determining how to proceed in order to succeed with another
alloying element since the functional (enhanced radiopacity) experiments were
considered to be successful.
In AISI 347 and AISI 348 austenitic stainless steels, tantalum and niobium are added
because they are carbide formers and may be used [4] singly or together, in amounts of
up to ten times the carbon content in these alloys, to stabilize the grade against chromium
carbide formation. This decreases the susceptibility to chromium depletion, which would
lead to intergranular corrosion in harsh environments. In the UNS $31673 matrix, there
is no excess carbon, thus in the case of tantalum additions, there is a Laves-type TCPstructure, the FezTa eta phase, and at higher tantalum contents, some indications of the
formation of ferrite. The tungsten additions clearly formed ferrite, along with an
unidentified phase, especially at higher contents. Tantalum and tungsten are known
ferrite-formers [6], so the presence of ferrite in these alloys it to be expected.
Approach
The precious metals, iridium, platinum and gold can be considered as potential
alloying elements. Biocompatibility would thus not be a first-order issue if one of these
were to be selected as an unspecified additional element in UNS $31673 alloys.
34
STAINLESSSTEELS FOR MEDICALAND SURGICALAPPLICATIONS
Solubility of these elements in the simple quaternary described above (18 w Cr, 14 w Ni,
M, balance Fe) is excellent; we find, at 900*C, iridium and platinum are completely
soluble in this quaternary and gold can be dissolved up to 37 w. The maximum amounts
of noble elements envisioned to enhance radiopacity of coronary stents made by adding
such unspecified additional elements to UNS $31673 are well below these solubility
levels. When considering the use of iridium, platinum and gold as alloying elements,
platinum was selected, in spite of its high price, since iridium was even more expensive
and gold does not act as an austenite stabilizer [8].
UNS $31673 alloys must meet ASTM requirements for ferrite content and inclusion
content. The presence of topologically close packed phases (TCP) in such alloys is
unacceptable because of their effect on alloy ductility
The approach applied by DuPont [9] and described recently by its originators
Morinaga et al [10] is used to predict the onset of TCPs; New PHACOMP was developed
for and is principally used in the development ofNi-base superalloys. The original
PHACOMP, which New PHACOMP would replace, emerged during the 1960s in
response to TCP (sigma phase) problems with superalloys in high-temperature service
and is attributable to Boesch and Slaney [11]. An overview of the state of the art of the
original PHACOMP technology, circa 1987, is given by Sims [12]. Barrows and
Newkirk [ 13] provide a model that describes the gamma phase-sigma phase boundary
(solvus) more closely than earlier works, in terms of concentration and temperature. The
methodology of [ 10] was developed in the early 1980s to address problems encountered
in using Ni3A1 (gamma-prime) arising from the use of aluminum in nickel-base
superalloys. The so-called cluster model surrounding nickel aluminide, with its formula
MNil2A16, led Morinaga et al to obtain the effect of a transition metal M on the electronic
structure of the nickel aluminide in terms of energy levels (eigenvalues), which are
expressed in electron volts (eV) relative to the Ni3A1 Fermi level; this level is arbitrarily
set to zero for both Ni and Ni3A1. These eigenvalues, due to the d-orbitals of the
transition metals, are referred to as Md levels, or parameters, expressed in eV. In [10], a
list of elements typically used in superalloys and steels is provided, along with their
respective Md parameters. The New PHACOMP approach defines the average Md,
which is the compositional value for an alloy, as
Md(avg) = Y~[ x~ ( M d ) J
(5)
where Xi is the atomic fraction of component element i and (Md)i is the value of Md for
each element i constituting the alloy. It was shown by Morinaga et al that the values of
Md(avg) roughly approximate the single phase gamma boundaries in a large number of
ternary alloys. Further, as an example, it was shown that when Md(avg) was greater than
0.925 eV at 1200~ the TCP phases were precipitated out of the gamma phase.
It was decided to use New PHACOMP to determine whether TCPs would form on
adding certain unspecified additional elements to a UNS $31673 matrix. At the time, the
Md parameters for the three noble metals under consideration (Ir, Pt, Au) had not been
published and assumed values were utilized, based on the Md parameters available. The
35
Md for Ir was subsequently published in [10]; Morinaga has also provided l~ the Md
parameters for Pt (0.764 eV), Au (0.627 eV), Pd (0.779 eV), and Ag (0.659 eV), along
with corresponding bond orders (Bo) o f these elements, as in [ 10], with Bo parameters of
0.920 eV (Pt), 0.528 eV (Au), 0.751 eV (Pd), and 0.391 eV (Ag).
Using appropriate Md parameters from the above discussion for the hypothetical
quatemary alloy discussed above, we find the average Md parameters are greater for Ta
and W but approximately equal for Ir, Pt, and Au
Table 8 - Md(avg) for Hypothetical Quaternary with 0 w and 5 w Ta, W, Ir, Pt, Au
5 wt% Ta
5 wt% W
5 wt% Ir
quaternary
5 wt% Pt
5 wt% Au
Md(avg) =
0.918 eV
Md(avg) =
0.909 eV
Md(avg) =
0.897 eV
Md(avg) =
0.895 eV
Md(avg) =
0.895 eV
Md(avg) =
0.893 eV
Using appropriate Md parameters in TCP-free F138 alloy (Carpenter BioDur 316LS),

average Md parameters increase with the BCC elements but remain stable with Ir, Pt, Au
Table 9 - Md(avg) for BioDur 316LS with 0 w and 5 w Ta, W, Ir, Pt, Au
5 wt% Ta
+ 316LS
Md(avg) =
0.936 eV
5 wt% W +
316LS
Md(avg) =
0.927 eV
5 wt% Ir +
316LS
Md(avg) =
0.915 eV
BioDur
316LS
Md(avg) =
0.915 eV
5 wt% Pt +
316LS
Md(avg) =
0.913 eV
5 wt% Au
+ 316LS
Md(avg) =
0.911 eV
For Ir only in a 316LS base, we have the following average Md parameters

Table 10 - Md(avg) for BioDur 316LS with 0 w to 60 w lr
BioDur
316LS
Md(avg)
0.915 eV
=
7.5 w i t +
316LS
Md(avg) =
0.915 eV
15 w l r +
316LS
Md(avg)
0.915 eV
=
30wit+
316LS
Md(avg) =
0.914 eV
45 w l r +
316LS
Md(avg) =
0.914 eV
60wlr+
316LS
Md(avg) =
0.913 eV
These 100 g ingots o f platinum containing alloys were cast, rolled, annealed, and
machined to shape. X-ray diffraction was used to determine the presence o f either TCP
phases or ferrite. Whereas tantalum, with an Md of 2.224 eV, resulted in an alloy
containing a TCP (eta phase), and tungsten, with an Md o f 1.655 eV, resulted in an alloy
containing ferrite, the diffraction results showed an absence o f ferrite or TCPs in the
BioDur 316LS containing platinum. Radiopacity measurements, as with tantalum and
tungsten, were as expected; sufficient enhancement in radiopacity of the resulting
coronary stents would be provided by approximately 5.0 w Pt. Thus it was decided to cast
a 50 kg ingot in order to try all applicable manufacturing processes. Later, a further series
10 Morinaga, M., Nagoya University, Nagoya, JP, personal communication with Charles
Craig, Interventional Technologies, Inc., San Diego, CA, October 2001.
36
of small ingots, with platinum contents up to 10 w, was cast. These were then processed
as before and subjected to the same analysis; no indications of TCPs were found and
radiopacity results compared well with expectations. Tubes were then manufactured from
the 5 w Pt foil. These tubes were examined by optical and transmission electron
microscopy (TEM). No indications were found of any of these tubes containing TCPs.
Although enhancing the radiopacity of the UNS $31673 alloys by intentionally
adding platinum as unspecified additional elements appears to be new, alloying stainless
steels with platinum group metals for other purposes is not new, dating back to 1911 [13].
More modem efforts to use such materials appear to be based on extensive, corrosionrelated research conducted in the former Soviet Union beginning in the late 1940s, which
first reached a large audience outside that community in a translated textbook [14] during
the late 1960s. McGill [15, 16] has reviewed the effects of platinum group metals on the
corrosion and mechanical properties of stainless steels.
Approximately 14 kg of the first 50 kg ingot containing 5 w platinum in 316LS was
rolled into sheet and then into foil and tubing suitable for prototype fabrication of a nextgeneration LP Stent. This was examined for pitting resistance and susceptibility to
corrosion in accordance with F138 and F139 and is reported by Covino et al [17], where
it is shown that this alloy fully complies with the pitting and corrosion requirements of
F138 and F139.
Summary and Conclusions
The addition of platinum to a commercially-available alloy complying to F138,
BioDur 316LS, has provided enhancement in the radiopacity of the resulting stainless
steel, referred to here as PERSS. PERSS may not qualify under F138 or F139 due to the
intentional addition of dense elements, but the criteria of those specifications is used for
reference purposes. For this new alloy to be useful for coronary stents the modified alloy
must still be capable of being processed and to survive in vivo. Corrosion behavior
(intergranular corrosion susceptibility; pitting resistance), as well as metallurgical and
mechanical properties, to include freedom from formation of delta ferrite and low
microinclusion content, must be similar to BioDur 316LS, as specified by F138 and
F 139. These aims are achieved by BioDur 316LS stainless steels alloyed with platinum.
References
[1]
Serruys, P. W. and Rensing, B., Eds., Handbook of Coronary Stents,

Martin Dunitz, Ltd., London, 2001.
4 th edition,
[2]
Sahagian, R. "Critical Insight: Marking Devices with Radiopaque Coatings,"

Medical Device & Diagnostic Industry, Vol. 21, No. 5, 1999, pp. 160-166.
[3]
Radisch, H. "Comparisons of the LP Stent with Other Stent Designs - 2 n a

edition," internal report, Interventional Technologies, Inc., San Diego, CA,
1999.
[4]
Hubbell, J. H. and Seltzer, S. M. Tables of X-ray Mass Attenuation Coefficients
37
andMass Energy-Absorption Coefficients (Version 1.03), National Institute

of Standards and Technology (NIST), 1997 [Online]. Available:
http://physics.nist.gov/xaamdi [2002, February 17], NIST, Gaithersburg, MD.
[5]
ASM International, ASM Handbook| Volume 17: Nondestructive Evaluation and

Quality Control, ASM International, Materials Park, OH, 1989, p. 323.
[6]
Beddoes, J., and Parr, J. G. Introduction to Stainless Steels, 3 ra edition, ASM

International, Materials Park, OH, 1999, pp. 137-138.
[7]
Okamoto, H. Desk Handbook: Phase Diagrams for Binary Alloys, ASM

International, Materials Park, OH, 2000.
[8]
Moema, J., and Paton, R., "The Influence of Gold and Ruthenium on the
Austenitic Stability of Type 316L Stainless Steels," internal report by Mintek
(Randburg, SA), for Interventional Technologies, Inc., San Diego, CA, 2001.
[9]
DuPont, J. N., "A Combined Solubility Product/New PHACOMP Approach for

Estimating Temperatures of Secondary Solidification Reactions in Superalloy
Weld Metals," Metallurgical and Material Transactions, Vol. 29A, 1998,
pp. 1449-1456.
[10]
Morinaga, M., Murata, Y., and Yukawa, H., "Recent Progress in the New
PHACOMP Approach," Materials Design Approaches and Experiences
(Proceedings of Symposium sponsored by the High Temperature Alloys
Committee of the Structural Alloys Division (SMD) of TMS (The Minerals,
Metals & Materials Society), Zhao, J. -C., Fahrrnann, M., and Pollock, T. M.,
Eds., TMS, Warrendale, PA, 2001, pp. 15-28.
[11]
Boesch, W. J., and Slaney, J. S., "Preventing Sigma Phase Embrittlement in

Nickel Base Superalloys," Metal Progress, July 1964, pp. 109-111.
[12]
Sims, C. T., "Prediction of Phase Composition," Superalloys 11, Sims, C. T.,

Stoloff, N. S., and Hagel, W. C., Eds., Wiley, New York, 1987, pp. 217-239.
[13]
Monnarz, P., "Contribution to the Study of Ferrochrome Alloys, with Special

Consideration Given to Acid Resistance," Metallurgie, Vol. 8, No. 7, 1911,
pp. 160-176.
[14]
Tomashov, N. D., and Chemova, G. P., Passivity and Protection of Metals

Against Corrosion, Plenum Press, New York, 1967.
[15]
McGill, I. R., "Platinum Metals in Stainless Steels: A Review of Corrosion and

Mechanical Properties," Platinum Metals Review, Vol. 34, No. 2, 1990, pp. 8597.
38
[16]
McGill, I. R. "Platinum Metals in Stainless Steels: Part II: Further Corrosion and
Mechanical Properties," Platinum Metals Review, 34, No. 3, 1990,
pp. 144-154.
[17]
Covino, B. S., Jr., Craig, C. H., Cramer, S. D., Bullard, S. J., Ziomek-Moroz, M.,
Jablonski, P. D., Turner, P. C., Radisch, H. R., Jr., Gokcen, N. A., Friend, C.
M., and Edwards, M. R., "Corrosion Behavior of Platinum-Enhanced
Radiopaque Stainless Steel (PERSS| for Dilation-Balloon Expandable
Coronary Stents," Stainless Steels for Medical and Surgical Applications,
ASTMSTP 1438, G. L. Winters and M. J. Nutt, Eds., ASTM International,
West Conshohocken, PA, 2003.
Markus Windler, l and Rainer Steger ~
Mechanical and Corrosion Properties of Forged Hip Stems made of High-Nitrogen

Stainless Steel
Reference: Windier, M., and Steger, R., "Mechanical and Corrosion Properties of
Forged Hip Stems made of High-Nitrogen Stainless Steel," Stainless Steels for
Medical and SurgiealApplications, ASTM STP 1438, G. L. Winters and M. J. Nutt, Eds.,
Abstract: The purpose of the study was to determine the mechanical and corrosion
properties of forged, commercially available, femoral hip stems. Two different stem
designs were selected (the MS-30, sizes 6 and 16, and the MUELLER straight stem, sizes
7.5 and 15.0). They were made of high-nitrogen stainless steel according to ISO 5832-9
and ASTM F1586. The samples for determining the mechanical properties were taken
from the proximal part of the prostheses. The metallurgical and the potentio-dynamic
corrosion behaviours were investigated in three different transverse cross-sectional
planes. The mechanical properties in the forged condition were as follows: Yield strength
1051 - 1172 MPa, tensile strength 1138 - 1238 MPa and elongation 15 - 19%. Vickers
hardness values of between 354 and 393 HV were measured. The breakdown potential
fell from +1000 mV(scE) (wrought bar) to +600 to +950 mV(scE) (forged hip stems) for
pH 4 and 37~ The rotational bending fatigue value for 10 million cycles was 587 MPa.
Keywords: high-nitrogen stainless steel, corrosion properties, mechanical properties,

mierostructure
Nomenclature
Breakdown potential - The least noble potential where pitting or crevice
corrosion, or both, will initiate and propagate [ASTM Standard Terminology Relating to
Corrosion Testing (G15)].
1Director, Materials Research, and Research Scientist, respectively, Material Research

Department, Sulzer Orthopedics Ltd., P.O. Box 65, 8404 Winterthur, Switzerland
39
www.astm.org
40
Introduction
Stainless steel has been employed as surgical implant material for many years.
The quality of the steel in the pioneering days was poor, and numerous deficiencies such
as fatigue failures and severe corrosion were reported [1]. The breakthrough of stainless
steel as an implant material came with the development of the material with the
designation AISI 316L. This type of steel is still used today, primarily for the products of
osteosynthesis like plates, nails and screws [2,3]. Endoprostheses made of this type of
steel since the 1960s, e.g., the Charntey hip stem, have been implanted and retained
successfully in the body for many years [4-6]. Reports of corrosion phenomena and
fatigue failures of this steel in-vivo have been made in the literature [2,4,7]. A new type
of steel was introduced in the field of orthopedic surgery in the mid-1980s [4]. The
chromium and manganese contents of the alloy were raised and up to 0.5% nitrogen was
also introduced to improve the corrosion resistance and increase the strength [8]. This
material was standardized under the designation "Wrought High Nitrogen Stainless Steel"
1992: ISO 5832-9 and ASTM F1586 - 95. The application for this highly alloyed
stainless steel in the field of orthopedic surgery have been modular, cemented hip
prostheses and femoral heads, which articulate against polyethylene. In this study, we
analysed the mechanical, metallurgical and corrosion properties of forged hip stems.
Material
The smallest and the largest size of the "MS-30" hip stem design, as well as the
smallest and third largest size of the "MUELLER straight stem" design were investigated
(Table 1). All femoral stem components were taken from the regular production of Sulzer
Orthopedics Ltd., Switzerland, and were made from the highly alloyed stainless steel
"high-nitrogen stainless steel," tradename ProtasuVM-S30. As far as the chemical
composition and the mechanical properties are concerned, the material fulfilled the
requirements of ISO 5832-9 and ASTM Standard Specification for Wrought Nitrogen
Strengthened -21 Chromium - 10Nickel -3 Manganese -2.5 Molybdenum Stainless Steel
Bar for Surgical Implants (F 1586). The stainless steel bars with a diameter of 22 mm
were mill annealed and formed into hip prostheses by means of the "drop-forging
process." The transverse cross-section (anterior-posterior) of the MS-30 hip stem
becomes larger as the size increases, whereas the cross-section (anterior-posterior) of the
MUELLER straight stem remains the same even though the size increases (Figure 1).
Table 1 -Investtgated Hip Stem Design, Size and Number ofused Specimen
Hip stem design
Size
No. of Specimen
MS-30
6 (smallest)
10
MS-30
16 (largest)
10
MUELLER straight stem
7.5 (smallest)
10
MUELLER straight stem
15.0
25
WINDLER AND STEGER ON FORGED HIP STEMS
41
Figure 1 - M S - 3 0 hip stem, size 6 (top) and MUELLER straight stem, size 7.5 (bottom).
Microstructure analysis and corrosion tests were done at planes A, B and C
Methods
Tensile Properties
The test specimen for the tensile properties were machined from the proximal
prosthetic stem of forged blanks. The gauge diameter of each tensile specimen was 6 mm,
except for those samples made from MUELLER stems, size 7.5. The throat diameter of
these stems was 4 mm. Six tests were performed for each stem type in accordance to
European Standard, Metallic Materials; Tensile testing; Part 1: Method of testing
(10 002). The elongations were determined for 5D gauge lengths. The mechanical values
were statistically analysed with the Student t-test (p < 0.05).
Microstructure Analysis
The microstructure analysis and also the corrosion tests were made on three
transverse planes of the hip stems (A - C). Plane A was located 15 mm below the face of
the taper; plane B was two-thirds in length between distal tip and proximal shoulder of
the prosthesis; and plane C was 25 mm from distal tip of the stem (Figure l). The
42
metallographical sections were etched and examined with 500-times magnification under
a high power light microscope (Leica DMRX, Wetzlar, Germany).
Hardness Measurement
The Vickers hardness (HV10) was determined at the two ends of the tensile
specimen. The measurements were made in accordance to ISO, Metallic materials Vickers hardness t e s t - Part 1: Test methode (6507-1). All hardness measurements were
made 3 times and averaged.
In addition, the Vickers hardness was measured on all three metallographical
planes of the MS-30 stem, size 16, and the MUELLER straight stem, size 15.0. The
measuring positions were located anterior, posterior, lateral, medial and in the center.
Potentio-dynamic Corrosion Test

Potentio-dynamic polarization curves of the hip stems were determined per ASTM
Standard Reference Test Method for Making Potentiostatic and Potentiodynamic Anodic
Polarization Measurements (G5) using an EG&G (VersaSat [I) Potentiostat, EG&G
(SoftCorr III) Software and an EG&G K46 corrosion cell (EG&G Instruments, Princeton,
USA). The measurements were made in the cross-section planes A, B and C (Figure 1).
The coupon surfaces were ground to 1200 grit and masked with a self-dried polymer
(Erne Galvanotechnik, Dattlikon, Switzerland, Product: Abdecklack No. 1). The test
surfaces were measured in size and immersed in 1 litre of sodium chloride solution
(0.9%) at 40~ pH 4 (adjusted with HCI) and with nitrogen bubbling during the whole
process. A cathodic pre-polarization (2 minutes at 100 ktA/cm2) was made 250 mV
negative to the rest potential to clean the immersed surface. After waiting 30 minutes to
allow the system to come to equilibrium, the test was started with a scan rate of 10
mV/min. Saturated calomel electrodes (SCE) were used for reference purposes. All the
tests were made twice, reusing the old sample after refinishing.
Rotating Beam Fatigue

A high cycle rotating beam fatigue test were used to determine the fatigue
properties with constant force mode. The test conditions were: 100 Hz for 10 million
cycles in air at room temperature and stress ratio R = -1. All specimens had a diameter of
4 mm and were grit blasted (Ra 1 - 2 ~tm). Sixteen (16) specimens were machined from
the proximal part of the MUELLER stems~ size 15.0. Probit analysis was used to
calculate the fatigue strength at 50, 90, 95 and 99% survival.
Results
Tensile Properties
The tensile properties are summarized in Table 2 and Figures 2, 3. With the
MS-30 stem, the tensile strength reduces significantly, namely to the extent of 8.1%
(p<0.001), from size 6 to size 16, and 10.5% in the case of the yield strength
(p<0.00l). The Vicker hardness also declines by 9.8% (p<0.001), but there is no
recognizable difference in the elongation (p=0.75). In contrast to this, no statistical
difference could be established between the investigated MUELLER straight stems
size 7.5 and 15.0. This also applies for the yield strengths values (p>0.3), the
hardnesses (p>0.3) and the elongations (p>0.7).
The relations of hardness and tensile properties is shown in Figure 4. The
yield strength and tensile strength rise linearly with increasing hardness.
Table 2 - Tensile Properties of Forged Hip Stems, Design MS-30 Size 6 and 16 and
MUELLER Straigth Stem Size 7.5 and 15.0
Design
Size
Yield Strength [MPa]
Tensile Strength [MPa]
Elongation [%]
Reduction of Area [%]
Vicker Hardness
MS-30
6
1172 19
1238
15 3
59.0 1.5
393 11
MS-30
16
1051 + 16
1138:~ 11
16 2
62.0 1.5
354 5
1300
1300
1250
1250
1200
g, 12oo
1150
-~ 1150
1100
r~
MUELLER
7.5"
1115
1197
19 3
66.7 + 1.0
370 + 5
MUELLER
15.0
1105 14
1186
18 3
"60.0 0.6
373 4
t II Size 7.5
[] Size 15.0
11oo
1050
1050
1000
1000
TS
YS
Figure 2 - Mechanical strength of forged

MS-30 hip stems, size 6 and 16
TS
YS
Figure 3 - Mechanical strength of forged

MUELLER hip stems, size 7.5 and 15.0
43
44
STAINLESSSTEELS FOR MEDICAL AND SURG!CAL APPLICATIONS
1300
TS = 2.56x + 233
1200
1100
YS R23. l ~ ; ; 50
r/3
1000
TS
YS
900
.
340
350
.
360
370
380
Vicker Hardness
390
400
Figure 4 - Relation between mechanical strength and Vicker hardness.

NOTE." TS = Tensile Strength, YS = YieM Strength
All the metallographical sections of the "high-nitrogen stainless steels" had

an austenitic matrix. No delta ferrite could be observed up to 100x magnification.
As a result of the forging process, the structure is deformed badly in the vicinity of
the flash. Two types of inclusions were observed in all the transverse cross-sections
of the prostheses. Type 1 are fine, homogeneously distributed [Cr0 Mo]-carbide
with a diameter of less than 1 ~m. Type 2 are larger inclusions with a diameter of
up to 10 ~tm and identified as [Cr, Nb]-nitrides. No inclusions could be observed at
the grain boundary. The microstructure in the necks of all of the stems (plane A)
was recrystallized in the anterior and posterior regions of the taper (Figure 5) and
was elongated near the flash in the medial and lateral regions (Figure 6).
Vicker Hardness
The average hardnesses measured on the tensile specimens were between 354 HV
for the MS-30 size 16 and 393 HV for the MS-30 size 6 (Table l). The values for the
MUELLER hip stems were 370 HV (size 7.5) and 373 HV (size 15.0).
The hardness values measured in different areas in the three cross-section planes
are shown in Table 3. The highest values were always measured near the flash (medial
and lateral) and in the center of the femoral stems. Compared with the tensile specimen
the Vicker hardness measured in plane B was between 7 and 10% higher.
Figure 5 -Microstructure in plane A (neck

region) which is fully recrystallized
Figure 6 -Microstructure in plane A (neck

region) with elongated grains near the flash
Table 3 - Vicker Hardness Values Measured at Three Sectional Planes A - C and

Different Locations. Hardness Values from the Tensile Specimen were at the Bottom
Plane
A
A
A
B
B
B
C
C
C
Position
anterior / posterior
medial / lateral
center
medial / lateral
center
medial / lateral
center
tensile specimen
45
MS-30, size 16
356
433
408
361
385
393
346
405
418
354 -4-5
MUELLER, size 15.0

327
403
4l8
398
407
402
406
449
424
373 -4-4
46
Corrosion Test
The breakdown potentials ranging from +800 to +980 mV(scE) was measured on
the investigated planes A - C o f the hip prosthesis designs (Figure 7). No differences
could be established with respect to the corrosion-resistance as a function o f the
prosthesis size and/or location. The current density in the passive range between 0.1 to 1
i.tA/cm2.
1.E-02
1 .E-03
<
1.E-04
Conditions
0.9% NaCI
pH4, 40~
scan rate 10mV/min
1.E-05
1 .E-06
r.)
1 .E-07
1.E-08
-800
-600
.
-400
-200
.
200
.
400
.
600
800
1000
1200
Potential [mV(SCE)]
Figure 7 - Potentio-dynamic corrosion with forged MUELLER hip stem, size 7.5. The
breakdown potential in the cross-sectional plans A to C ranging between +840 and

+980 mV(scE)
Rotation Beam Fatigue

The plot o f all fatigue specimens are in Figure 8. At 107 cycles, the fatigue
strength o f forged high-nitrogen stainless steel were calculated as follows:
50%
90%
95%
99%
survive
survive
survive
survive
692
634
618
587
MPa
MPa
MPa
MPa.
750
47
700
650
600
550
500
,
0.1
10
100
Million Cycles
Figure 8 - Rotation banding fatigue curve in air at room temperature. Black spots
represent non fractured samples and white spots fractured samples
Discussion
The minimum mechanical properties for the investigated, highly alloyed nitrogen
stainless steel in the forged state are:
9
9
9
9
Tensile Strength
Yield Strength
Elongation
Fatigue Strength
TS
min.
YS
min.
E
min.
(107 cycles)
1100 MPa
1000 MPa
10%
587 MPa
The mechanical properties of forged hip prostheses can be determined easily in

tensile and fatigue tests. On a daily basis, hip stem components are mechanically loaded
in every single step over many years in the patient. Consequently, it is important robe
conversant with the fatigue limit of the final product. However, the documentation of the
endurance fatigue limit of the device was not the objective of this investigation and has
not been measured in this study. It is possible that this alloy may exhibit greater reduction
in its fatigue strength under such conditions compared to CoCrMo alloys.
Smethurst [9] investigated the properties of cold-worked high-nitrogen stainless
steel (Ortron 90). The tensile properties were a bit lower (yield strength 810 MPa, tensile
strength 1150 MPa, elongation 15%) and similar values for Vicker hardness (HV 413)
and fatigue strength (583 MPa at 107 cycles) than this study.
A literature comparison with other metallic materials, which are employed for
endoprostheses, shows that CoCrMo cast alloy has the lowest mechanical strength of
them all. This applies, first and foremost, to the fatigue behaviour. The reasons for this
48
are the coarse as-cast structure and sporadic microporosities. With the aid of the isostatic
process (HIP), however, it is possible to reduce the number and size of the pores. As far
as the mechanical properties are concerned, high-nitrogen stainless steel" is comparable
with those of the wrought CoCrMo alloys [10]. This applies to both high-carbon and lowcarbon CoCrMo alloy.
Alpha/beta and beta titanium alloys, which are employed primarily for cementless
applications, exhibit lower tensile strength but similar fatigue strength [ 10]. In addition,
the fatigue behaviour of this titanium alloy is notch-sensitive [ 11].
Truman [8] did potentio-dynamic corrosion tests with various steels in sodium
chloride solution at 25~ The pitting potential with the high-nitrogen stainless steel were
highest with +950 mVIscz) as compared with 317S 16 (+320 mV) and 316S 16 (+140 mV).
Those results were confirmed by Smethurst [9], who investigated this material in the
cold-worked state (Ortron 90). In this study the breakdown potential of the forged "highnitrogen stainless steel" is slightly lower than in the initial soft-annealed state. Incoming
inspections of the raw material are absolutely essential to maintain the high quality of the
forged components. The breakdown-potential of the investigated products is comparable
with that of the CoCrMo alloys.
CONCLUSIONS
Hip prostheses made from the material "high-nitrogen stainless steel," which are
manufactured according to a forging process, have high mechanical properties in
combination with high ductile characteristics.
The more this highly alloyed steel is formed, the higher its mechanical properties
and hardness. The structure is badly deformed in the vicinity of the flash. There are no
segregations on the grain boundaries.
The corrosion-resistance of this steel is good. Nevertheless, only bar material with
high breakdown potential should be used as raw material for forged products. The
corrosion resistance declines as a result of the forging process. As a result, the forging
process should be optimzed with a view to combining the mechanical properties with
high corrosion resistance.
High-nitrogen stainless steel has also been employed for cemented hip prostheses
in Europe for more than 15 years. Up to now, there have been no reports of any negative
results such as fractured stems or corrosive surface changes, which could be associated
with this material.
References
[ 1]
[2]
Frank, E. and Zitter, H., Metallische Implantate in der Knochenchirurgie, Springer

Verlag, 1971.
Cook, S.D., Renz, E.A., Barrack, R.L., Thomas, K.A., Harding, A.F., Haddad,
R.J., Jr., and Milicic, M., "Clinical and Metallurgical Analysis of Retrieved
WlNDLER AND STEGER ON FORGED HIP STEMS
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
49
Internal Fixation Device," Clinical Orthopaedics and Related Research, No. 194,
1985, pp. 236 - 247.
Disegi, J.A. mad Eschbach L., "Stainless Steel in Bone Surgery," Injury, Vol. 31,
2000, Supplement 4, pp. D2-D6.
Wroblewski0B.M., Revision Surgery m Total Hip Arthroplasty, Springer Verlag,
1990.
Sochart, D.H, Porter, M.L., and Lancashire, W., "The Long-Term Results of
Charnley Low-Friction Arthroplasty in Young Patients Who Have Congenital
Dislocation, Degenerative Osteoarthritis or Rheumatoid Arthritis," Journal of
Bone and Joint Surgery, Vol. 79-A, 1997, pp. 1599 - 1617.
Callaghan J.J., Forest E.E., Olejniczak, J.P., Goetz, D.D., and Johnston, R.C.,
"Charnley Total Hip Arthroplasty in Patients Less Than Fifty Years Old," Journal
of Bone and Joint Surgery, Vol. 80-A, 1988, pp. 704- 714.
Wright, T.M., Burstein, A.H., and Bartel, D.L., "Retrieval Analysis of Total Joint
Replacement Components: A Six-Year Experience," Corrosion and Degradation
of Implant Materials, ASTM STP 859, A.C. Fraker and C.D. Griffin, Eds.,
American Society for Testing Materials, Philadelphia, 1985, pp. 415 - 428.
Truman, J.E,, "An Austenitic Stainless Steel of Imporved Strength and Corrosion
Resistance," Stainless Steel Industry, Vol. 6, No, 23, 1978, pp. 21 - 23.
Smethurst, E., "A New Stainless Steel Alloy for Surgical Implants Compared to
316S12," Biomaterials, 1981, Vol. 2, pp. 116- 119.
Brehme, J., and Biehl, V., "Metallic Biomaterials," Handbook of Biomaterial
Properties, J. Black and G. Hastings, Eds., Chapman & Hall, 1998, pp. 135 - 214.
Grover, H.J., "Metal Fatigue in Some Orthopedic Implants," Journal of Materials,
Vol. 1, No.2, 1966, pp. 413 - 420.
John A. Disagi ~and Lyle D. Zardiackas 2

Metallurgical and Mechanical Evaluation of 316L Stainless Steel Orthopaedic Cable
Reference: Disegi, J. A. and Zardiackas, L. D., "Metallurgical and Mechanical
Evaluation of 316L Stainless Steel Orthopaedic Cable," Stainless Steels for Medical
and SurgicalApplications, ASTM STP 1438, G. L. Winters and M. J. Nutt, Eds., ASTM
Abstract: Stainless steel cable systems have been designed for a multitude of orthopaedic
implant applications. Various implant fixation, eeretage, and tension banding techniques
may be used with cable constructs to stabilize a variety of simple and complex bone
fractures. Multifilament cable products, however, can pose special challenges regarding
metallurgical evaluation and mechanical property testing. The present project investigated
the metallurgical and mechanical properties of a new 316L stainless steel cable system
designed for general trauma surgery. Important cable attributes such as strand
configuration, cable construction, and finishing options are discussed. The advantages of
inductively coupled plasma spectroscopy analysis are presented for the compositional
verification of solid cable specimens. Metallographic methods for the examination of
nonmetallic inclusion limits and grain size are described for a cable construct composed of
75 fine wires. The microstructure associated with laser remelted cable ends is documented
via light microscopy. Unique tension test methods are described for the determination of
failure load and elongation of 1.1 and 1.7 mm diameter cable sizes. Test results are
compared with generalized requirements compiled in ASTM F 2180 Standard
Specification for Metallic Implantable Strands and Cables.
Keywords: metal (for surgical implants), wire, strand, cable, stainless steel, surgical
implants
Stainless steel cable systems are designed for a multitude of orthopaedic implant
applications. Typical clinical indications for cable [1 ] include fracture fixation
stabilization of olecranon, patella, femur, humerus, pelvic, acetabular, and ankle fractures.
Additional clinical uses include acromioclavicular dislocations, prophylactic banding
during total joint procedures, and temporary reduction during open reduction procedures.
1Materials Development Manager, Synthes (USA) Technical Center, 1301 Goshen
Parkway, West Chester, PA, 19380.
2Professor and Coordinator of Biomaterials and Professor of Orthopaedic Surgery,
University of Mississippi Medical Center, 2500 North State Street, Jackson, MS,
39216.
50
Copyright*2003by ASTM International
www.astm.org
DISEGI AND ZARDIACKAS ON 316L STAINLESS STEEL ORTHOPEDIC CABLE
51
The implants and instruments utilized with cable systems for orthopaedic trauma represent
sophisticated designs since the clinical techniques are very specialized. Orthopaedic cable
typically displays improved tensile failure load [2], enhanced fatigue endurance limit [3],
and better handling characteristics when compared to monofilament wire of the same
diameter. Intraoperative advantages for the surgeon include the absence of sheared
monofilament wire ends that can cut through surgical gloves and cause potential disease
transmission problems during surgery. Cable products, however, can pose special
challenges regarding metallurgical evaluation and mechanical property testing. The present
project investigated the metallurgical and mechanical properties of a new 316L stainless
steel cable system designed for general trauma surgery.
Materials and Methods
Cable Construction
The starting coil material used for the manufacture of the cable product examined in
this study met the prevailing requirements specified in ASTM Specification for Wrought
18 Chromium-14 Nickel-2.5 Molybdenum Stainless Steel Bar and Wire for Surgical
Implants (F 138).
Proper terminology related to stranding and cabling configurations is essential in order
to discuss construction and design requirements for orthopaedic cable. Round wire is the
individual product form that is utilized to produce a strand, Strand is a group of wires
helically twisted together while cable is a group of strands helically twisted together. The
construction designation for strands and cables is known as MxN, where M represents the
number of strands in the cable and N represents the number of wires in each strand. Some
examples of strand construction are Ix7 and lx3 while some common cable constructions
are 7x7 and 7x19, The lay or twist is the helical form taken by the wires and by the strands
in a cable. The lay may be right or le~ depending whether the orientation is in the same or
opposite direction as the thread on a right-hand screw. Pitch is the distance parallel to the
axis of the strand or cable in which the wire or strand makes one complete turn about the
axis. A lx7 strand is comprised of 1 wire surrounded by 6 wires with a left-hand lay and a
uniform pitch of 8 to 16 times the nominal strand diameter. A lx19 strand is characterized
by an inner lx7 strand with a fight-hand lay and a uniform pitch of 8 to 12 times the
nominal strand diameter. The outer 12 wires have a lea-hand lay with a uniform pitch of 9
to 11 times the nominal strand diameter. Other types of specialized strand configurations
may be used depending on design criteria and intended applications. Cabling
configurations refer to the manner in which the strands are constructed to achieve a final
cable diameter. Some common cable configurations include 7x7 and 7x19 construction.
The 7x7 cable consists of an inner lx7 strand with a right-hand lay and a uniform pitch of
12 to16 times the nominal strand diameter. The outer lx7 strands have a let's-hand lay with
a uniform pitch of 8 to 16 times the nominal strand diameter. Overall, the 7x7 cable has a
right-hand lay with a uniform pitch of 7 to10 times the nominal cable diameter. The strand,
pitch, and lay configuration for a 7x19 cable is fairly complex and beyond the scope of this
presentation. Various other cable constructions may be used, including hybrid designs that
represent non-standard configurations. The present study evaluated a cable comprised of a
uniquely designed lx19 inner core surrounded by eight lx7 outer strands. The fine wire
52
size was varied to provide either a 1.1 mm or 1.7 mm cable diameter. A cross-sectional
illustration of the cable containing 75 wires is shown in Figure 1.
Figure 1 - Cross-sectional schematic of 1. 7 mm cable containing 75 wires (30X).

Cable manufacturing operations include laser or tungsten inert gas (TIG) thermal
treatments to secure the ends of the cable while hollow crimp blocks are used to provide a
means for threading, tensioning, and crimping the cable. Auxiliary implants may include
eerclage positioning pins and cerclage buttons. Cerclage pins maintain proper cable
position when threaded through a bone plate hole while cerclage buttons help to augment
screw purchase, especially for unicortical screw fixation around a prosthesis. An
orthopaedic cable system will also include specialized surgical instruments such as a cable
passer, calibrated tensioner, crimper, and cutter. An example of a stainless steel cable
system with cerclage positioning pins and a cerclage button to stabilize a stainless bone
plate is shown in Figure 2.
Figure 2 - Stainless cable system with bone plate (IX).
53
Metallurgical Evaluation
Laser treated cable ends were cut, cold mounted in epoxy resin, polished with a series
of polishing suspensions to a final metallographic finish using 0.05 micron aluminum oxide
(A1203), and etched in a solution of ferric chloride (FeCI3) saturated hydrochloric acid
(HCt) with a small amount of nitric acid (HNOs). The microstructure in the vicinity of the
remelted cable end was examined by light microscopy. Composition was determined by
cutting a 40 mm long cable length, folding the cable length in two, and inserting the folded
cable length into a special fixture. Five samples from each cable lot were analyzed.
Manganese (Mn), silicon (Si), nickel (Ni), chromium (CO, and molybdenum (Me) content
were analyzed using a Spectre inductively coupled plasma (ICP) spectroscopy with a
spark stand attachment, which allows direct sampling of solids rather than acid dissolution.
Prior to analysis, the ICP equipment was calibrated with primary standards used for the
routine analysis of implant quality 316L stainless steel. Mean and standard deviation
values were calculated for each element analyzed in the product analysis study. A 15 mm
long specimen was cut, cold mounted, polished, and rated for inclusion content in the
unetched condition per ASTM Test Methods for Determining the Inclusion Content of
Steel (E 45). Five samples per lot (n = 5) were rated for nonmetallic inclusion content.
Five 15 mm long samples per lot were obtained for transverse grain size measurements,
etched with FeC13 saturated HCI plus a small amount of liNe3, and rated according to
ASTM Test Methods for Determining Average Grain Size (E 112).
Tension Testing
Either 1.1 mm or 1.7 nun diameter cable test samples about 400 mm in length were
provided for uniaxial tension testing in a vertical plane according to ASTM Test Methods
for Tension Testing of Metallic Materials (E 8). The cable ends were laser welded to
prevent untwisting during the tension test. A 254 mm gage length was tested at a 2.00
mm/sec stroke rate to failure. A plot of load versus deflection was obtained during the
tension test. The 0.2% yield strength and reduction of area were not determined. A
minimum of five samples (n = 5) representing various implant cable lots were tested.
Failure load and % elongation were recorded. The mean and standard deviation values
were calculated for each cable lot. Tension testing was performed with a Model 810 MTS
servohydraulie test machine equipped with load cells calibrated with standards traceable to
the National Institute of Standards and Technology (NIST).
Results and Discussion
Metaltographic examination of the laser processed cable ends indicated that a relatively
uniform dendritic mictrostructure was present. The dendritic microstructure appeared
similar to recast austenitic stainless structures that are produced by fast melting followed
by a high rate of solidification. Voids and thermal oxidation were not observed in the
remelted regions. Inert gas shrouding during the laser remelting process produced a
homogeneous microstructure that was free of significant oxidation. Laser thermal
treatment has the added advantage of providing highly localized remelting with a small
54
heat affected zone (HAZ). A light microscopic image of the laser processed 316L stainless
steel microstructure is shown in Figure 3.
Figure 3 - Remelted structure of 1. 7 ram diameter laser treated cable end (I OOX).
ICP compositional results were straightforward and no unusual problems were
experienced. The check analysis values for Mn, Si, Ni, Cr and Mo were within the
product analysis tolerance limits specified in F 138. Product analysis tolerances cover
variations between laboratories in the measurement o f chemical content. ICP results for
three cable lots are compared to the certified heat analysis of the starting coil stock in
Table 1.
Table 1 - I C P Product Analysis Results for 1.1 mm Cable
Element
Lot A
Heat
Product
(n = 2)
(n = 5)
Weight %
Lot B
Heat
Product
(n = 2)
(n = 5)
Lot C
Heat
Product
(n = 2)
(n = 5)
Mn
Si
Ni
Cr
Mo
1.77
0.50
14.76
17.46
2.82
1.77
0.46
14.73
17.50
2.80
1.76
0.46
14.72
17.44
2.79
1.85 + 0.03
0.44 + 0.02
14.54+0.30
18.32 + 0.70
2.52 _+0.07
1.84 + 0.06
0.35 + 0.03
14.33 +0.12
17.28 + 0.12
2.61 + 0.06
1.88 _+0.00
0.39 _+0.01
14.71 _+0.16
17.60 _+0.19
2.61 -+ 0.07
DISEGI AND ZARDIACKASON 316L STAINLESS STEEL ORTHOPEDIC CABLE
55
The alloy suppliers usually report the heat composition values in duplicate on the primary
certificate of tests so standard deviations were not calculated for the mean values.
Product analysis test results (n = 5) for Mn, Si, Ni, Cr, and Me indicated that excellent
standard deviations were obtained. The use of the ICP spark system is considered an
advantage because analysis time is significantly reduced and errors that may be introduced
during classical wet chemical analysis are avoided. Standardization with solid primary
standards also provides better control of interelement effects when compared to the use of
individual elemental standards. Also, improved detection limits are generally observed
with ICP when compared to X-ray fluorescence (XRF) procedures.
Transverse metallographic mounts were used to examine the inclusion content of the
stranded fine wires. During the polishing operation about 20-25% of the fine wires became
repositioned in the longitudinal plane and this orientation was used to evaluate the
nonmetallic inclusion content. Nonmetallic inclusion ratings in the unetched condition
were within the limits specified in F 138. Only thin and heavy Type D oxides were
observed when a series of eighteen lots of 1.1 nun diameter and sixteen lots of 1.7 nun
diameter cable were rated for nonmetallic inclusions. The repetitive fine wire die drawing
operations tend to deform and disperse Type A, Type B, and Type C inclusions whereas
the Type D oxides tend to resist deformation during manufacturing,
Grain size in the transverse plane was difficult to rate at 100X magnification due to the
combination of cold work and small wire diameter. The nominal wire diameter was either
0.10 nun or 0.15 mm depending on the cable size that was examined. E 112, Section 10,
Comparison Procedure, defines the maximum value (finest grain size) as ASTM 8 for a
twinned austenitic structure and ASTM 10 for an untwinned austenitic structure. Grain
size was rated finer than ASTM 10 for the majority of the cable lots that were examined,
The stainless cable met the F 138 grain size requirement ofASTM 5 or finer that is
specified for bar and wire product. However, the measurement of grain size was not
straightforward and testing limitations were identified for orthopaedic cable composed of
small diameter wire.
Preliminary tension test results (n = 5) indicated that failure load and elongation values
were not reproducible and unacceptable standard deviations were obtained for both the
1.1 mm and 1.7 mm diameter cable sizes. The early tests were performed with 400 nun
long cable specimens and conventional wire tension grips. Examination of the load versus
deflection curves indicated that cable slippage was occurring in the wire grips, Initial
discussions with a cable producer, Pioneer Cable Technology, Marquette, MI, suggested
that longer test specimens and specialized cable grips with roller assemblies would be
required to provide reproducible tension testing results. Special cable grips were
constructed and 1450 mm long tension test specimens with laser welded ends were
utilized in order to establish a standardized tension test procedure. A long cable specimen
was needed so that both ends of the cable could be threaded around the 25 mm diameter
roller and locking assemblies that are attached to each cable grip. The modified tension
test arrangement eliminated slippage problems and provided excellent data reproducibility.
It was also noted that careful specimen alignment was critical to ensure uniform cable
elongation within the gage section. Elongation was determined by monitoring the amount
ofcrosshead separation during tension testing to failure (not peak load) because it was not
possible to measure accurately the cable length change after failure. Figure 4 shows a
cable specimen inserted into the specially fabricated tension test grips.
56
Figure 4 - Roller grip assembly usedfor cable tension testing (2X).

Mean failure loads, mean % elongation, and standard deviations for random production
lots of 1.1 mm and 1.7 mm diameter cables are compiled in Table 2.
Table 2 - Cable Failure Load and Elongation Results (n = 5)
Cable Diameter
(mm)
1.1
1.7
Lot
Mean Failure
Load, N
A
B
C
D
E
F
G
H
1195+_2.9
1215+12.1
1212 + 5.6
1193 + 1.2
2774 + 13.5
2751 + 18.4
2770+33.4
2801 + 7.4
Mean
Elongation, %
3.9+-0.13
4.2+0.08
4.2 +0.14
4.4+0.13
5.1 + 0.18
5.1 +0.16
4.2+0.17
4.6 + 0.09
The magnitude of the cable failure load and elongation are highly dependent on whether
the fine wire filaments are manufactured in the strand annealed or cold drawn condition.
Strand annealing parameters and the amount of cold work introduced into the fine wire
during the last wire drawing operation will have a significant effect on the finish
mechanical properties attained for cable. Engineering specifications for the implant 316L
stainless cable sizes under investigation define minimum failure loads that must be met on
a consistent basis. Minimum ultimate tensile strength (UTS) is not specified because of
57
errors that may be introduced in calculating the true cross-sectional area of the cable. The
option of using swaging operations to size the cable during manufacture and the tightness
of the strand are factors that can influence the accuracy of the cross-sectional area
measurements. Theoretical formulas for various strand configurations are included in the F
2180 standard but cable designs may not always meet idealized cross-sectional area
calculations. The desire to standardize UTS limits for cable must be balanced against the
improved accuracy that is obtained when area measurement errors are eliminated.
Conclusions
A metallurgical and mechanical property evaluation of 1.1 and 1.7 mm diameter 316L
stainless steel cable intended for surgical implants revealed the following
1. Inert gas shrouded laser remelting is a satisfactory method for captivating the ends of
multifilament cable and this high temperature process provided a homogeneous
dendritic structure that was free of thermal oxidation.
. Standard metallographic techniques are satisfactory for the examination of nonmetallic
inclusions although only type D globular oxides were identified. A grain size rating
finer than ASTM 10 was generally observed but the combination of cold work and
small wire diameter contributed to grain size measurement difficulties.
3. ICP spectroscopy analysis yielded reproducible Mn, Si, Cr, Ni, and Mo
compositional values that were within ASTM F 138 product analysis tolerance limits.
4. Special roller grip assemblies, extra long cable specimens, and careful alignment
techniques are required to provide reproducible tension test results.
References
[1]
[2]
[3]
Synthes (USA), "The Orthopaedic Cable System Technique Guide," Brochure

GP1369-A, 1999, pp. 2-20.
Dall, D.M., "Reconstruction of the Femur in Total Hip Replacement Using
Multifilament Cerclage Cables Fastened with Crimp Sleeves," In: Bone Implant
Grafting, J. Older, Ed., Springer-Verlag, New York, 1992, pp. 135-155.
Lie, A., O'Connor, D.O., and Harris, W.H., "Comparison of Cerclage Techniques
Using a Hose Clamp Versus Monofilament Cerclage Wire or Cable," Journal of
Arthoplasty, Vol. No. 12, 7, 1997, pp. 772-776.
Processing, Quality Assurance Concerns,

and MRI Testing
J. Zachary Dennis, l Charles H. Craig, 1Herbert R. Radisch, Jr., 1 Edward J. Pannek, Jr., 1
Paul C. Turner, 2 Albert G. Hicks, 3 Matthew Jenusaitis, l Nev A. Gokcen, 4 Clifford M.
Friend, s and Michael R. Edwards 5
Processing Platinum Enhanced Radiopaque Stainless Steel (PERSS ~ ) for Use as
Balloon-Expandable C o r o n a r y Stents
Reference: Dennis, J. Z., Craig, C. H., Radisch, H. R., Jr., Pannek, E. J., Jr., Turner, P.
C., Hicks, A. G., Jenusaitis, M., Gokcen, N. A., Friend, C. M., and Edwards, M. R.,
"Processing Platinum Enhanced Radiopaque Stainless Steel (PERSS e) for Use as
Balloon-Expandable Coronary Stents," Stainless Steels for Medical and Surgical
Applications, ASTM STP 1438, G. L. Winters and M. J. Nutt, Eds., ASTM International,
West Conshohocken, PA, 2003.
Abstract: A platinum-enhanced variant of UNS $31673 stainless steel has been
developed for use in fabricating balloon-expandable coronary stents. PERSS | is
sufficiently radiopaque to be detected by current radiographic methods and maintains
compliance with ASTM requirements. Boston Scientific Corporation/Interventional
Technologies (BSC/IVT) has implemented a pilot program for the development of
PERSS with the intent of achieving vertical integration of the manufacturing process.
PERSS ingots containing 5 wt% platinum were cast by vacuum induction melting, refined
by vacuum arc remelting, upset in a forge, then hot and cold-rolled into sheet and foil.
The sheet was cold-rolled on a 20-high cluster mill to a foil thickness of less than 0.25
mm (0.01 inches). While the final product otherwise meets the ASTM specification for
UNS $31673 alloys, intentional addition of unspecified additional elements precludes its
compliance with same.
Keywords: coronary stent, stainless steel, platinum, vacuum induction melting, vacuum
arc remelting, Sendzimir mill, PERSS
i ManufacturingEngineer, Research & DevelopmentEngineer, Vice President (R&D), Vice President

(Process Development), and General Manager, respectively, Boston Scientific
Corporation/InterventionalTechnologies, 3574 RuffinRoad, San Diego, CA, 92123.
2 Associate Director for Thermal Treatment Technologies,Albany Research Center, U.S. Department of
Energy, 1450 Queen Ave., Albany, OR, 97321.
3 Technical Manager, Rolled Products Division, The Arnold Engineering Co., 300 N. W. Street, Marengo,
IL, 60152.
4 Consultant, 385 Palos Verdes Dr. W, Palos Verdes Estates, CA, 90274,
Cranfield Postgraduate Medical School, Shrivenharrl, Swindon SN6 8LA, United Kingdom.
6 Registered trademark: Boston Scientific Corporation/InterventionalTechnologies, San Diego, CA.
61
www.astm.org
62
Introduction
Balloon-expandable coronary stents are typically made from thin-walled tubes of

surgical grade stainless steel made in accordance with ASTM Specification for Wrought
18 Chromium-14 Nickel-2.5 Molybdenum Stainless Steel Sheet and Strip for Surgical
Implants (UNS $31673), F139. This material has sufficiently low carbon content to
ensure freedom from intergranular corrosion susceptibility, while pitting resistance is
obtained by maintaining a specific weight percentage of chromium and molybdenum.
Stents made from commercially available variants of UNS $31673 stainless steel that is
made to F139 such as BioDur~ 7 316LS are quite suitable in terms of their structural
properties and their biocompatibility, but may not be suitable in terms of fluoroscopic
radiopacity, a functional property that determines whether the stents are visible
radiographically once they have been deployed inside the coronary artery. The PERSS
alloy was developed to provide a solution to this problem by an unspecified additional
element, platinum, which addresses the functional property without compromising the
structural or biocompatibility properties. Development of the PERSS alloy is described
by Craig et al [ 1 ].
Processing of the PERSS alloy is strongly influenced by design-mandated concerns
over dimensional control of the final thickness of the foil and over maintaining its final
grain size. BSC/IVT uses welded tubes made from this alloy to fabricate stents, which
are made by rolling foil into a tube, laser-welding the seam, then drawing it to the
required diameter of the stent. BSC//VT fabricates stents by a chemical etching process,
which requires tubes of an extremely consistent wall thickness and grain size in order to
produce implant grade medical products. Grain size significantly affects the strength and
ductility of the stent and is controlled for cold-worked and annealed structures principally
by the amount of cold working [2]. The effect of the annealing temperature is much less,
especially when the amount of prior cold work is large, Thus, in order to produce a
specific grain size after annealing, as is required to produce the appropriate mechanical
properties for stents, careful control of the cold working before annealing is essential.
These material conditions must be met before the tube is chemically etched.
Based on constraints of thickness and grain size, processing requirements for
manufacturing the foil were determined and are addressed here. Figure 1 shows the
processing steps for PERSS alloys prior to tube production and stent fabrication. The
alloy is formed by Vacuum Induction Melting (VIM) a commercially available stainless
steel, BioDur 316LS, in rod form, along with the chosen unspecified additional element,
platinum, and any additional specified elements such as chromium and molybdenum
required to maintain PERSS within the compositional specifications ofF139. The alloy
is refined through Vacuum Arc Remelting (VAR) and molded into an ingot. The ingot is
taken through a forging process where it is formed into a billet. The billet is formed into
a sheet by hot-rolling in a 2-high rolling mill and cold-rolling in a 4-high rolling mill.
The foil is then formed from the sheet by a 40% final reduction in thickness by a 20-high
Sendzimir cluster rolling mill (Z-mill).
7Registered trademarkCarpenterTechnologyCorporation,Reading, PA.
DENNIS ET AL. ON PERSS AS CORONARY STENTS
Casting
Casting
Forging
H .o, H co,H
Rolling
Rolling
63
Z-MiU
Figure 1 - Block diagram of the stent material production process prior to tubeforming
Background
VIM of the stainless steel was chosen in order to reduce the level of nonmetallic
inclusions in the alloy, which can tend to promote unwanted corrosion. VIM is a
metallurgical process that uses an induction furnace inside a vacuum chamber to melt and
cast steel (as well as other alloys). VIM consists of heating the alloy components together
in a crucible that is surrounded by a water-cooled copper coil. High frequency current
passes through the coil and melts the materials within the crucible as well as causing a
powerful electromagnetic stirring action. The use of a vacuum helps to minimize the
amount of impurities present in the alloy by keeping oxides and other nonmetallic
products from forming.
A second melting of the VIM-produced alloy may be accomplished by VAR in order
to reduce the chemical segregation in the ingot, as well as producing further reductions in
the level of nonmetallic inclusions. VAR consists of maintaining a high current DC arc
between the electrode made from the VIM-produced PERSS alloy and a molten metal
pool of the alloy that is contained in a water-cooled copper crucible. The VAR process,
as with the VIM process, is kept under vacuum to maintain alloy cleanliness and
eliminate impurities. The remelting process produces an ingot with good internal
structure and excellent chemical homogeneity.
The reduction of the ingot to plate was carried out by initially hot upsetting the ingot
and then hot rolling. Forging the molded ingot into a billet is performed by compressing
the ingot between two fiat dies, or "upsetting". The forging process changes the
microstructure of the workpiece from a cast to a wrought structure, i.e., from a chemically
homogenous ingot with non-uniform grains to a wrought billet with uniform grains. Hot
rolling is performed above the recrystallization temperature of the alloy. A billet is
heated and drawn through a pair of hardened steel rollers that reduces the thickness of the
material over several passes to produce a plate form. The grains initially elongate and
subsequently recrystallize into smaller, more uniform grains, providing greater strength
and ductility than is provided by the metallurgical structure of the billet.
For the reduction of the plate to thin sheet, it was necessary to cold roll the alloy at
room temperature, because this gave better dimensional tolerances and a superior surface
finish. Cold rolling is performed to reduce plate thickness without allowing the grains to
reerystallize.
The final cold rolling of the PERSS sheet into foil requires a 40% reduction in
thickness with the foil thickness required to be very consistent both along the width and
the length of the foil. Normal rolling mills are affected by roll deflection, defined as a
tendency for the rolls to bend outward in response to the roll forces. This causes a crown
to be formed on the rolled material; the center is observed to be thicker than the outer
64
edges. This effect can be countered by using a larger roll and giving it a barrel shape
(camber) to offset the effects of roll deflection. Larger rollS, however, are more
susceptible to roll flattening, where the rolls bulge into an oblong shape in response to the
roll forces. Roll flattening can cause defects in the final material and limits the amount
that the material can be reduced.
The Z-mill is of a class of rolling mills known as "cluster" mills (see Figure 2). Two
small-diameter rolls that contact the metal are supported by a group of larger rolls. The
smaller diameter rolls enable the mill to perform the 40% reduction of the material
without suffering the effects of roll flattening. The smaller diameter rolls also reduce the
roll force and power requirements, and help prevent horizontal spreading of the material.
The larger supporting rolls prevent the working rolls from deflecting, so a consistent foil
thickness can be maintained. These rolls are particularly useful in rolling thin sections
and materials that are hard. Since stainless steels have high work-hardening coefficients,
these mills have often been applied in the rolling of fine gages of stainless steels. Further
details of Sendzimir mills may be found in [3].
Figure 2 - 20-high Z-Mill roll configuration

Fo~ Production
The initial processing of the alloy was performed at the U. S. Department of Energy's
Albany Research Center (ARC) located in Albany, Oregon. It consisted of the melting,
forging and rolling of the alloy to sheet approximately 0.060 in thick. The alloy was then
rolled to product size, 0.006 in thick, by The Arnold Engineering Company, located in
Marango, IL. BioDur 316LS stainless steel rod, chromium, molybdenum, and platinum
were melted together in a VIM furnace. The VIM system used in the development
65
program is shown in Figure 3, and the ingot produced from the VIM process is shown in
Figure 4. The ingot has approximate dimensions o f 15 cm diameter by 20 cm long. The
composition o f the platinum enhanced stainless steel ingot was determined and is
presented in comparison to the typical composition of BioDur 316LS in Table 1 below.
Table 1 - Compositions of UNS $31673, BioDur 316LS, and PERSS alloys
Element
Symbol
ASTM F139
Carbon
0.030 wt% max.
Mn
Si
Cr
Ni
Mo
2.0 wt% max.

0.75 wt% max.
17.0 - 19.0 wt%
13.0 - 15.0 wt%
2.25 - 3.0 wt%
Manganese
Silicon
Chromium
Nickel
Molybdenum
Platinum
Pt
BioDur 316 LS
0.024 wt%
1.80
0.44
17.66
14.66
2.78
wt%
wt%
wt%
wt%
wt%
PERSS
ingot #50
0.023
wt%
1.54 wt%
0.45 wt%
18.67 wt%
13.25 wt%
2.94 wt%
5.32 wt%
Figure 3 - Vacuum induction meltingfurnace at Albany Research Centers
s Figure 3 - 8 courtesy Albany Research Center.
66
Figure 4 - Vacuum induction melt PERSS ingot

To further refine the material and improve its quality the VIM ingot was subjected to
the VAR process. The ingot was quartered and welded together to form a consumable
electrode and then secured in an evacuated chamber. The amount of current passing
through the bars was gradually increased from 1500 A at 26 V to a maximum of 4800 A
at 32 V. After VAR, the re-solidified ingot had an approximate diameter o f 15 cm and an
approximate length of 20 cm.
To prepare the material for the hot-rolling process, the ingot was forged into a
rectangular block (billet). The ingot was heated to 1230~ for a soak time o f five hours
and transferred to the forge shown in Figure 5. The material was upset through a series of
compressions, reheating the material between actions of the forge to produce a billet
approximately 9.5 cm x 17 cm x 22 cm.
The process of hot and cold- rolling the billet into plate form in a 2-high rolling mill
took place in several stages, with a typical reduction o f 10% per pass. The billet was
rolled into a slab at an initial temperature o f 1230~ and reheated between the subsequent
passes to maintain the elevated temperature. The slab was rolled into a plate with a final
thickness of 1.33 cm (0.522") and was of sufficient consistency that it was not necessary
to re-flatten the material on the forge. The 2-high rolling mill used in the development
program is shown in Figure 6, while Figure 7 shows the hot plate emerging from the mill.
Figure 8 shows the sheet of material obtained after cold-rolling on the 2-high rolling
67
mill. The material was annealed at 1040~ for 14 minutes before fan-assisted cooling to
room temperature. The plate was transferred to a 4-high rolling mill and cold-rolled by
an extensive series of 5% reductions with occasional fifteen-minute anneals at 1040 ~
The 4-high rolling mill is shown in Figure 9. The sheet that was obtained through the
first part of the cold-rolling process had a thickness o f 1.63 mm (0.064"). The cold-rolled
sheet was coiled and secured for a vacuum batch anneal at 950~ The strip was cleaned
and trimmed and the thickness further reduced by cold-rolling to a thickness o f 0.69 mm
(0.027") on the 4-high mill.
Prior to the final reduction in the Z-mill, the strip o f PERSS material was trimmed to a
width o f 15.88 cm (6.25") and strip annealed at 1065~ at approximately 2 m per minute
(6 feet per minute) in a horizontal furnace. The material was then loaded onto the Z-mill
and reduced to a final thickness o f 0.15 mm (0.0063"). A final anneal was performed at
1050~ at approximately 1 m per minute (3 feet per minute) in the horizontal furnace.
Figure 10 shows a Z-mill such as that used for the process development described above.
Figure 5 - Forging press at Albany Research Center
68
Figure 6 -
Figure 7 -
2~4-high rolling mill at Albany Research Center
Hot-rolled PERSS plate emergingfrom 2-high mill at Albany Research Center
Figure 8 -
Cold-rolled PERSS sheet after 2-high rolling at Albany Research Center
Figure 9 -
4-high rolling mill at Arnold Engineering
69
70
Figure 10 - 20-high cluster mill at Boston Scientific/Interventional Technologies
Summary and Conclusions

A platinum-enhanced variant of UNS $31673 stainless steel has been developed by
BSC/1VT. PERSS has an enhanced radiopacity compared to UNS $31673 stainless steels
without additional unspecified elements, which makes it ideal for coronary stent
applications. Procedures described in this paper discuss the process of taking the initial
PERSS ingot and converting it into a form that can be used with BSC/1VT's chemical
etching process to produce stents.
BSC/IVT has developed a pilot program for internal production of its new PERSS
alloy. The Z-mill was shown to be capable of providing the precise reduction in
thickness necessary for the foil to be used in the stent manufacturing process.
Acknowledgments
The authors would like to express their gratitude to the U.S. Department of Energy's
Albany Research (Portland, OR), The Arnold Engineering Company (Marengo, IL), and
Intergrated (sic) Industrial Systems (Yalesville, CT) for assistance in the processing of
this material. Without their contributions, the timely completion of this research would
have been much more difficult.
References
[1]
Craig, C. H., Radisch, H. R., Jr., Trozera, T. A., Turner, P. C., Govier, R. D.,
Vesely, E.J., Jr., Gokcen, N. A., Friend, C. M., and Edwards, M. R.,
"Development of a Platinum Enhanced Radiopaque Stainless Steel
(PERSS|
Stainless Steels for Medical and Surgical Applications, ASTM
STP 1438, G. L. Winters and M. J. Nutt, Eds., ASTM International, West
Conshohocken, PA, 2003.
71
[2]
Humphreys, F. J., and Hatherly, M., Recrystallization and Related Annealing

Phenomena, Pergamon, Oxford, 1996, pp. 205-206.
[3]
Lahoti, G. D., and Semiatin, S. L., in Metals Handbook, 9~hEdition, Volume 14,
Forming and Forging, ASM International, Materials Park, OH, 1988, pp. 343360.
Markus Windler, I Rainer Steger, ~and Gary L. Winters 2
Quality Aspects of High-Nitrogen Stainless Steel for Surgical Implants
Reference: Windier, M., Steger, R., and Winters, G. L., "Quality Aspects of HighNitrogen Stainless Steel for Surgical Implants," Stainless Steelsfor Medical and
SurgicalApplications, ASTMSTP 1438, G. L. Winters and M. J. Nutt, Eds., ASTM
Abstract: High-nitrogen stainless steel has been used in orthopedic implants because of
its high strength, good corrosion resistance and biocompatibility. In this study six
annealed wrought high-nitrogen stainless steel sample bars from five different
manufacturers were characterized in terms of chemical composition, mechanical
properties, hardness, metallurgical structure and corrosion resistance. The metallurgical
investigation, which determined the microstructure, grain size and inclusion content, was
made with a light microscope, image analysis and SEM with EDX. The potentio-dynamic
corrosion behaviour in the cross-sectional plane was measured at pH 4 and 40~ No
major differences were found in the chemical compositions, mechanical properties,
hardness or metallurgies of the six sample materials. The only difference between the
materials was established in the corrosion test, where the breakdown potential varied
between -150 and +900 mV~scE). However, no correlation could be established between
the corrosion results and chemical composition, microstructure, or type / number of the
inclusions. Additional heat treatment improved the material with low breakdown
potential. It is recommended that minimum requirements should be established for the
breakdown potential and incorporated in the current material standards.
Keywords: high nitrogen stainless steel, corrosion properties, mechanical properties,
microstructure, inclusions
1Director, Materials Research, and Research Scientist, respectively, Material Research

Department, Sulzer Orthopedics Ltd., P.O. Box 65, 8404 Winterthur, Switzerland.
2Materials Engineer, Material Research Department, Sulzer Orthopedics Inc., Spectrum
Drive 9900, Austin, TX, 78717, U.S.A.
72
www.astm.org
WINDLER ET AL. ON HIGH-NITROGEN STAINLESS STEEL
73
Nomenclature
Breakdown potential - The least noble potential where pitting or crevice corrosion, or
both, will initiate and propagate [ASTM Standard Terminology Relating to Corrosion
Testing (G15)].
Introduction
Stainless steel has been employed as surgical implant material for many years. The
quality of the steel in the pioneering days was poor and numerous deficiencies such as
fatigue failures and severe corrosion were reported [ 1]. The breakthrough of stainless
steel as an implant material came with the development of the material with the
designation AISI 316L. This type of stainless steel is still used today, primarily for the
products of osteosynthesis like plates, nails and screws [2,3]. Endoprostheses made of this
type of steel since the 1960s, e.g. the Charnley hip stem, have been implanted and
retained successfully in the body for many years [4-6]. Reports of corrosion phenomena
and fatigue failures of this stainless steel in-vivo have been made in the literature [2,4,7].
A new type of stainless steel was introduced in the field of orthopedic surgery in the mid
t 980s [4]. The chromium and manganese contents of the alloy were raised and up to
0.5% nitrogen was also introduced to improve the corrosion resistance and increase the
strength [8]. This material was standardized under the designation "Wrought High
Nitrogen Stainless Steel" 1992: ISO 5832-9 and ASTM F1586 - 95. This highly alloyed
stainless steel is employed primarily for cemented hip prostheses and is well established
in Europe. As a result of the increasing demand for material in the mid-1990s, more and
more steelworks began to manufacture this stainless steel.
This study was made to document the variation in quality of this high alloyed
stainless steel, which is produced by various manufacturers. In addition to a quantitative
microstructural analysis, the corrosion resistance was studied in the potentio-dynamic
test. The biocompatibility of a metal is influenced decisively by the corrosion resistance.
Reports have been made of the negative biological effects of corrosion products by
various metals [ 1,9,10].
Material
Six wrought high-nitrogen stainless steel sample bars (A, B-l, B-2, C, D, and E)
from five different manufacturers ("A" - "E") were compared in this evaluation. The
commercially manufactured bars, which had diameters of 22 to 33 mm and were annealed
(1050~ 1 h, quenched with water), met the requirements of ISO 5832-9 and ASTM
Standard Specification for Wrought Nitrogen Strengthened-21 Chromium- 10Nickel3Manganese -2.5Molybdenum Stainless Steel Bar for Surgical Implants (F1586) for
chemical composition and mechanical properties (Table 1). The values in Table 1 were
taken from the manufacturer's material certificates. This material will be used for femoral
hip stems which underwent afterwards a forging process.
74
Methods
Metallographic sections of all the bars were prepared in the longitudinal and
transverse directions and etched and studied with high powered optical microscope
(Leica DMRX, Wetzlar, Germany). Typical structural pictures with a magnification of
500x were made of all materials in the scanning electronic microscope (JSM-IC 848 A,
Joel, J-Tokio). With the aid of image analysis software (Soft Imaging System, Analysis
3.0, Mtinster, Germany), the inclusions in the steel were determined according to type,
number and size. The chemical elements of the inclusions were determined by means of
EDX Analysis (Energy Dispersive Analysis) (TRACOR Analysator TN 5421, Noran
Instruments Ltd., Middleton, WI, USA). A structural analysis consisted of 5 images,
each with a size of 200/am x 150 Mm.
An X-ray diffraction analysis was made on the transverse section of Material "E" to
identify the inclusions (Goniometer, Cu K~ rays with % = 0.15418 nm).
Table 1 - Chemical Composition and Mechanical Properties of Studied Materials
"A . . . .
B- 1. . . . B-2 . . . . C . . . .
D....
E"
ISO 5832-9
Diameter
Io22mm o33mm e33mm o22mm o22mm 033 mm
n.s.
Carbon [%]
0.034
0.02
0.032
0.043
0.03
0.034
0.08 max.
Chromium [%]
20.9
20.2
21.25
21.27
20.5
20.71
19.5 to 22
Nickel [%]
9.55
10.15
10.55
9.15
9.8
9.82
9 t o 11
Manganese [%]
4.12
3.65
4.05
3.51
3.99
3.81
2 to 4.25
Molybdenum [%] 2.15
2.4
2.28
2.36
2.1
2.17
2to 3
Nitrogen [%]
0.39
0.55
0.43
0.41
0.37
0.39
0.25 to 0.5
~Niobium [%]
0.31
0.38
0.27
0.36
n.m.
0.27
0.25 to 0.8
Phosphorus [%]
0.021
0.023
0.025
0.011
0.018
0.016 0.025 max.
Sulphur [%]
<0.002 0.0001
0.001
0.001
0.001
0.001
0.01 max.
Copper [%]
0.013
0.08
0.23
0.06
0.1
0.06
0.25 max.
Silicon [%]
0.34
0.73
0.19
0.14
0.25
0.15
075 max.
Iron [%]
Balance Balance Balance Balance Balance Balance
Balance
Properties
TS [MPa]
880
1017
902
900
869
868
740 rain.
YS (0.2%) [MPa]
506
600
565
539
621
482
430 rain.
Elongation [%]
39
35
41
36
35
46
35 rain.
Brinell Hardness
241
276
252
254
251
229
n.s.
Grain Size No.
10
8
7-8
10
5 - 10
8
>5
NOTES: TS = tensile strength, YS = yield strength, n.s. = not specified, n.m. = not
measured by the vender. The Brinell Hardness values were measured in our lab and were
not from the vendors' certificates.
75
Hardness Measurement
The Brinell hardness survey was made in the center of each metallographic
transverse section. The Brinell hardness (2.5/187.5) was determined according to ISO
Metallic materials - Brinell hardness test - Part 1: Test method (ISO 6506-1), using an
EMCO Tester, M4A-025, Hallein, Austria. Each survey consisted of three measurements.

A potentio-dynamic polarization curve of each sample were determined per ASTM
Standard Reference Test Method for Making Potentiostatic and Potentiodynamic Anodic
Polarization Measurements (G5) using an EG&G (VersaSat II) Potentiostat, EG&G
(SoftCorr III) Software and an EG&G K46 corrosion cell (EG&G Instruments, Princeton,
USA). The samples were prepared so that the measurement surface of each sample were
transverse planes of the original bar. The sample surfaces were ground to 1200 grit. A
surface area of 1 cm 2 was immersed in 1 litre of sodium chloride solution (0.9%) at 40~
pH 4 (adjusted with hydrochlorid acid) and with nitrogen bubbling during the whole
process. A cathodic pre-polarization (2 minutes at 100 ~tA/cm 2) was made 250 mV
negative to the rest potential to clean the immersed surface. After waiting 30 minutes to
allow the system to come to equilibrium, the test was started with a scan rate of 10
mV/min. Saturated calomel electrodes (SCE) were used for reference purposes. All the
tests were made twice reusing the old sample after refinishing.
Corrosion Test after Heat Treatment

Four specimens each with a length o f 20 mm were separated from the material
produced by manufacturer "E" (022 mm). These specimen were subjected to different
heat treatments at temperatures of 850~ 900~ 950~ 1000~ 1050~ and I 100~ in
a High-Therm VMK 1400 air fumace (Linn, Hirschbach, Germany). After a holding time
of 1h, the specimens were quenched in air. Electrochemical tests were performed on the
transverse section of each sample. Two tests were performed for each heat treatment
reusing the old sample after refinishing. Transversal cross sections were made after heat
treatment to check grain growth, as well as qualification of inclusion content by light
microscope.
Results
All metallographic sections of the high-nitrogen stainless steels had an austentic
matrix with a grain size of between 7 and 10. No delta ferrite could be observed with a
magnification of up to 100x. The structure of the transverse sections was i sotropic and the
structure of the longitudinal sections was elongated in the rolling direction. Three forms
76
of inclusions were observed in the longitudinal and transverse cross sections. Type 1
inclusions were fine and homogeneously-distributed with a diameter of less than 1 p.m.
They were identified as [Cr, Mo]-carbides after qualitative EDX analysis detected the
presence of the elements Cr and Mo in these features. Type 2 inclusions were larger,
more elongated in the rolling direction, had a diameter o f up to 10 ~m and consisted o f
Cr, Nb and N (Figure l). On the basis of the X-ray diffraction analysis, these inclusions
were tetragonal [Cr, Nb]-nitrides; Card Number No. 25-0240 according to JCPDS [ 11 ].
Type 3 inclusions were black, non-metallics up to 15 p,m in diameter and had a
composition of aluminum (Figure 2). These inclusions were alumina-slag and resulted
from the deoxidation process. The number of Type 2 and Type 3 inclusions are shown in
Figures 3 and 4.
Figure 1 - Microstructure, Material "B-2"
Figure 2 - Microstructure, Material "B-2"
in the longitudinal cross-section. Type 2

inclusion consisted of [Cr, Nb]-nitride
in the transverse cros-section. Type 3

inclusion consisted of alumina-slag

Typical potentio-dynamic plots are shown in Figure 5. The rest potential (or open
circuit potential) varied between -700 mV and -400 mV for all the tests, and a corrosion
rate of about 1 ~tA/cm 2 was observed in the passive range. The breakdown potential for
the different bars was noticeably different (Table 2). The breakdown potential varied over
a range o f - 1 5 0 m V and +900inV. The plot of material "C", had severe transients above
WlNDLER ET AL. ON HIGH-NITROGEN STAINLESS STEEL
77
-100 mV and the breakdown potential could only be determined in these cases with
difficulty.
300
O
250
"~ 200
~
Z
15o
100
50
0
oijl;,,
6oil ,tlltitl:l
8O
70 -F -
[] Lon~udml
9 Transversal [[~_~
~,
[] Longitudinal
...... .Transversal
B1
B2 C
D
Manufacturer
40
30
20
10
0
Figure 3 - Number o f Type 2 inclusions

[Cr, Nb]-nitride in the longitudinal and
transverse cross sections
B1
B2 C
D
Manufacturer
Figure 4 - Number o f Type 3 inclusions

(slag) in the longitudinal and transverse
cross sections
1.E-02
1 .E-03
<
O
"r
1 .E-04
Conditions
9 0.9%NaC1
//
9 pH 4, 40~
1. scanrate 10mV/min
. / q Material "E" [
Aft")/
'
'
1.E-05
1.E-06
I ate "A"l
1 .E-07
1 .E-08
-800
-600
-400
-200
200
400
600
800
1000
Potential [mV(scE)]
Figure 5 - Potentio-dynamic corrosion test o f high-nitrogen stainless steel. Breakdown
potential f o r Material "A " = +910 m V(scE) and Material "E " = +200 m VcscE)
78
Corrosion Test after Heat Treatment

The breakdown potentials of the heat-treated specimens are listed in Table 2. The
heat treatments up to 900~ did not have any influence on the breakdown potential. An
improvement could only be realized from 950~ upwards, but with numerous transients
in the passive range. The potentio-dynamic corrosions curves had no transients after heat
treatments of > 1000~ with breakdown potentials in the range of +850 to +920 mV(scE)
(Figure 6). A small grain growth was recognized after the heat treatment >1050~ and
about 50% of the type 2 inclusions (Cr,Nb-nitrides) had been dissolved into the matrix.
Table 2 - Breakdown Potential for Different Qualities of High-Nitrogen Stainless Steel
Material
"A"
"B-I"
"B-2"
"C"
"D"
"E"
"E"
850~ / 900~
"E"
950~
"E"
1000~ - 1100~
Breakdown Potential
[mV(scE)]
+910 / +800
+520/+380
+890/+820
approx. +640 / approx. +810
+230 / -150
+160 / +200
+180 to +220
Remarks
pitting corrosion
crevice corrosion
pitting corrosion
transients above -100 mV
crevice corrosion
crevice corrosion
approx. +600
crevice corrosion
transients above +100 mV
+850 to +920
pitting corrosion
Discussion
With high-nitrogen stainless steel, the substantial difference in the breakdown

potential of various material lots and manufacturers cannot be explained on the basis of
the microstmctural analysis. The material is not easy to manufacture - a statement, which
is substantiated by the following facts. The material lots "C" and "D" were the first times
that the respective manufactures had attempted to make this material. In all probability,
the material qualities, i.e., corrosion resistance, fluctuated markedly because the
fabrication operations were not optimal. In view of the quality problems, Manufacturer
"E" does not market this material any more.
Inclusions in the stainless steel can have a marked effect on the corrosion
resistance. In our opinion, the larger [Cr, Nb]-nitrides have a greater influence on the
corrosion resistance than the very small chromium-molybdenum carbides. As a result of
the additional heat treatment > 1000~ the [Cr, Nb, N]-inclusions may have been
dissolved into the matrix, with a positive effect on the corrosion resistance. In general,
inhomogeneities in the material can be reduced by means o f heat treatment.
WlNDLER ET AL. ON HIGH-NITROGENSTAINLESS STEEL
79
1.E-02
IConditio_._____~
1 .E-03
1.E-04
9 0.9% NaC1
9 pH4, 40~
9 scan rate 10mV/rr~a
[ '~E"900~
1 .E-05
1.E-06
'%" 1100~
1 .E-07
1.E-08
-800
-600
.
-400
-200
0
200
400
Potential [mV(scE)]
600
800
1000
Figure 6 - Potentio-dynamic corrosion test of Material "E" after heat treatments at

900~ 950~ and llO0~
The corrosion resistance in the stainless steels could be affected by the presence of
sigma-phase (~1) and / or delta-ferrite (6) [12]. ~-phase can occurs if the material were
held for some time between about 700 and 810~ The crystal structure is tetragonal, hard
and brittle. 8-ferrite can form directly from austenite during the cold working process. If
the rolling/heating operations, which the mill used to balance grain size, mechanical
properties and the requirements to be annealed were not adequate, then stainless steels
may form small amounts of ~-phase and 6-ferrite. Both phases usually cannot be found in
the optical microstructure, but they can be removed by annealling.
With good material lots, the corrosion resistance of high-nitrogen stainless steel is
distinctly higher than those of the conventional stainless steels, type ISO 5832-1 or AISI
316L. The breakdown potential for AISI 316L was measured in the range of +100mV to
+350mV(scE) by various authors depending on pH and temperature [ 13,14]. In the main,
the differences are attributable to the different structural and deformation conditions. In
comparison with the stainless steel alloys, the breakdown potential by CoCrMo alloys
ranges from +400 to +600mV~scE) [15,16]. These values are valid for CoCrMo cast alloy
[ASTM Standard Specification for Cobalt -28Chromium -6Molybdenum Casting Alloy
and Cast Products for Surgical Implants (F75)] as well as for wrought or forged CoCrMo
alloys [ASTM Standard Specification for Wrought Cobalt - 28Chromium -
80
6Molybdenum Alloy for Surgical Implants (F1537) or ASTM Standard Specification for
Cobalt -28Chromiurn -6Molybdenum Alloy Forgings for Surgical Implants (F799)].
The majority of the potentio-dynamic corrosion tests described in the literature were
made in sodium chlorid solutiOns or Ringer's solution at 37~ and pH 7 (near
physiological conditions). Corrosion phenomena, which are mainly attributable to crevice
corrosion, have been found by the analysis of implants [17]. Fontana [18] developed the
basic model for crevice corrosion. According to this theory, metal consumes oxygen in a
cathodic reaction and oxygen cannot be replaced in a crevice because the small
dimensions of the crevice impeds the diffusion of oxygen into the crevice. At the final
stage the pH of the crevice drops to a lower value and the unprotected metal dissolves.
The tests in this study were therefore conducted at pH 4 to simulate the acidic condition
of the crevice.
Conclusions
As far as the chemical composition and the mechanical properties were concerned,
all six bar materials of the high-nitrogen stainless steel specimen fulfilled the
requirements of the standards ISO 5832-9 and ASTM F1586.
A critical difference was found for the corrosion resistance, whereby the best
material lot had a breakdown potential of about +900 mV and, in contrast, the poorest
material lot exhibited a breakdown by a potential of-150 mV.
The breakdown potential can be improved considerably by means of a subsequent
heat treatment at _>1000~ The reason for this, however, was not indicated in this work.
Nevertheless, it is interesting to find that an additional heat treatment had a positive effect
on bars that have already been mill annealed.
From the point of view of corrosion resistance, it is possible to differentiate
between "Good Material Lots" and "Poor Material Lots" with the aid of a simple
potentio-dynamic corrosion test. This corrosion test, together with minimum
requirements, should be incorporated in the ISO and ASTM standards for the material
"High-Nitrogen Stainless Steel" in future.
References
[ 1]
[2]
[3]
[4]
Frank, E., and Zitter, H., Metallische Implantate in der Knochenchirurgie, Springer
Verlag, 1971.
Cook, S.D., Renz, E.A., Barrack, R.L., Thomas, K. A., Harding, A.F., Haddad, R.J.,
Jr., and Milicic, M., "Clinical and Metallurgical Analysis of Retrieved Internal
Fixation Device," Clinical Orthopaedics and Related Research, No. 194, 1985, pp.
236 - 247.
Disegi, J.A., and Eschbach, L., "Stainless Steel in Bone Surgery," Injury, Vol. 31,
2000, Supplement 4, pp. D2-D6.
Wroblewski, B.M., Revision Surgery in Total Hip Arthroplasty, Springer Verlag,
1990.
[5]
[6]
[7]
[8]
[9]
[ 10]
[11]
[ 12]
[13]
[14]
[ 15]
[ 16]
[17]
[18]
81
Sochart, D.H., Porter, M.L., and Lancashire, W., "The Long-Term Results of
Charnley Low-Friction Arthroplasty in Young Patients Who Have Congenital
Dislocation, Degenerative Osteoarthritis or Rheumatoid Arthritis," J. Bone and
Joint Surgery, Vol. 79-A, 1997, pp. 1599 - 1617.
Callaghan, J.J., Forest, E.E., Olejniczak, J.P., Goetz, D.D., and Johnston, R.C.,
"Chamley Total Hip Arthroptasty in Patients Less Than Fifty Years Old," J. Bone
and Joint Surgery, Vol. 80-A, 1988, pp. 704 - 714.
Wright, T.M., Burstein, A.H., and Barrel, D.L., ~Retrieval Analysis of Total Joint
Replacement Components: A Six-Year Experience," Corrosion and Degradation of
Implant Materials, ASTM STP 859, A.C. Fraker and C.D. Griffin, Eds., American
Society for Testing Materials, Philadelphia, 1985, pp. 415 - 428.
Brehme, J., and Biehl, V., "Metallic Biomaterials," Handbook of Biomaterial
Properties, J. Black and G. Hastings, Eds., Chapman & Hall, 1998, pp. 135 - 214.
Merritt, K., and Brown, S.A.; "Biological Effects of Corrosion Products from
Metals," Corrosion and Degradation of lmplant Materials, ASTM STP 859, A.C.
Fraker and C.D. Griffin, Eds., American Society for Testing Materials,
Philadelphia, 1985, pp. 195 - 207.
Black, J., Biological Performance of Materials: Fundamentals of Biocompatibility,
Marcel Dekker, Inc., 1999.
Mineral Powder Diffraction File, JCPDS International Centre for Diffraction Data,
Swarthmore, U.S.A., 1986.
"Atmospheric and Aqueous Corrosion," Stainless Steels, ASM Specialty Handbook,
J.R. Davis, Ed., ASM Intemational, Materials Park, Ohio, U.S.A., 1994, pp. 133 180.
Sheety, R.H., Gilbertson, L.N., and Jacobs, C.H., "The New Surface Finish - A
Method of Improving the Properties of 316L Stainless Steel," [abstract],
Transaction of the Society of Biomaterials, New York, 1987, p. 233.
Ogundele, G.I., and White, W.E., "Polarization Studies on Surgical-Grade Stainless
Steels in Hanks' Physiological Solution," Corrosion and Degradation of Implant
Materials, ASTM STP 859, A.C. Fraker and C.D. Griffin, Eds., American Society
for Testing Materials, Philadelphia, 1985, pp. t 17 - 135.
Kuhn, A.T., "Corrosion of Co-Cr Alloys in Aqueous Environments," Biomaterials,
Vol. 2, 1981, pp. 68 - 77.
Devine, T.M., and Wulff, J., "The Comparative Crevice Corrosion Resistance of
Co-Cr Base Surgical Implant Alloys," J. Electroehem. Soc.: Electrochemical
Science and Technology, Vol. 123, 1976, pp. 1433 - 1437.
Witlert, H.-G., Brob/ick, L.-G., Buchhom, G.H., Jensen, P.H., Krster, G., Lang, I.,
Ochsner, P., and Schenk, R., "Crevice Corrosion of Cemented Titanium Alloy
Stems in Total Hip Replacements," Clinical Orthopaedics and Related Research,
No. 333, 1996, pp. 51 - 75.
Fontana, M.G., and Greene, N.D., Corrosion Engineering, 3TdEdition, McGrawHill, New York, 1986.
Terry O. Woods ]
MRI Safety and Compatibility of Implants and Medical Devices
Reference: Woods, T. 0., "MRI Safety and Compatibility of Implants and Medical
Devices," Stainless Steels for Medical and Surgical Applications, ASTM STP 1438, G. L.
Winters and M. J. Nutt, Eds., ASTM International, West Conshohocken, PA, 2003.
Abstract: Since MRI (magnetic resonance imaging) scanners became available in the
1980's, they have rapidly become one of the most common clinical imaging tools. The
MR environment produces unique safety and compatibility issues for materials used in
implants and medical devices. The principal issues for MR safety and compatibility are
magnetically induced displacement force and torque, radio frequency (RF) heating, and
image artifact. This paper defines and discusses the potential hazards produced by
implants and devices in the MR environment, with an emphasis on stainless steels. It
also describes MR safety and compatibility test methods developed by ASTM and
summarizes ongoing MR standards development work.
Keywords: MR1, MRI safety, MR1 compatibility, implants, medical devices
Introduction
MR1 (magnetic resonance imaging) scanners became commercially available in the
mid 1980s. The high quality of the diagnostic information provided has led to their
adoption as one of the most common clinical imaging tools. Demand for greater image
resolution and faster scan times has driven the evolution of scanner technology over the
last 15 years with several-fold increases in field strength and temporal and spatial
gradient fields over those present in early scanners. Interventional MR (magnetic
resonance) procedures, performed inside or near the scanner, are the newest facet of MR
technology to place increased demands for safety and compatibility on stainless steel
surgical instruments and other intraoperative devices like patient monitors and
anaesthesia equipment. It is likely that everyone in the developed world will be exposed
to an MR1 scan at some time in his or her life, so the issues of MR safety and
compatibility of implants and medical devices are ones that directly impact millions of
patients.
The MR environment produces unique safety and compatibility issues for materials
used in implants and medical devices. The FDA proposed definitions for MR safety and
compatibility in "A Primer on Medical Device Interactions with Magnetic Resonance
Imaging Systems," which was released for comment in February 1997 [1]. These
definitions were published in the ASTM Standard Test Method for Evaluation of MR
~Mechanical Engineer, Food and Drug Administration, Center for Devices and
Radiological Health, 9200 Corporate Boulevard, HFZ-150, Rockville, MD, 20850.
82
www.astm.org
WOODS ON MRI SAFETY
83
Artifacts from Passive Implants (F2119). A device is considered MR safe if it presents

no hazards to the patient or other individuals. An MR compatible device is MR safe and
in addition does not signific~/ntly affect the diagnostic information nor have its operations
affected by the MR scanner. For both definitions, the MR conditions in which the device
was tested must be specified because a device that is found to be safe or compatible under
one set of conditions may not be found to be so under more extreme MR conditions. This
is a particularly important warning currently because most devices that are claimed to be
MR safe or compatible have been tested in 1.5 tesla scanners. Recently, 3 tesla scanners
with much harsher MR environments have been introduced and it is likely that some
devices that were safe or compatible in a 1.5 tesla system will no longer be safe or
compatible in a 3 tesla scanner.
Just a few years ago, a needle biopsy was the most complicated interventional MR
procedure. Today however, MR guided neurosurgery is commonly performed in some
centers because the MR gives a unique capability to identify tumor boundaries
intraoperatively. The rapid and widespread development and increasing complexity of
procedures performed in the MR suite is exponentially increasing the demand for MR
safe and compatible devices. Development of these interventional and intraoperative MR
capabilities requires a fully configured MR safe operating room, one which potentially
contains many stainless steel components. The types of applications where stainless
steels might be found in a device in the MR environment include passive implants (stents,
aneurysm clips, sutures), active implants (neurostimulators), external passive devices
(patient positioning devices, electrical leads, surgical hand tools, iv poles, equipment
carts, gas bottles), and external electrically active devices (patient monitors, anaesthesia
equipment, defibrillators). These medical devices range in complexity from the bearings
in a castor on a cart to a life-sustaining implant like an endovascular stent-graft, and yet
in the wrong circumstances, both could potentially interact with an MRI scanner to cause
serious problems for the patient.
The principal issues for MR safety and compatibility are magnetically induced
displacement force and torque, radio frequency (RF) heating, and image artifact.
Gradient field induced voltages are also a potential problem for implants. Adverse events
ranging from deaths caused by magnetically induced displacement force and torque to
third degree burns and images containing large void areas due to the presence of
magnetic materials have happened in the past and continue to happen. Specific examples
are presented below. Fortunately, serious accidents are not common, though they do
continue to happen every year. A number of organizations including the FDA, the
Radiological Society of North America, the American College of Radiologists, the
National Electrical Manufacturers Association (NEMA) MR Technical Committee, and
ECRI (formerly the Emergency Care Research Institute) are working to increase MR
safety awareness through publications, courses, web sites and standards development
activities [2-4].
Standards Development Efforts

ASTM Standard Specification for the Requirements and Disclosure of Self-Closing
Aneurysm Clips (F1542-94), was the first standard to outline a test method for
determining MR safety, though it addresses only the magnetically induced displacement
84
force. In 1998, ASTM Committee F04 on Medical and Surgical Materials and Devices,
Subcommittee 15, Materials Test Methods, formed a task group to develop standards
addressing MR safety and compatibility of implants and medical devices. The group has
completed the ASTM Standard Test Method for Measurement of Magnetically Induced
Displacement Force on Passive Implants in the Magnetic Resonance Environment
(F2052), the ASTM Standard Test Method for Measurement of Radio Frequency Induced
Heating near Passive Implants During Magnetic Resonance Imaging (F2182), and F2119
for determining image artifact. The group is also working on a test method for measuring
magnetically induced torque on passive implants. These standards are the first general
test methods addressing MR safety and compatibility of medical devices. Their scope is
currently limited to passive implanted devices, but the test methods are also appropriate
for a range of active implants as well as passive and active devices that are not implants.
ISO 9713 Neurosurgical implants - Self-closing intracranial aneurysm clips, requires
MR testing to determine magnetically induced forces and torques and image artifacts, but
gives no test methods. IEC 60601-2-33, Particular Requirements for the Safety of
Magnetic Resonance Equipment for Medical Diagnosis, defines operating limits for MR
scanners. In particular, it defines safety limits for specific absorption rate (SAR) in the
head, whole body and local tissue [5]. These limits help to define the allowable amount
of RF induced heating.
MRI Effects on Materials and Devices
All materials are either ferromagnetic, paramagnetic, or diamagnetic. As defined in

ASTM F2052, a ferromagnetic material has ordered and parallel magnetic moments that
produce magnetization in one direction; a paramagnetic material has a relative
permeability which is slightly greater than unity, and which is practically independent of
the magnetizing force; and a diamagnetic material has a relative permeability less than
unity. The most common ferromagnetic materials are iron, cobalt, and nickel [6].
Magnetic susceptibilities for even the most weakly magnetic ferromagnetic materials are
orders of magnitude greater than the magnetic susceptibilities ofparamagnetic and
diamagnetic materials [7]. Common examples ofparamagnetic materials are titanium,
aluminum, and platinum [6, 8, 9]. Many metals, copper, for instance, and most
nonmetals are diamagnetic [6].
Martensitic and ferritic stainless steels are ferromagnetic. Ferromagnetic stainless
steels, like the 400 series Class 4 martensitic stainless steel and the Class 6 ferritic
stainless steel in the ASTM Standard Specification for Stainless Steel Billet, Bar, and
Wire for Surgical Instruments (F899), are not MR safe. Austenitic stainless steels are
paramagnetic at room temperature [8]. They include the austenitic stainless steels in the
ASTM Standard Specification for Wrought 18 Chromium-14 Nickel-2.5 Molybdenum
Stainless Steel Bar and Wire for Surgical Implants (UNS $31673) (F138), the ASTM
Standard Specification for Wrought 18 Chromium-14 Nickel-2.5 Molybdenum Stainless
Sheet and Strip for Surgical Implants (UNS $31673) (F 139), the ASTM Standard
Specification for Wrought Nitrogen Strengthened 22 Chromium-12.5 Nickel-5
Manganese-2.5 Molybdenum Stainless Steel Bar and Wire for Surgical Implants (F 1314),
the ASTM Standard Specification for Wrought 18 Chromium-14 Nickel-2.5
Molybdenum Stainless Steel Surgical Fixation Wire (UNS $31673) (F 1350), the ASTM
WOODS ON MRI SAFETY
85
Standard Specification for Wrought Nitrogen Strengthened-21 Chromium- 10 Nickel-3

Manganese-2,5 Molybdenum Stainless Steel Bar for Surgical Implants (F 1586), and the
Class 3 austenitic stainless steel specified in ASTM F899.
These austenitic stainless steels are safe for some MR applications, though cold work
[ 10-12] and heat treatments [6, 13] during manufacturing may produce regions of
magnetic martensite or ferrite that can render a device unsafe in the MR environment.
For instance, it has been shown that martensite transformation produced during heavy
cold working of 316 stainless steel can increase the susceptibility from 0.003 to 9 [8],
transforming a paramagnetic and potentially MR safe material into an unsafe
ferromagnetic material. Because processing can change the magnetic properties of the
material, it is important that the MR safety and compatibility testing for any device be
performed on a finished device. It is also important to consider possible lot to lot
variations in magnetic properties of starting materials and possible effects of processing
changes on the magnetic properties of a finished device when determining the frequency
and sample size needed for MR safety and compatibility testing. In spite of these
caveats, there are many applications for which the appropriate austenitic stainless steel is
an acceptable material for a device or device component in the MR environment.
Magnetically Induced Displacement Force

A magnetic displacement force is produced when a magnetic object is exposed to a
spatial gradient in a magnetic field. This force is responsible for the "projectile effect"
familiar to anyone who has held a small magnet near a paper clip or other small magnetic
object. Even though this magnetic field effect is well-known to virtually everyone,
accidents caused when magnetic objects become projectiles in the large magnetic fields
around MR scanners continue to happen. This is in part due to the fact that the field
strength and gradient fields in MR scanners are many times greater than the magnetic
fields to which most people are accustomed. The field strength in a 3 tesla MR scanner,
the largest currently available clinical system, is more than 50,000 times the earth's
magnetic field strength
Objects ranging from scissors and IV poles to forklift tines and "sandbags" containing
steel shot have been pulled into the bores of MR scanners, sometimes causing serious
injuries [1]. An iron filing in a patient's eye was displaced by a scanner's magnetic field,
causing the patient to lose vision in that eye permanently [ 14]. Numerous accidents have
occurred when gas cylinders were pulled into MR scanners. In July 2001, a child
undergoing an MR scan was killed when he was struck in the head by an oxygen bottle
that was pulled into the scanner [15]. This accident occurred the same month an article
was published presenting five projectile cylinder accidents which had taken place in two
institutions, four of which had occurred in the previous three years [16]. The authors
suggested that in spite of MR safety education, projectile cylinder accidents may be
increasing because more patients are being scanned while they are on life support
systems [ 16]. Ironically, even an object as seemingly innocuous as a pillow can become
a projectile in a magnetic field. A pillow containing coil springs that were not detected
during pre-scan screening by a hand-held magnet was attracted into an MR scanner,
fortunately without causing any injuries [ 17]. Later measurements showed that 16% of
the mass of the pillow consisted of metal. Experiments were performed to determine
86
velocities developed by objects moving from the creep point (the point furthest from the
magnet at which the object would accelerate within 30 s of being placed there) into the
bore of the 1.5 tesla scanner. The 1 kg pillow reached a maximum velocity of 33.7 km/h
after undergoing a maximum acceleration o f 9.9 g, while a 91 g metal glasses case
reached a maximum velocity of 77.8 km/h [ 17].
Large magnetic fields can also affect the operation of devices. At least one patient
with an implanted cardiac pacemaker died when the MR system interrupted the
functioning of the pacemaker, and an implanted insulin infusion pump was found to be
non-functional after being exposed to an MR scanner [ 1]. Recent work in our laboratory
has shown that the flow rate produced by some portable oxygen regulators is altered at
the entrance to the bore o f a 2 tesla MR scanner [18].
The magnetically induced displacement force, F, is a function of the magnitude of the
magnetic field gradient and the magnetic moment of the object (which is in turn a
function of the object's magnetization and volume). Or, F = V(moB), where m is the
dipole moment and B is the magnetic field strength [19]. Since the magnetization, M, is
proportional to m, it is also true that Foc V(MoB). For diamagnetic and paramagnetic
materials and for ferromagnetic materials below their magnetic saturation point, the force
is proportional to the product of the magnetic field strength and the gradient of the
magnetic field strength, ( IBI IVB[) [9]. For ferromagnetic materials that have reached
saturation, the force is proportional to the gradient of the magnetic field strength, VB [9].
The displacement force must be measured at the location of the maximum spatial
gradient or the location o f the maximum product of the spatial gradient and the field
strength so it is necessary to determine that location. Magnet system designers often
incorporate shielding, which contains the field within a smaller volume around the
magnet. This shielding can produce greater gradients in the magnetic field, which will
produce greater displacement forces on magnetic objects. Shielding can also complicate
the spatial distribution o f the field and its gradient. For a scanner with a horizontal static
field, Bo, along the long axis (Z axis) o f the magnet, the maximum spatial gradient should
occur in the Z direction near or inside the entrance to the magnet bore. This is due to the
fact that the field lines run along the Z axis until they reach the portal of the magnet and
then begin to bend around in the X and Y directions.
The magnetically induced deflection force may be measured using the method
described in ASTM F2052. In general, a device can be considered to experience an
acceptable level of magnetically induced deflection force if the measured force is less
than the weight of the device.
Magnetically Induced Torque

A torque is produced as an object attempts to align itself with the magnetic field
produced by the imaging magnet, just as a compass needle aligns itself with the earth's
magnetic field. This effect has also caused the death of at least one MR patient.
Magnetically induced torque rotated an intracranial aneurysm clip in a patient who had
entered the MR scan room, but who had not entered the scanner and so had not entered
the most extreme field conditions. The aneurysm clip tore adjacent blood vessels and the
patient hemorrhaged and died [20].
WOODS ON MRI SAFETY
87
The torque is a function of the strength of the magnetic field and the characteristics,
both material and geometric, of the magnetically active object. It is related to the
geometry, composition, distribution of mass, and magnetic properties of the object and
the strength of the magnetic field. The torque may be written as:
T=mxB,
(1)
where m is the dipole moment and B is the magnetic field strength [19].
So, for an object with a given dipole moment, the maximum magnetic torque
produced on an object will occur when the object is in the region of the highest magnetic
field strength, that is, in the imaging volume.
For an MRI scanner with a horizontal magnetic field:
B = Bo ~
(2)
m = (mx~ +my~ + m z 2 )
(3)
T = (rex x + my ~ + mz ~ ) Bo ~ = Bomy~ -Bomx~
(4)
T = Tx ~ + Ty )3
(5)
and
SO,
or
Therefore, for a scanner with a horizontal magnetic field, the torque may have
components in the x and y directions. It is very important to note that while the direction
of B is always in the z direction, the direction o f m varies as the orientation of the device
varies with respect to the scanner. So, as the orientation of an object changes with
respect to the scanner, the magnitude of the torque (and the magnitudes of the
components of the torque in the x and y directions) will change.
If an object has a spherical symmetry, it will have no "main axis," and there will be no
preferred orientation with the magnetic field and no torque experienced, even if the object
has substantial (but uniform) magnetic susceptibility. However, a spherically symmetric
object would still experience a magnetically induced displacement force.
A number of methods have been used to measure the magnetically induced torque.
There is currently no published standard test method, however one method is the subject
of an F04.15 draft standard test method. In all of these test methods, the maximum
magnetically induced torque is compared to the maximum torque on the device produced
by gravitational forces. In addition to the other MR safety constraints, the maximum
magnetically induced torque acting on a device must be less than or equal to the
gravitational torque for a device to be determined to be safe in the MR environment.
RF Heating
RF heating is heating of the body produced by currents induced by the pulsed RF
imaging gradients applied during scanning. The potential for this problem is greatly
increased by the presence of metallic implants or other medical devices, particularly ones
with the shape of a long wire, for instance in an electrode for a nerve or cardiac
stimulator [21 ]. Patients have been burned severely when they were positioned
88
improperly within an MR scanner so that conductive loops were formed, for instance
when clasping their hands together, or when lying so that there was a small point contact
between a patient's calves. Many burns have been reported involving leads that were
coiled to produce loops and then were allowed to touch the patient inside the scanner,
including burns severe enough to require skin grafts [ 1, 22]. There also have been many
incidents with pulse oximeters, some requiring skin grafts, and some severe enough to
require amputation of portions o f fingers or toes [23, 24].
As with torque testing, a number o f methods have been used to determine RF heating.
ASTM F2182, a test method for measuring RF induced heating of passive implants was
published in 2002. This method is the only published standard test method for
determining RF induced heating of medical devices. The FDA follows the
IEC60601-2-33 guidelines for acceptable levels o f RF heating during clinical scanning.
Image Artifact
In MR images, artifacts may appear as void regions where no signal is seen or as
geometric distortions of the true image [8]. There are three primary mechanisms for
production of an MR imaging artifact by an implant [25, 26]. A ferromagnetic material
may produce a static field in addition to the uniform magnetic field produced by the
magnet. This will perturb the relationship between position and frequency that is
essential for accurate image reconstruction [26]. Magnetic objects may produce
appreciable distortion if they have a magnetic susceptibility sufficiently different than
that for tissue. The presence of such objects can lead to macroscopic imperfections of the
magnetic field that lead to nonparallel flux lines [25]. The extent of the artifact depends
upon the magnetic susceptibility, mass, shape, orientation, and position o f the object, as
well as the method used for image processing. Finally, an implant may exhibit an
induced eddy current due to the incident RF magnetic field. This will alter the RF field
near the implant, thereby creating distortion. The extent of this artifact depends upon the
factors listed above as well as the conductivity of the implant.
The image artifact does not generally affect the MR safety o f a device, though there
has been at least one report of surgery that was performed on a patient based on an
artifact that was present in an MR image [1]. The amount and location of the artifact
with respect to the location of the region the clinician wishes to image does impact the
MR compatibility of a device. It is also important to note that a complete absence of
artifact is not always desirable. Manufacturers have gone to considerable effort to design
MR safe biopsy needles that have a small enough artifact to allow the region of interest to
be visible but enough artifact to allow the needle to be visualized as it is inserted so that it
may be precisely positioned in the tissue that is to be sampled. ASTM F2119 identifies a
protocol for determining image artifact using standardized pulse sequences.
Discussion and Conclusions
With the advent and rapid proliferation of MR scanning technology, medical devices
have been faced with new safety issues. The principal issues affecting MR safety and
compatibility are magnetically induced displacement force and torque, RF induced
heating and image artifacts. This paper identifies test methods that can be used to
WOODS ON MRI SAFETY
89
determine the safety and compatibility of a device in the MR environment. The current
ASTM test methods, which are the only existing published standard test methods
addressing MR safety and compatibility of medical devices, are written for passive
implanted devices. There is a great need for continued work to expand the scope of these
test methods and to develop additional test methods for electrically active implants and
devices that are not implants.
Stainless steels have been and will continue to be widely used in medical devices. If
properly chosen and fabricated, devices composed of or containing austenitic stainless
steel can be safe in the MR environment. However, because of potential hazards caused
by the presence of medical devices in and near MR scanners, manufacturers and users
must take extreme care to ensure that any device composed of or containing stainless
steel is MR safe and that "identical" devices from different lots of raw material o r
different production runs are also MR safe.
References
[ 1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[ 10]
[ 11]
[12]
Center for Devices and Radiological Health, "A Primer on Medical Device
Interactions with Magnetic Resonance Imaging Systems,"
http://www.fda.gov/cdrh/ode/primerf6.html, February 7, 1997.
Center for Devices and Radiological Health,
http://www.fda.gov/cdrh/safety/mrisafety.html, "MRI Safety Web Site."
ECRI, "The Safe Use of Equipment in the Magnetic Resonance Environment,"
Health Devices, Vol. 30, 2001, pp. 421-444.
E. Kanal, Special Cross-Specialty Categorical Course in Diagnostic Radiology:
Practical MR Safety Considerationsfor Physicians, Physicists, and Technologists

2001. RSNA, Oak Brook, IL, 2001.
L.A. Zaremba, "Regulatory Issues in MR Safety," in Special Cross-Specialty
Categorical Course in Diagnostic Radiology: Practical MR Safety
Considerations for Physicians, Physicists, and Technologists 2001, E. Kanal, Ed.
RSNA, Oak Brook, IL, 2001, pp. 103-110.
R.M. Bozorth, Ferromagnetism. IEEE, Inc., New York, NY, 1978.
L.H. Van Vlack, Materials Sciencefor Engineers. Addison-Wesley, Reading,
MA, 1970.
J.F. Schenek, "The Role of Magnetic Susceptibility in Magnetic Resonance
Imaging: MRI Magnetic Compatibility of the First and Second Kinds," Medical
Physics, Vol. 23, 1996, pp. 815-850.
E.M. Purcell, Electricity and Magnetism, Berkeley Physics Course, Volume 2.
McGraw-Hill, New York, 1985.
J.T. McFadden, "Metallurgical Principles in Neurosurgery," Journal of
Neurosurgery, Vol. 31, 1969, pp. 373-385.
M. Dujovny, N. Kossovsky, R. Kossowsky, R. Valdivia, J. S. Suk, F. G. Diaz, K.
Berman, and W. Cleary, "Aneurysm Clip Motion during Magnetic Resonance
Imaging: In Vivo Experimental Study with Metallurgical Factor Analysis,"
Neurosurgery, Vol. 17, 1985, pp. 543-548.
M. Dujovny, N. Kossovsky, R. Kossowsky, F. G. Diaz, and J. I. Ausman,
"Magnetic Aneurysm Clips: Correlation with Martensite Content and
90
[13]
[14]
[15]
[ 16]
[ 17]
[ 18]
[ 19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
Implications for Nuclear Magnetic Resonance Examination," Surgical Forum,

Vol. 34, 1983, pp. 525-527.
R.M. Brick, A. W. Pense, and R. B. Gordon, Structure and Properties of
Engineering Materials, 4th ed., McGraw Hill, New York, 1977.
CDRH, MDR-R100222, http://www.fda.gov/cdrh/maude.html, 1/8/85.
D.W. Chen, "Boy, 6, dies of skult injury during M.R.I.," New York Times. New
York, July 31, 2001, pp. B1, B5.
G. Chaljub, L. A. Karmer, R. F. Johnson III, J. R. F. Johnson, H. Singh, and W.
Crow, "Projectile Cylinder Accidents Resulting from the Presence of
Ferromagnetic Nitrous Oxide or Oxygen Tanks in the MR Suite," American
Journal ofRoentgenology, Vol. 177, 2001, pp. 27-30.
B. Condon, D. M. Hadley, and R. Hodgson, "The ferromagnetic pillow: a
potential MR hazard not detectable by a hand-held magnet," The British Journal
of Radiology, Vol. 74, 2001, pp. 847-851.
T.O. Woods, L. W. Grossman, A. A. Graham, and J. N. Johannessen, "The Effect
of a 2T Magnetic Field on Flow Delivered by Oxygen Regulators," 2002 FDA
Science Forum - FDA: Building a Multidisciplinary Foundation, Washington,
DC, 2002.
R.J. Thome and J. M. Tarrh, MHD and Fusion Magnetics, Field and Force
Design Concepts, John Wiley & Sons, New York 1982.
R.P. Klucznik, D. A. Cartier, R. Pyka, and R. W. Haid, "Placement of a
Ferromagnetic Intracerebral Aneurysm Clip in a Magnetic Field with a Fatal
Outcome," Radiology, Vol. 187, 1993, pp. 855-856.
J.A. Nyenhuis, J. D. Bourland, and A. V. Kildishev, "Safety Considerations of
Rapidly Switched Gradients in MR Environments," in Special Cross-Specialty
Categorical Course in Diagnostic Radiology: Practical MR Safety

Considerations for Physicians, Physicists, and Technologists 2001, E. Kanal, Ed.
RSNA, Oak Brook, IL, 2001, pp. 103-110.
CDRH, MDR-M709484, http://www.fda.gov/cdrh/maude.html, 5/1/95.
E.M. Bellon, et. al, "MR Artifacts: A Review," American Journal of
Roentgenology, Vol. 147, 1986, pp. 1271-1281.
F.G. Shellock, "Biological Effects and Safety Aspects of Magnetic Resonance
Imaging," Magnetic Resonance Quarterly, Vol. 5, 1989, pp. 243-261.
Corrosion
Lukas Eschbach, Gianni Bigolin, Werner Hirsiger, and Beat Gasser ~
Fatigue of Small Bone Fragment Fixation Plates Made from Low-Nickel Steel
Reference: Eschbach, L., Bigolin, G., Hirsiger, W., and Gasser B., "Fatigue of Small
Bone Fragment Fixation Plates Made from Low-Nickel Steel," Stainless Steelsfor
Abstract: Today's implant quality stainless steels contain up to 16 wt% nickel although
nickel ions are the most widespread skin contact allergens. Previously sensitized persons
may develop allergic reactions when nickel is released from stainless steel medical implants. New low-nickel stainless steels combine the benefits of excellent mechanical
properties with virtual absence of nickel.
Miniature bone plates and corresponding 2.0 mm screws for the fixation of small
bone fragments were produced of a low-nickel stainless steel. The implants were tested in
a static reverse-bending setup and under dynamic conditions, and compared to commercially pure (CP) titanium and standard 316L implant steel counterparts. The low-nickel
plate could withstand over 200 cycles of bending, whereas the titanium plate broke at 26
cycles. This confirms the higher tolerance of the low-nickel plate to multiple contouring
during surgery.
Nevertheless, high cycle fatigue tests under physiologic conditions showed that the
low-nickel steel plates exhibit lower resistance to cyclic loads than titanium and 316L
plates. SEM investigations of the fatigue fractures confirmed that the cracks preferentially propagate along grain boundaries leading to intergranular fracture of the low-nickel
steel. It is suggested that the intergranular crack initiation facilitates the early failure under high cycle fatigue conditions, whereas plastic bending properties are not affected.
The tendency to intergranular crack initiation in the low-nickel steel could stem from
surface deformation (work hardening) introduced during machining and related embrittlement in the surface zone.
Keywords: Low-nickel steel, fatigue resistance, bending properties, bone plate, bone
screw.
1 Scientific co-worker, technical co-worker, technical co-worker, and head of basic

research, Dr Robert Mathys Foundation, Bischmattstrasse 12, CH-2544 Bettlach,
Switzerland.
93
www.astm.org
94
Introduction and General Overview of Low-NickeL Steels
High Nitrogen Stainless Steels

Today's implant quality stainless steels contain 9 to 16 wt% nickel although nickel
ions are the most widespread skin contact allergens, and more than 20% of the European
female population suffers from nickel allergy [ 1]. Previously sensitized persons may develop allergic reactions and significantly more aseptic or septic complications when
nickel is released from stainless steel medical implants [2]. Therefore, titanium implants
are used in patients with known nickel allergies. Titanium and titanium alloys have better
biocompatibility and corrosion resistance than stainless steel, but titanium disadvantages
include lower mechanical strength, reduced toughness, and higher price.
High nitrogen stainless steels were developed during the 1980s after the successful
introduction of the pressurized electro-slag-remelting process. They were first thought to
increase the performance of petrol, offshore and power generation facilities, because of
the positive effect of nitrogen on strength and corrosion resistance. However, nitrogen as
an alloying element in steel has the additional benefit of stabilizing the austenitic phase.
In fact, nitrogen has been found to be 18 times more efficient than nickel in stabilizing
the gamma phase [3]. It is therefore possible to replace nickel by nitrogen in austenitic
stainless steels. The resulting virtually nickel-free steels have attracted the attention of
producers of medical implants as well as watch case manufacturers, as possible problems
with nickel allergies can be omitted.
Some low-nickel stainless steel compositions have been developed for medical or
implant applications during the last years (Table 1). All compositions have a high manganese content, which is necessary for increased nitrogen solubility. In alloys with highest manganese contents, nitrogen is even brought into solution without pressure. For the
production of the lower manganese composition, remelting under pressurized nitrogen is
needed to bring enough nitrogen into the melt. Nitrogen replaces nickel in advanced lownickel stainless steels, is a powerful austenite stabilizing element, and therefore prevents
the formation of secondary magnetic phases.
Table 1 - Typical composition (wt%) of known low-Ni steelsfor medical applications.
Balance: Fe
Low-Ni steel
Cr
Mn
Mo
Ni
Carpenter BioDur Tr~ 108 2

Boehler P558 3
VSG P2000 4
Aubert & DuvalNonic M1 5
21
17
18
18
23
11
14
22
1
3
3
3
< 0.05
<0.2
Nb
_<0.25
<0.05
2 Carpenter Technology Corporation, Reading, PA.

3 Boehler Edelstahl GMBH, Kapfenberg, Austria.
4 VSG Energie- und Schmiedetechnik, Essen, Germany.
5 AUBERT & DUVAL, Neuilly-sur-Seine, France.
_<0.08
0.2
< 0.15
< 0.08
> 0.90
0.5
0.75-1.00
0.60-0.90
ESCHBACH ET AL. ON SMALL BONE FRAGMENT FIXATION PLATES
95
Mechanical Properties
Nitrogen strengthened low-nickel stainless steels typically show excellent static
mechanical properties, which are clearly better than today's standard implant stainless
steels [1,4,5]. In the annealed (soft) condition, the strength is much higher than in the
conventional 316L implant stainless steel, whereas elongation values lie in the same
range (Table 2) [1]. It is well known that the strength o f austenitic stainless steels can be
increased by cold deformation, whereas the elongation properties are reduced. This is
usually done with implant raw material when higher strength is required. Low-nickel
stainless steels exhibit a stronger potential for work hardening than conventional stainless
steels (Figure 1). Earlier studies have shown that the strengthening effect of cold deformation increases with increasing nitrogen content [6]. This opens new possibilities for
high-strength implant applications or for the dimensional reduction (miniaturizing) of
implant systems where anatomically limited space is an issue.
Table 2 - Typical mechanical properties of standard and low-nickel stainless implant
steel (17% Cr, 12% Mn, 3% Mo, and 0.9% N) in the annealed condition [1].
Tensile yield strength, TYS

Ultimate tensile strength, UTS
Elongation at break
316L
Implantsteel
(ISO 5832-1)
Low-Ni steel
220 - 260 MPa

500 - 540 MPa
55- 65 %
610 - 720 MPa

980 - 1120 MPa
55 - 65 %
Not much is known about the fatigue properties of low-nickel steels. Although the
static mechanical properties of these steels are very promising, there are some indications
that fatigue strength is below that of 316L implant material. In one study, this fact was
related to a manufacturing process that was not optimal [7].
2800
2600
2400
2200
2000
lOOO
1600
c~
r
1400
1200
N
"~
1000
800
16L
600
4OO
2O0
0
20
40
60
80
100
degree ofcoldwork[%reduction in cress section]
Figure 1 - Increase of tensile yield strength as an effect of cold work [1].
96
Corrosion Resistance and Biocompatibility

The low-nickel steels are very corrosion resistant because o f their high chromium,
molybdenum, and nitrogen content [8]. The pitting resistance equivalent (PRE) is an empirical formula to quantify the effect of different alloying elements on localized corrosion
resistance. It can, for example, be derived from measurements of the critical crevice corrosion temperature. For standard nickel containing stainless steel the PRE is defined as
%Cr + 3.3 x %Mo, This sum shall be at least 26 for implantable compositions, according
to ASTM F 138 and ISO 5832-1, where this formula is stated as compositional requirement. For the high-nitrogen steel BioDur 108, the PRE has been defined as %Cr + 3.2 x
%Mo + 8 x %N [9]. For compositions similar to the Boehler P558 or the VSG P2000, the
PRE has been found to be %Cr + 3.3 x %Mo + 20 x %C + 20 x %N [1]. This results in
PRE values between 30 and 45 for the nitrogen containing steels. Regardless of the formula used to calculate the PRE, it is clear that nitrogen provides significantly increased
corrosion resistance. The improved corrosion resistance has been confirmed in laboratory
tests and represents a reduction of metallic ions and corrosion products available to interact with the biological environment. In anodic polarization measurements, the pitting
potential of BioDur 108 was found to be higher than that of 316L implant steel but lower
than in commercially pure titanium [10]. The repassivation potential on the other hand
was lower than in 316L, which could indicate some limitations if the passive layer is destroyed.
Wear corrosion resistance of Boehler P558 was compared to standard implant
steels in a recent study [11]. P558 performed much better than implant quality 316L (ISO
5832-1, ASTM F 138) and Rex734 (ISO 5832-9, ASTM F 1586) in slurry wear tests, pinon-disk, and crevice corrosion temperature measurements. The resistance against pitting
corrosion was found to be equivalent to that of Rex734.
Although low-nickel stainless steels are still under evaluation with regard to surgical implant applications, the biocompatibility appears to be satisfactory. Investigations on
cell toxicity have been made, and no adverse reactions were detected [12 - 14]. Furthermore, an animal study focused on the different effects of implant quality stainless steel,
pure titanium, and low-nickel steel on blood circulation [ 15]. Using the hamster dorsal
skinfold chamber model and intravital microscopy, it was demonstrated that reducing the
nickel content in stainless steel implants has a positive effect on local microvascular parameters.
Further Considerations
Unfortunately, the high strength and toughness of low-nickel stainless steels are responsible for poor machinability. The ability to optimize machining parameters and the
use of near net shapes represent future challenges for implant manufacturers. Trials of
forging and Metal Injection Molding [16] of low-nickel steels have successfully been
carried out.
Introducing new materials for implants always asks the question o f adverse reactions provoked by direct contact of implants of different materials (galvanic coupling).
Implant quality 316L and low-nickel stainless steel plate and screw samples were investigated in pure and mixed fretting corrosion combinations [17]. The mixed combinations
97
showed even less fretting than both pure combinations. Based on this result, it was coneluded that no additional problems o f fretting corrosion are to be expected using this lownickel stainless steel in combination with conventional implant quality stainless steel.
Low-nickel stainless steels typically contain a residual nickel content of < 0.1%.
Although this content is very low compared to all stainless implant steels, the biological
effect of this residual nickel content on a strongly sensitized patient is unknown. Additionally, there are occasional cases of chromium allergy. For these patients, low-nickel
steels will not be a reasonable alternative to titanium implants.
The price (low-nickel steels are expected to cost about twice the price of 316L implant steel raw material [ 18]), availability, and machinability of these new low-nickel
implant steels will be influential in determining the widespread use in bone surgery.
However, the scope of this study was to evaluate the feasibility o f producing an
implant system from new low-nickel steel and to compare the relevant mechanical properties to standard implant materials.
Low-Nickel Steel as a Candidate Material f o r Small Bone Plates
Synthes | Compact Lock Plates 6 are small implants for the fixation ofmidface and
mandible fractures as well as for osteotomies and reconstructions (Figure 2). These plates
have an overall length of about 40 mm. Corresponding bone screws have an outer diameter of 2.0 mm, are self-drilling and self-tapping due to their special tip design, and
have a threaded conical head that locks in the plate hole, providing angular stability. Today, this fixation system is made o f CP titanium (plates) and titanium alloy Ti-6A1-7Nb
(screws). Known disadvantages of the titanium system include limited formability of the
plate (multiple contouring is not possible, the titanium plate risks to break), low cutting
performance o f the self-drilling screws in hard bone because of the limited hardness of
Ti-6AI-7Nb, possible tendon irritations, and too high an amount of tissue ingrowth on the
titanium surface, resulting in difficulties at explantation of temporary implants.
Figure 2 - Compact Lock Plate and 2.0 mm screw prototypes.

6 Synthes AG, Chur, Switzerland.
98
The purpose of this study was to produce the implants of the Compact Lock 2.0
system with low-nickel stainless steel to overcome all the mentioned disadvantages without causing new allergic problems. The prototype system was then to be compared with
the current titanium system and prototype plates made of implant quality 316L stainless
steel. Reverse bending tests were proposed to confirm the superior intraoperative contourability (multiple bending treatment) and high cycle fatigue tests to ensure the longterm durability of the new material during service.
Materials and Test Methods
Compact Lock Plate Prototypes

Compact 6-hole plates and corresponding 2.0 mm screws (Figure 2) for the fixation
of small bone fragments were made of Carpenter BioDur 108 low-nickel stainless steel
(tensile properties of the raw material see Table 3, transverse micrographs see Figure 3)
using conventional machining techniques. The plates were compared to identical plates
made of CP titanium and standard 316L implant steel (Table 3).
Table 3 - Mechanical properties of raw material according to supplier's test certificates
(BioDur 108) and internal recertification results (other materials).
Tensile yield
strength
BioDur I08 round bar (screws)
BioDur 108 strip (plates)
316L implant steel strip (plates)
CP titanium sheet (plates)
938MPa
642MPa
not measured
553;MPa
Ultimate tensile Elongation at

strength
break
1227MPa
1001MPa
837MPa
680MPa
31%
46%
17.5%
23.9%
Figure 3 - Transverse micrograph of the BioDur 108 raw material used for production of
plates and screws; (a) 1.5 mm strip raw material (b) 4 mm diameter round bar.
99
Test Setup
In a first setup, the bone plates were tested for tolerance against multiple reverse
bending procedures. The plates were fixed on one side and plastically bent up and down
in a range o f + 10 mm by a motor-driven actuator at a cantilever of 22 mm (Figure 4).
The velocity of the bending device was 200 mm/min (5 cycles/min) and deformation was
continued until plate rupture. The fractures were always placed through the middle of the
central hole where the highest bending moment occurs.
Figure 4 - Test set-up for reverse bending testing.

The high cycle fatigue properties o f all three plate types were compared in a combined compression-flexion setup (Figure 5). The plates were fixed with 6 screws (0,3 Nm
fixation torque) onto cylinders ofpolyoxymethylene (POM). A gap of 6 mm between the
POM parts simulated the situation of load bearing through the plate without bony support. Dynamic compression load at a minimum to maximum load ration o f R = 0.1 was
applied vertically on the POM cylinders with a frequency of 6 Hz. The complete setup
was placed in a container with 1% NaCl-solution and kept at 37~ to simulate a physiologic situation. Tests were performed at different maximum loads, using new specimen
for every test, and were conducted until failure of the plate or 2.5 mio cycles, respectively.
Figure 5 - Test set-upfor dynamic bending testing,frontal and lateral view.
100
Results
Reverse Bending and Fatigue Tests

The results o f the reverse bending test are summarized in Figure 6. CP titanium
plates broke after 26 + 7 cycles (3 samples, failure after 24, 25, and 29 cycles), whereas
BioDur 108 plates could withstand a mean number o f 226 + 46 cycles (208, 226, and 245
cycles). Plates of 316L stainless steel were not investigated in this setup. The results confirm the significantly (p = 0.001) higher tolerance against multiple plastic deformation of
stainless steel compared to CP titanium. The higher damage tolerance o f steel plates prevents from possible fractures during contouring operations.
300
25O
200
"~ 150
~ 100
50
CP Titanium
Low.Nickel Steel BIoDur 108
Figure 6 - Number of cycles to failure in reverse bending test.

The results of the fatigue tests are summarized in the Woehler diagram (Figure 7).
Differences in fatigue resistance are mainly found at the highest loads: At 20 N maximum load, the first BioDur 108 plate broke at about 5 000 cycles, whereas the best 316L
plate withstood 676 000 cycles. At this load, the titanium plates showed intermediate
fatigue resistance with failures between 60 000 and 85 000 cycles.
At lower maximum loads, CP titanium and 316L plates showed comparable fatigue
resistance, but BioDur 108 plates were still inferior.
Scanning Electron Microscopy

The fracture surfaces of the BioDur 108 plates were analyzed by scanning electron
microscope (SEM) and compared to 316L stainless steel plates to find out more about
possible reasons for early fatigue failures. For this SEM investigation, plates were used
that had been tested at highest fatigue loads (20 N maximum load, 2 N minimum load).
The BioDur 108 plate was the one that broke after 4978 cycles already, the 316L plate
was the best one that broke after 676 584 cycles.
All fatigue fractures occurred at one o f the central holes, where the cross section of
the plate is reduced because o f the hole, and multidirectional loads are found as a consequence of the tight fixation by screws (Figure 8).
1 01
22
20
.....................
18
................................
........
:---'~,
........
.......
~ r
....................
"~
". . . . . . . . . . . . . . . . . . . . . . . . .
l-----~
"%
, ..................
............
'
........ ~.>~
............
Z
1
-,,.
, %
O
..J
e~
16
...............................
!.................................
'.................................
"----"-~z-I
|
!
..................
l i !i i!!!ii !i i i i i i i i i i i !i!i i i i !!i!i i i!i i i i i i i i
14
I
I ..................
|
.............
I
I
I
!
|
.. .. .. .. .. . . . . . .
12
1000
Ii
104
,,
,,
los
lOe
2,5 millions:
Ted
LL
107
Stop
Number of Cycles
Figure 7 - Woehler-diagram: Highest fatigue resistance at high loads is f o u n d in 316L

plates, f o l l o w e d by C P titanium p l a t e s a n d B i o D u r 108 plates. A t l o w e r loads, differences
in fatigue resistance are smaller.
Figure 9 a provides an overview of the fracture surface of the BioDur 108 plate.
Crack initiations were found on both sides of the plate. The cracks propagated from the
upper and lower side to the center of the plate, where the remaining overload fracture
surface is located. A second crack zone was usually found on the opposite side of the
fatigue failure. The examination of the crack initiation site (Figure 9 b) o f these secondary cracks at the plate surface was very helpful in characterizing the type of fracture. It
was found, that the fatigue cracks in BioDur 108 have a great percentage ofintercrystalline propagation (Figure 10 a). It seems that the grain boundaries are a site of preferred
fatigue crack propagation in the BioDur 108 stainless steel. In addition, a spreading of the
fractures along parallel gliding planes can be observed (Figure 10 b).
The secondary, incomplete crack in the 316L stainless steel plate is shown in an
upper (Figure 1 la) and a lower plate side view (11 b). The fatigue cracking results in a
strong deformation o f the thread in the plate hole. The crack initiation zone is characterized by multiple small crack sites. The image o f the same zone at stronger magnification
(Figure 12 a) makes clear that cracking is not intercrystalline such as in the BioDur 108
alloy, but much more transcrystalline. Nevertheless, the fracture characteristic of the
316L plate seems similar in terms of spreading of cracks along parallel gliding planes
(Figure 12 b).
102
Figure 8 -
Upper side view of the fatigue-tested BioDur 108 plate (a) and 316L stainless
steel plate (b).
F i g u r e 9 - BioDur 108 plate: (a) General view of the fracture surface. Crack initiation
was at the top and the bottom of the plate and the craeks propagated from both sides to
the center of the plate. (b) Crack initiation zone on the upper plate side.
Figure
Figure
10 -
11 -
103
Surface of BioDur 108 specimen near the initiation site of fatigue cracks:
(a) Intercrystalline fatigue crack propagation;
(b) Spreading of cracks along gliding planes.
316L stainless steel plate: Crack initiation zone on the upper plate side (a)
and on the lower plate side (b).
104
Figure 12 - Surface of 316L specimen near the initiation site offatigue cracks:
(a) Transcrystalline fatigue crack propagation;

(b) Spreading of cracks along gliding planes.
Since it could be observed that the fatigue cracks in BioDur 108 show a much
greater percentage of intergranular fracture, the metallographic specimens were analyzed
by SEM too. Especially the grain boundaries were checked for possible foreign phases or
precipitation particles. But it was impossible to find any reason for the weakness of grain
boundaries.
Miniature bone plates were produced of new BioDur 108 low-nickel stainless steel
to overcome known disadvantages of titanium, such as poor ductility and formability.
Reverse bending tests confirmed that the low-nickel steel plates are significantly more
tolerant to multiple bending procedures than titanium plates.
Nevertheless, high cycle fatigue tests under physiologic conditions showed that the
low-nickel steel plates exhibit lower resistance to cyclic loads than titanium and 316L
plates of identical size and design. Although the static tensile strength of low-nickel steel
is clearly superior to both pure titanium and implant quality 316L stainless steel, this new
steel showed lowest performance under dynamic conditions, especially at high loads.
SEM investigations of the fatigue fractures confirmed that a different fracture mode is
characteristic for the BioDur 108 plates: a great percentage of the cracks propagate along
grain boundaries leading to intergranular fracture.
It was thought that the high nitrogen content of over 0.90 % could lead to the precipitation of nitrides on the grain boundaries, and hence facilitate intergranular fracture.
But it was not possible to detect foreign phases at the grain boundaries by SEM investigation of the fracture surfaces and the metallographic specimens.
105
Low-nickel steels with high nitrogen content seem to be prone to reduced high cycle fatigue resistance due to preferred intergranular fracture. This is in accordance with
previous reports confirming the inferior fatigue resistance compared to other implant materials [7].
Crack initiation is a critical step under high cycle fatigue conditions, much more
critical than under plastic bending conditions. It is likely that the intergranular crack initiation facilitates the early failure under high cycle fatigue conditions, whereas plastic
bending properties are not affected. This would explain the different outcome of the
BioDur 108 plates in the two tests presented here. The tendency to intergranular crack
initiation could come from surface deformation (work hardening) introduced during machining and related embrittlement in the surface zone. Further development in the processing and machining of low-nickel steel products could help to improve the microstructure of these materials and overcome the possible problem of reduced high cycle fatigue
resistance documented in this study.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
Speidel, M. O., and Uggowitzer, P. J., "Biocompatible Nickel-Free Stainless

Steels to Avoid Nickel Allergy," Materials in Medicine, Proceedings of Materials
Day 1998, ETH Zttrich, Department of Materials, pp. 191-208.
Hierholzer, S. and Hierholzer, G., "Internal Fixation and Metal Allergy- Clinical
Investigations, Immunology and Histology of the Implant Tissue Interface,"
Thieme Medical Publishers, New York, 1992.
Speidel, M. O., "Stickstoffiegierte Stahle - eine l]'bersicht," Ergebnisse der Werkstoff-Forsehung Band 4, Thubal-Kain, Ztirich, 1991, pp. 1-17.
Uggowitzer, P. J., Magdowski, R., and Speidel, M. O., "Nickel Free High Nitrogen Austenitic Steels," ISIJInternational, The Iron and Steel Institute of Japan,
Vol. 36, No. 7, 1996, pp. 901-908.
Stein, G. and Hucklenbroich, I., "Biokompatible stickstofflegierte Austenite ftir
Anwendungen am und im menschlichen Krrper," Werkstoffwoche '98, Band IV,
Symposium 4, WerkstoffeJ~r die Medizintechnik, H. Planck and H. Stallforth,
Eds., Wiley-VCH, Weinheim, 1999, pp. 3-9.
Menzel, J., Kirschner, W., and Stein, G., "High Nitrogen Containing Ni-Free Austenitic Steels for Medical Applications," ISIJInternational, The Iron and Steel Institute of Japan, Vol. 36, No. 7, 1996, pp. 893-900.
Fink, U., Weik, T., and Gtimpel, P., "Nickel-Free Austenitic Steels as a Future
Implant Material," 13th Europ. Conf. on Biomaterials, GOteborg, Sweden, Sept. 47, 1997, p. 140.
Cigada, A., Rondelli, G., Vicentini, B., Brunella, F., and Dallaspezia, G., "Corrosion Behaviour of High Nitrogen Stainless Steels for Biomedical Applications,"
Compatability of Biomedical Implants, San Francisco, California, USA, 23-25

May, 1994, pp.185-195.
106
[9]
[10]
[11]
[ 12]
[ 13]
[14]
[15]
[16]
[ 17]
[ 18]
Brown, R. S. and Gebeau, R. C., "Strength and Corrosion Resistance of BioDur

108 Alloy," Proceedings of the Sixth World Biomaterials Congress, Kamuela,
Hawaii, May 15-20, 2000, p. 828.
Disegi, J. A., Zardiackas, L. D., and Mitchell, D. W., "Anodic Polarization
Evaluation of Nickel-Free Implant Quality Stainless Steel," Proceedings of the
Sixth World Biomaterials Congress, Kamuela, Hawaii, May 15-20, 2000, p. 816.
Thomann, U. I. and Uggowitzer, P. J., "Wear-Corrosion Behavior of Biocompatible Austenitic Stainless Steels," Wear, Vol. 239, 2000, Pr~"48-58.
Carpenter Technology Corporation Alloy Data, BioDur 108 Alloy, Carpenter
Technology corp., Reading, PA, 1999.
Internal Report on Cytotoxicity of Low-Nickel Steels, Robert Mathys Foundation,
Bettlach, Switzerland, 1998.
Hoch6rtler, G., Bernauer, J., and Kriszt, K., "Corrosion Behaviour and Nickel
Release of Alloys with Various Nickel Contents," Materials for Medical Engineering, Proceedings of Euromat 99, Vol. 2, H. Stallforth and P. Revell, Eds.,
Wiley-VCH, Weinheim, 2000, pp. 204-209.
Kraft, C. N., Burian, B., Perlick, L., Wimmer, M. A., Wallny, T., Schmitt, O., and
Diedrich, O., "Impact of a Nickel-Reduced Stainless Steel Implant on Striated
Muscle Microcirculation, A Comparative In Vivo Study," Journal of Biomedical
Materials Research, Vol. 57, 2001, pp. 404-412.
Uggowitzer, P. J., Speidel, M. O., and Wohlfromm, H., "PANACEA Provides the
Answer to Ni Allergy," MPR (Metal Powder Report), Sept. 1998, pp. 48-52.
Eschbach, L., Marti, A., and Gasser, B., "Fretting Corrosion Testing of Internal
Fixation Plates and Screws," Materials for Medical Engineering, Proceedings of
Euromat 99, Vol. 2, H. Stallforth and P. Revell, Eds., Wiley-VCH, Weinheim,
2000, pp. 193-198.
Kramer, K.-H., "Implants for Surgery - A Survey on Metallic Materials," Materials for Medical Engineering, Proceedings of Euromat 99, Vol. 2, H. Stallforth and
P. Revell, Eds., Wiley-VCH, Weinheim, 2000, pp. 9-29.
Lyle D. Zardiackas l, Scott Williamson 1, Michael Roach I, and Jay-Anthony Bogan I

Comparison of Anodie Polarization and Galvanic Corrosion of a Low-Nickel
Stainless Steel to 316LS and 22Cr-13Ni-SMn Stainless Steels
Reference: Zardiackas, L. D., Williamson, S., Roach, M., and Bogan, J. A.,
"Comparison of Anodic Polarization and Galvanic Corrosion of a Low-Nickel
Stainless Steel to 316L S and 22Cr-13Ni-5Mn Stainless Steels," Stainless Steelsfor

ASTM International, West Conshohocken, PA 2003.
Abstract: Over the years, numerous biomedical implants have been produced from
austenitic stainless steels due to their exceptional mechanical and corrosion properties.
High nickel content has been used to stabilize the austenitic microstructure, but the
presence of these nickel concentrations has been implicated as a cause of metal sensitivity
reactions in a minor portion of the population. In a partial response to the possibility of
hypersensitivity reactions, low-nickel stainless steels with a high nitrogen content to
stabilize the austenitic microstructure have been developed. In this study, the anodic
polarization and galvanic corrosion behavior of one such low-nickel alloy (BioDur | 108
Carpenter Technology) was compared to BioDur | 316LS and BioDur | 22Cr-13Ni-5Mn.
For anodic polarization, duplicate samples with a surface ground to 600 grit and duplicate
samples with a metallographic polish were tested in accordance with ASTM G5 at a
sweep rate of 0.6 V/hr, initiated 200 mV below the rest potential. Samples for galvanic
testing were metallographically polished, the rest potentials were recorded prior to test
initiation, and duplicate coupled tests were run for a period of 48 hours for each alloy
combination. The galvanic current and couple potential were recorded in accordance
with ASTM G71. Anodic polarization results revealed BioDur | 108 to have a pitting
potential similar to BioDur | 22Cr-13Ni-5Mn and significantly higher than that of
BioDur | 316LS. In addition, substantially higher pitting levels on BioDur | 316LS were
also noted during post-anodic metallographic examination. None of the alloy
combinations revealed substantial galvanic coupled currents indicating little susceptibility
to galvanic corrosion. No evidence of pitting or crevice corrosion was found on any of
the galvanically coupled surfaces.
Keywords: corrosion, galvanic, stainless steel, low-nickel
1professor and Coordinator of Biomaterials and Professor of Orthopaedic Surgery, senior

materials engineer, materials engineer and senior materials engineer, respectively,
University of Mississippi Medical Center, SOD / Biomaterials, 2500 North State
Street, Jackson, MS, 39216.
2Carpenter Technology Corporation, Reading, PA, USA.
107
www.astm.org
108
Introduction
Austenitic stainless steels are widely used in the production of biomedical implants.
They provide high strength, good corrosion resistance, and are cost effective [1 ].
However, up to this time medical grade stainless steels have used a significant amount of
nickel to maintain their anstenitic microstructure. Nickel is reported to account for as
much as 90% of the clinically observed sensitivity to metals and may lead to other
complications when released from stainless steel implants [1,2]. Fretting and wear, in
addition to other forms of corrosion, of stainless steel devices may lead to metallic ion
release into the tissues surrounding the implants. Histological analyses of these tissues
have shown a nickel concentration as high as 2% [3]. Carpenter Specialty Alloys has
developed a low-nickel austenitic stainless steel (BioDur | 108) in partial response to
concerns over the possibility of nickel induced sensitivity reactions. A nitrogen
concentration of at least 0.85% is used to maintain the austenitic microstructure of this
low-nickel alloy. In addition, this high nitrogen content is also reported to contribute to
the high strength and good corrosion resistance of the material [4,5].
In physiological solutions, several studies have documented greater pitting corrosion
susceptibility of implant quality 316L stainless steel compared to other austenitic
stainless steels used for implant devices [4,5,6]. Potentiodynamic studies on 316L steel
in salt solutions reveal a small passive range, a low breakdown potential, and a large
hysteresis loop between the forward and reverse scans which is indicative of a material
susceptible to pitting corrosion [6,7]. Implant retrieval studies have shown that localized
corrosion is generally not sufficient to result in the mechanical failure of 316L devices.
However, the release of corrosion products over the term of implantation may lead to
tissue inflamation, primarily due to nickel ion release, and possible device loosening
contributing to the need for implant removal [8]. Nevertheless, the corrosion resistance
for implant quality 316L remains generally acceptable for many implant applications [9].
High manganese anstenitie stainless steels, such as 22Cr-13Ni-5Mn, benefit from a
combination of molybdenum and elevated nitrogen levels as well as chromium to raise
their pitting potential [10]. Cyclic polarization studies in Ringer's solution reveal 22Cr13Ni-5Mn to exhibit a much larger passive region, higher pitting potentials, and a very
small hysteresis compared to 316L, indicating superior pitting resistance [4,6,11]. Initial
studies on the corrosion properties ofBioDur | 108 stainless steel suggest excellent pitting
and crevice corrosion resistance with similar oxide breakdown potential to those of 22Cr13Ni-5Mn stainless steel [4,5].
The direct contact or electrical couple of two different metals leads to the formation of
a galvanic couple [ 10,12]. Differences in composition and processing of implant quality
austenitie stainless steels may lead to galvanic corrosion effects if dissimilar metals are in
contact. Initial studies of mixed metal combinations ofBioDur | 108 and implant
stainless steels have shown good results in fretting tests [ 1]. However, previous studies
have shown that even mixing implant plates and screws of the same material from
different manufacturers may lead to galvanic effects [13].
The purpose of this study was to characterize the anodic polarization and galvanic
corrosion ofBioDur | 108 (alloy A), as compared to both BioDur | 316LS (alloy B) and
BioDur | 22 Cr-13 Ni-5 Mn (alloy C) due to their long and successful history as implants.
ZARDIACKAS ET AL. ON ANODIC POLARIZATION AND GALVANIC CORROSION
109
Samples were prepared from a single lot of 8 mm centerless ground bar stock in the
cold worked condition of each alloy. Transverse and longitudinal samples of each alloy
were mounted in diallyl phthalate, ground, and polished to a final finish of0.051am A1203
for optical metallography. The inclusion content was evaluated on longitudinal samples
according to ASTM standard "Test Methods for Determining the Inclusion Content of
Steel" (E45 - 97 Method A using Plate l-r). After examination for inclusions, the
longitudinal samples were etched at 3 volts for 20 seconds in a solution of 45g KOH in
100 mL of distilled water to evaluate for free ferrite and other foreign phases. Transverse
samples of alloy B were etched for grain size and grain structure determination using a
solution containing 30g FeC13, 30 mL HC1, and 1.8 mL of HNO3. Samples of alloy A and
alloy C were etched in a solution of 45 % KOH solution electrolytically at 3 volts to
reveal the grain structure. The values for the grain size were determined as defined by
ASTM standard "Test Methods for Determining Average Grain Size" (El 12 - 96) on
transverse samples as defined by ASTM standards "Wrought Nitrogen Strengthened -22
Chromium-12.5 Nickel-5 Manganese-2.5 Molybdenum Stainless Steel Bar and Wire for
Surgical Implants" (F 1314-01) and ASTM "Wrought 18 Chromium- 14 Nickel-2.5
Molybdenum Stainless Steel Bar and Wire for Surgical Implants (UNS $31673)" (F 13800).
Quantitative compositional evaluation of the major alloying elements of each material
was evaluated using inductively coupled plasma emission spectrometry (ICP) with a
spark attachment for analysis of solid materials (Spectro ICP with LISA3). Each alloy
was analyzed in duplicate using calibration curves developed using standard samples.
Results of analysis were compared to ASTM and/or internal melt limits as applicable to
ensure each alloy met specification.
Corrosion testing was performed by dynamic cyclical anodic polarization using a
Gamry4 PC3 potentiostat. Two transverse samples with a cross section of 0.5cm 2 were
mounted in Durofast 5 epoxy resin prior to polishing and two like samples were prepared
for grinding. Grinding was performed with 600 grit silicon carbide paper and polishing
was performed to a final finish with 0.05gm A1203. Polished samples were photographed
at 100x magnification prior to and after anodic polarization to document surface
condition. Samples were placed in a corrosion cell having a standard calomel reference
electrode, two carbon counter electrodes, and Ringers solution (pH 5.5) at 37~ The cell
was purged with N z for the two hours prior to initiation of polarization, and N 2 was
continuously bubbled through the solution for the duration of the test. During the initial
purging with N2, the rest potential was allowed to equilibrate. Anodic polarization scans
were initiated 200mV below the rest potential (EOC), tested at a sweep rate of 0.6V/hr,
and reversed when the potential reached 3.0 volts with respect to the reference (Eref) or
when the current density reached 5mA/cm2. Polarization plots of potential (E) vs. log
current density (A/cm 2) were recorded and Tafel extrapolations were prepared to
determine for Eco~ and Ico~ for each test.
3SPECTRO Analytical Instruments, Fitchburg, MA, USA.
4Gamry Instruments, Inc., Warminster, PA, USA.
5Durofast/Struers, Westlake, OH, USA.
110
Transverse samples with a 0.5cm 2 cross section were mounted in Durofast epoxy resin
prior to polishing to a final finish with 0.05~tm A1203 for galvanic testing.. Micrographs
were taken at 100x prior to and after testing to document the surface condition. Each
alloy (A, B, and C) was coupled to each of the other two alloys to determine the galvanic
potential and current. Each pair was placed in a corrosion cell containing a standard
calomel electrode and Ringers solution at 37~ (pH 5.5). The cell was purged with
nitrogen gas for two hours prior to initiation of the test and nitrogen gas was bubbled
through the solution though out the test to minimize crevice effects. After the initial two
hour purge, each pair of alloys was galvanically coupled and tested using the zero
ammeter technique as outlined in the ASTM standard "Guide for Conducting and
Evaluating Galvanic Corrosion Tests in Electrolytes" [G71-81 (Reapproved 98)]. The
rest potentials of each individual alloy were recorded immediately prior to coupling the
alloys and test initiation. Duplicate coupled scans were run for a period of 48 hours on
each alloy combination. This particular time period was chosen based upon initial testing
which showed no changes in potential or current after the 48 hour time period. The
galvanic current and couple potential were recorded over the duration of the test in
accordance with ASTM G 71.
Results
ICP compositional analysis results are provided in Table 1. All alloys were found to
be within ASTM specification or Carpenter internal melt limits for major alloying
elements. Metallurgical evaluation found alloy A to have a 2.0 thin globular oxide and a
0.5 thin alumina type inclusion rating: Alloy B was found to have an inclusion rating of
1.0 for thin globular oxides and a 0.5 for thin alumina. Alloy C was found to have a 2.0
thin globular oxide and a 1.0 heavy globular oxide inclusion rating. Alloy A was
evaluated to have the largest ASTM grain size at 5.0, followed by alloy B at 6.0, and
finally alloy C at 9.0, respectively.
Anodic polarization results are listed in Table 2, and shown in Figures 1 and 2. Alloy
C had the highest rest potential of the three austenitic stainless steels, followed by alloy
B, and finally alloy A. The Ico~ values for all alloys tested were similar, with alloy C
having a slightly more positive value than alloys A and B. Alloys A and C recorded
similar breakdown potentials, which were significantly greater than those of alloy B.
Potential scans for each alloy tested revealed a slight shift between the ground and
polished samples. Ground samples revealed slightly more negative rest potentials, and
slightly higher current density values than the polished samples. Of the three alloys, the
most pronounced shift between the ground and polished conditions was observed in alloy
B. Metallographic examination of the surfaces of alloys A and C (Figures 3a and 3c)
revealed significantly less evidence of pitting attack than those of alloy B (Figure 3b).
Galvanic couple results for each alloy combination of the stainless steels tested are
provided in Table 3 and Figures 4 and 5. The coupling of each combination of alloys
created an initial over-potential due to a net current caused by the differences in the rest
potentials of the individual alloys. This change in potential and current density, for the
coupled alloy combinations, stabilized over the first 24 hours of the test (Figures 4 and 5).
After the first 24 hours of the tests, the coupled current densities continued to approach
zero and the coupled potentials maintained a stable value. Metallographic examination of
Alloy C
Alloy B
Alloy A
Material
2.78
23.48
0.33
Si
0.002
0.005
0.006
0.020
0.020
0.005
Polish to 0.05
I Micron Alumina
Polish to 600 Grit
SiC paper
Polish to 0.05
Micron Alumina
Polish to 600 Grit
SiC paper
Polish to 0.05
Micron Alumina
Polish to 600 Grit
SiC oaoer
Condition
Open Circuit
Potential
(mV)
-365.5
-366.5
-406.8
-471.1
-248.9
-237.3
-298.1
-311.1
-224.2
-220.8
-224.0
-271.1
8.72E-08
5.35E-08
7.40E-08
1.40E-07
5.37E-08
6.07E-08
6.35E-08
9.48E-08
1.00E-07
7.71E-08
1.24E-07
1.15E-07
Ieorr (A/cm2)
--
0.050
--
Cu
Passivation
Potential
(mV @~A/cm')
- [email protected]
[email protected]
- [email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
9
[email protected]
[email protected]
[email protected]
-58(&6.20E-07
Table 2 - Anodic Polarization Data vs SCE (Ringer's @ 3 7 ~C, n =2)
14.09
0.75
Mn
0.42
17.78
0.03
Mo
Alloy C
Bal.
21.42
13.41
2.31
5.06
1 Analysis performed by Carpenter Technology Corporation
Bal.
Alloy B
20.95
Ni
0.50
Bal.
Alloy A
Cr
1.74
Fe
Alloy
Table 1 - Compositional Analysis of BioDur Alloys (%) (n=2)
0.33
0.037
1.06
N1
Breakdown
Potential
(mV @ A/cm2)
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
1050(-~3.13E-05
0.017
0.017
0.038
C1
0
z
0
0
"in
<
r-
N
-4
i-=
Z
0
O
Z
,t-
m
-,"I
~K
>
20
e~
u~
-8.0
- -
....
-7.0
Alloy C 600 Grit SiC
Alloy B 600 Grit SiC
-6.0
Alloy A 600 Grit SiC
-5.0
-4.0
Log Current Density (A/era2)
,/--
-3.0
-2.0
-1.0
0.0
finish in Ringer's solution at 37 ~
Figure 1 - Representative anodic polarization scans of each alloy ground to a 600 grit SiC surface
-9.0
-1.000
-0,500
0.000
0.500
1.000
1.500
2,000
Z
09
-'4
"o
if-
r'-"
c
~J
G)
r-
m
m
roo
"m
0
~J
m
Go
oo
o)
.-~
i--
I~O
',.w_,
- 1.000
-8.0
-7.0
. ~. . ~ . . , , . ~ -
Alloy C Polished
-6.0
-5.0
-4.0
-3.0
-2.0
- 1.0
0.0
Log Current Density (A/cm2)
finish in Ringer "s Solution at 37 ~
"-t,
O
23
~3
O
60
<
z
I'=
z
121
.-t
r-
12
0
Z
0
F
0
Z
GO
m
c)
Alloy B Polished
23
c2
Alloy A Polished
-. --- -------'.-
- -
....
F i g u r e 2 - Representative anodic polarization scans o f each alloy polished to a O.05ltm alumina
-9.0
-0,500
0.000
0,500
1.000
1.500
2.000
114
Representative micrograph
of the post-anodic polarization surface
of alloy A (IOOX).
Figure 3a -
of alloy B (IOOX).
Figure 3b -
of alloy C (I OOX).
Figure 3c -
Table 3 -
115
Galvanic Corrosion Coupling Results (n=2)
Coupled
Initial
In~ial
ln~ial
Initial
Final
Final Coupled
Alloys
#1 - #2
Uncoupled
Potential #1
(mV)
Uncoupled
Potential #2
(mV)
Coupled
Potential
(mV)
Coupled
Current
Densi~
Coupled
Potential
(mV)
Current
Densi~
(A/cm2)
(A/cm2)
A- B
A- B
-319
-264
-382
-335
-350
-288
1.81E-08
-2.88E-08
-142
-215
7.81E-10
-5.78E-10
A- C
A- C
-333
-330
-318
-315
-332
-302
-1.55E-09
-1.36E-09
-210
-213
,3
1.66E-09
1.34E-09
B -C
B-C
-243
-359
-383
-338
-307
-355
-5.32E-08
9.60E-09
-233
-330
-7,02E-10
1.28E-09
the samples after testing revealed little, i f any, evidence o f pitting attack on the sample
surfaces o f the alloys tested.
5.00E-08
- - " Alloy A - Alloy B
4.00E-08
.....
Alloy A - Alloy C
Alloy B - Alloy C
3.00E-08
2.00E-08
I.OOE-08
0.OOE+00
-I.00E-08
-2.00E-08
-3.00E-08
-4.00E-08
-5.00E-08
10
20
30
40
50
60
Time (Hours)
Figure 4 -
Representative scans showing the galvanically coupled current densities

of the alloy combinations in Ringers solution at 37 ~ over time.
116
0.000
- - - Alloy A - Alloy B
........ Alloy A - Alloy C
-0.050
--Alloy
B - Alloy C
-0.100
~ -o.15o
-0.200
>
-0.250
-0.300
-0.350
l0
20
30
40
50
60
Time (Hours)
Figure 5 - Representative scans showing the galvanically coupled potentials

of the alloy combinations in Ringers solution at 37 ~ over time.
The anodic polarization results of the current study are in agreement with previous
investigations by other researchers documenting a substantially higher breakdown
potential and a larger stable passive region in physiological solutions for 22Cr-13Ni-5Mn
stainless steel (alloy C) as compared to implant quality 316L stainless steel (alloy B)
[6,14]. However, these studies also reported little, if any, hysteresis between the forward
and reverse potentiodynamic scans of 22Cr-13Ni-5Mn stainless steel indicating a high
pitting resistance. Due to the higher scan reversal potential in the current study, which
resulted in driving the potential farther into the trans-passive region, the cyclical
polarization curves for alloy C exhibited a larger hysteresis. These results indicate a
greater susceptibility to pitting corrosion than that noted in the above referenced study.
These observations were further confirmed by post-anodic metallographic examination
revealing pitting on the surfaces of alloy C (Figure 3). Nevertheless, a much heavier
uniform pitting attack was shown in the post-anodic analysis of alloy B. Anodic
polarization results for alloy A showed breakdown potentials and a large hysteresis similar
to those for alloy C. The results shown here are in agreement with the manufacturer's
material data sheets for these alloys [5,11].
It has been suggested that stainless steel to stainless steel galvanic couples are stable
1 17
under static conditions [15]. The results of the zero-resistance galvanic corrosion testing
between each combination of the stainless steels indicated small initial currents for
galvanically coupled samples. The current diminished to a value approaching zero over
the first 24-hours of testing. The galvanicalty coupled potentials stabilized over the first
24-hours of the couple, and maintained a near constant value for the duration of the tests.
These results are in good agreement with a previous study which galvanically coupled
316L stainless steel (alloy B) to 22Cr-13Ni-5Mn stainless steel (alloy C) [ 14]. Under
dynamic conditions leading to accelerated corrosion processes, such as fretting, electrically
coupled components may experience galvanic corrosion effects [ 15]. In addition, studies
have shown galvanic corrosion effects may be accelerated in implant situations where a
small anodic area is coupled to a larger cathodic area. This effect could occur in situations
such as plate and screw combinations, where the screw is anodic with respect to a much
larger cathodic plate.
In conclusion, the results of the potentiodynamic portion of this study indicated that the
corrosion resistance of the low-nickel alloy A is similar to alloy C, and significantly better
than that of alloy B. Galvanic testing revealed little, if any, effects of coupling the
individual alloy combinations. Therefore, the use of mixed combinations of these alloys
appears to be adequate for many implant applications. However, the dynamic effects
encountered during fretting corrosion on these alloy combinations is still a potential
concern and is currently under investigation.
Acknowledgment
This research was supported by a grant from Carpenter Technology Corporation.
References
[1]
[2]
[3]
[4]
[5]
[6]
Disegi, J.A. and Eschbach, L., "Stainless Steel in Bone Surgery," Injury, Vol. 31,
Suppl. 4, 2000, pp. D2-D6.
Zardiackas, L. D., Roach, M., Williamson, S., Bogan, J. A., "Comparison of
Notch Sensitivity and Stress Corrosion Cracking of a Low-Nickel Stainless Steel to
316LS and 22Cr-13Ni-5Mn Stainless Steels," Stainless Steels for Medical and
Surgical Applications, ASTM STP 1438, G. L. Winters and M. J. Nutt, Eds.,
ASTM International, West Conshohocken, PA 2003.
Xulin, S., Atsuo, I., Tetsuya, T., and Hosino, A., "Fretting Corrosion Resistance
and Fretting Corrosion Product Cytocompatability of Ferritic Stainless Steel,"
Journal of Biomedical Materials Research, Vol. 34, 1997, pp. 9-14..
Gebeau, R.C. and Brown, R.S., "Tech Spotlight - Biomedical Implant Alloy,"
Advanced Materials and Processes, Vol. 159, No. 9, 2001, pp. 46-48..
Carpenter Technology Corporation Alloy Data, BioDur 108 Alloy, Carpenter
Technology Corp., Reading, PA. 2000.
Shetty, R.H., Ottersberg, W.H., "Metals in Orthopedic Surgery," Encyclopedic
Handbook of Biomaterials and Bioengineering, Vol. 1, 1995, pp. 509-540.
118
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
Ogundele, G.I., White, W.E., "Polarization Studies on Surgical-Grade Stainless

Steels in Hank's Physiological Solution," Corrosion and Degradation oflmplant
Materials, ASTM STP 859, 1985, pp. 117-135.
Beddoes, J., Bucci, K., "The Influence of Surface Condition on the Localized
Corrosion of 316L Stainless Steel Orthopaedic Implants," Journal of Materials
Science: Materials in Medicine, Vol. 10, 1999, pp. 389-394.
Fraker, A.C., "Corrosion of Metallic Implants and Prosthetic Devices," Metals
Handbook, Ninth Edition, Vol. 13, 1987, pp.1324-1335.
Sedriks, J.A., "Galvanic Corrosion," Stainless Steels, John Wiley and Sons, 1992,
pp. 344-347.
Carpenter Technology Corporation Alloy Data, Carpenter 22Cr-13Ni-5Mn,
Carpenter Technology Corp., Reading, PA., 1987.
Baboian, R., "Electrochemical Techniques for Predicting Galvanic Corrosion,"
Galvanic and Pitting Corrosion - Field and Laboratory Studies, ASTM STP 576,
1976, pp. 5-19.
Black, J., Biological Performance of Materials: Fundamentals of Biocompatibility,
Shetty, R.H., Jacobs, C.H., "Galvanic Corrosion Properties of 22-13-5/316L
Stainless Steel Couple in Physiologic Solution," [abstract] Transactions of the
Society of Biomaterials, Volume 10, 1987, p. 231.
Jacobs, J.J., Gilbert, J.L., Urban, U.M., "Current Concepts Review: Corrosion of
Metal Orthopaedic Implants," The Journal of Bone and Joint Surgery, Vol. 80 (2),
1998, pp. 268-282.
Ilia Tikhovski j, Holger Brauer l, Martina M61ders 2, Martin Wiemann 2, Dieter Bingmann 2,
Alfons Fischer I
Fatigue Behavior and In-Vitro Biocompatibility of the Ni-Free Austenitic HighNitrogen Steel X13CrMnMoN18-14-3
Reference: Tikhovski, I., Brauer, H., M61ders, M., Wiemann, M., Bingmann, D., and
Fischer, A., "Fatigue Behavior and In-Vitro Biocompatibility of the Ni-Free Austenitic
High-Nitrogen Steel X13CrMnMoN18-14-3," Stainless Steels for Medical and Surgical
Applications, ASTMSTP 1438, G. L.Winters and M. J. Nutt, Eds., ASTM International, West
Conshohocken, PA, 2003.
Abstract: Austenitic stainless steels generally have a favorable combination of strength
and ductility as well as a sufficient resistance against corrosion. This and the reported
biocompatibility lead to the use of 304- and later 316 L-type steels in medical
applications. Especially in orthopedics these steels were applied as implants for e.g.
fracture fixation as bone plates, intermedullary nails, and screws. But these steels contain
a high amount of Ni, which was attributed to cause Ni-allergies for an growing amount of
patients. Thus, alternatives were needed and - beside the already known CoCrMo-alloys implants of Ti and its alloys emerged increasingly into the medical market. The aim of
this paper is to introduce a new austenitic Ni-free CrMnMo-steel X 13CrMnMoN 18-14-3
(Material No.: 1.4452, brand nanae: P2000), which makes use of about 1% N in order to
gain a combination of high strength, high ductility, and a superior corrosion resistance.
In a first step the cyclic fatigue behavior in air and in Ringer solution in the solution
annealed state is investigated. This is accompanied by electrochemical testing in Ringer
solution as well as in-vitro cytotoxicity tests against MC3T3 ceils in bovine serum.
The tests revealed that the solution annealed X 13CrMnMoNI 8-14-3 at 5 Hz has an
50 % endurance limit of 346 MPa in air and of 302 MPa in Ringer solution, which is
markedly higher compared to solution annealed CrNiMo-steels. In addition it was found
that the CrMnMoN-steel shows no distinct susceptibility to stress-corrosion cracking in
the entire region of finite life between stress amplitudes of 400 to 550 MPa. The pitting
potential in Ringer solution was measured to be 1.1 V, which is in the range of common
Dipl.-Ing. and Research Scientist, Dipl.-Ing. and Research Scientist, Prof. Dr.-Ing.,
respectively, Werkstofftechnik, Universitaet Essen, Universitaetsstr. 15, 45117
Essen, Germany
2 Research Scientist, PD Dr. and Research Scientist, Prof. Dr., respectively, Institut fuer
Physiologie, Universitaetsklinikum, Universitaet Essen, Hufelandstr. 152, 45122
Essen, Germany
119
www.astm.org
120
Ni-containing high-Nitrogen steels as well as of CoCr20Mo6 alloys. No reduction of

MC3T3 cell adhesion could be observed.
Thus, the Ni-free CrMnMoN-steels might be a promising alternative to the CrNiMosteels in medical applications.
Keywords: High-Nitrogen Steel, Fatigue, Biocompatibility, Cytotoxicity
Nomenclature
ca.t total strain amplitude
ea,pl plastic strain amplitude
em,pJ cyclic mean plastic strain%
AT temperature change
Oa stress amplitude
HNS
MTT
AM
%
%
K
MPa
high Nitrogen steel X 13 CrMnMoN 18-14-3

Methylthiazoltetrazolium
Acetoxymethylester
Introduction
Austenitic stainless steels generally have a favorable combination of strength and
ductility as well as a sufficient resistance against corrosion. This and the reported
biocompatibility lead to the use of 304- and later 316 L-type steels in medical
applications. Especially in orthopedics these steels were applied as implants for e.g.
fracture fixation as bone plates, intermedullary nails, and screws. However, any implant
brings about a reaction within the adjacent tissue and the surrounding media leads to a
reaction of the implant surfaces [1]. Thus, these steels may corrode under combined
chemical-mechanical loads and release their metal ions into the body fluid. Due to the
fact, that they contain a high amount of Ni, this is attributed to cause Ni-allergies for a
growing amount of patients [2-7]. Thus, alternatives were needed. One way is to increase
the chemical stability of passive layers of Cr203-type on steels by adding Cr, Mo, and N
[8, 9]. The other way is to choose materials which have no Ni as alloying element like
most of the CoCrMo- and Ti-base alloys. The latter make use of a passive layer of TiO2type. This is chemically much more stable and has an excellent biocompatibility [10-12].
But, due to the fact that under fatigue and wear all these passive layers will crack there
will be a release of metal ions in any case. Thus, beside the already known Ni-free
CoCrMo-alloys implants of Ti and its alloys emerged increasingly into the medical
market. Parallel Ni-free stainless steels have been developed, which combine a high
strength and an ductility with good corrosion properties [13].
The aim of this paper is to introduce the fatigue properties of solid solution annealed
anstenitic high-Nitrogen Ni-free CrMnMoN-steel X13CrMnMoN18-t 4-3 (HNS, Material
121
TIKHOVSKI ET AL. ON NI-FREE AUSTENITIC HIGH-NITROGEN STEEL
No.: 1.4452), which makes use of about 1 % N in order to gain a combination of high
strength, high ductility, and a superior corrosion resistance. In addition the results of invitro cytotoxicity tests are shown and discussed.
Experimental Procedures
Material
The material investigated was delivered as drawn, solution annealed and ground bars
with a diameter of 30 mm. From these all specimens were cut either by sawing or by wet
cutting. The range of the typical chemical compositions of the high-nitrogen steel is given
in Table 1.
Table 1 - Chemical Composition of X13CrMnMoN18-14-3
Fe
Cr
Mn
Mo
Nb
bal.
16-20
12-16
2.5-4.2
_<0.25
_<0.15
0.75-1.0
Tensile tests at strain rates between 0.016 and 10 %/s (Figure 1) of the solution annealed
state revealed no distinct strain rate sensitivity. For a strain rate of 0.1%/s the mechanical
1200
,
,L
100
900
75
..=
.=.
6O
t~ 600
""
"''''IF'~
50 ...
t..
t_
rJ~
Tensile Strength Rm
rJ~
- 9
300
Yield Strength Rp0.2
..... 25
---*- Elongation A [ % ]
--,t- Reduction of Area Z [ % ]
F
0.016
0.1
0.4
0
10
Strain Rate ~ in %/s

Figure 1 - Tensile Properties of the HNS under different Strain Rates
122
properties are listed in Table 2.
Table 2 - Tensile Properties of X13CrMnMoN18-14-3

yield
strength
in MPa
tensile
strength in
MPa
elongation
to fracture
in %
reduction
of area in
%
590
1030
70
75
Metallography and Microscopy

For the light microscopical analyses of grain size and hardness eight discs of 30 mm
diameter and 5 mm thickness were cut from the bars. These specimens were ground down
to 1200 mesh size SiC followed by polishing with diamond paste (ATM, Altenkirchen,
Germany) down to 6 ~tm. The hardness testing was done according to Vickers with a load
of 98.0665 N leading to HV10 hardness numbers. On every disc up to 30 measurements
were taken along a line for the determination of the mean value and the standard
deviation. The average hardness was measured being 259 14 HV10.
After this the specimens were cleaned and further polished with diamond paste down
to 1 ~tm. Etching was done for 120 s at 55~ in V2A etchant (100 mL distilled water,
100 mL hydrochloric acid, 10 mL nitric acid, 0,3 mL pickling inhibitor). Polishing and
etching cycles were repeated three times in
order to gain a preparation result which is
suitable for automated image analyzing
(Figure 2). From these specimens the grain
size was derived according to ASTM
Standard "Grain Size, Average, Using
Semiautomatic and Automatic Image
Analysis" (E 1382-91) using ImageC
Software (Imtronic GmbH, Berlin,
Germany) attached to a digital camera
DIC 1300 (Micromotion, Landshut,
Germany), which was mounted on a light
microscope METALLUX 2 (Leitz,
Figure 2 - Austenite Grain Structure of the
Wetzlar, Germany). The intersections
Solid Solution Annealed HNS
between the measuring grit and the twin
boundaries were deleted manually in order
to get measuring values solely from the
austenite grain boundaries. This revealed an average grain size of 100 13 Jam. ASTM
grain size ranges from 3 to 4.
For transmission electron microscopy (TEM) small discs of 3 mm diameter and
TIKHOVSKI ET AL. ON NI-FREE AUSTENITIC HIGH-NITROGENSTEEL
123
0.2 mm thickness were cut out (ATM, Altenkirchen, Germany) of the bigger specimens
and ground to a thickness of 0.1 ram. Afterwards the center of these discs were thinned
electrolytically (Tenupol, Struers, Willich, Germany) at 40 V and 100 mA in A8 (950 mL
acetic acid, 50 mL perchloric acid) at 8 to 10~ in until a small hole was generated
suitable for TEM investigations.
Corrosion
Potentiodynamic corrosion tests according to ASTM Standard "Standard Reference

Test Method for making Potentiostatic Anodic Polarization Measurements" (G5-94) at
37~ in Ringer solution showed a pitting potential at 1.1 V. In order to achieve this
cylindrical specimens were ground down to 6 pm diamond paste and afterwards electro
polished. The Ringer solution was continuously irrigated with N so that the O z content
was below 1 mg/L. The resting potential was measured until there was no change of the
potential bigger than 1 mV over a 300 s time period. Subsequently the specimens were
activated at -2 V for 60 s. Then the potentiodynamic test was started at 50 mV below the
resting potential with 600 mV/h until the transpassive region was reached. Afterwards the
potential was decreased with the same speed in order to measure the repassivation
behavior. In comparison a standard 316L-type of steel X2CrNiMo 18-15-2 (Material-No.:
1.4441) was investigated in the solid solution annealed state as well.
Cytotoxicity
In-Vitro Biocompatibility - A total of 30 discs (diameter 5.5 mm, thickness 1+/-0.3)
manufactured from I-INS were electro polished (Polier- und Entgratungselektrolyt, Graul,
75417 M0hlacker, Germany), cleaned in 5% HNO3 (80~ for 2 h), defatted in acetone
(3 times, 10 min each) and ethanol (70%, v/v) and finally heat sterilized at 220~ for 3 h.
All further handling was under sterile conditions. Size matched discs made from copper
(defatted and sterilized) were used as negative control material in some experiments. Size
and shape of all metal discs allowed them to be inserted into moulds of a 96-well cell
culture plate, such that the vessel wall tightly fitted the metal. This was advantageous for
all cell culture testing procedures performed under conditions of direct cell-material
contact.
Cell Line - The well-established osteoblast-like cell line MC3T3-E1 [14] was used
throughout (passage number 5-9). Cells were maintained at 37~ in a humidified 5% CO 2
atmosphere using aMEM (ICN Biomedicals Inc., Aurora, OH, USA) supplemented with
20 % fetal calf serum, 2 mmol/L glutamine, penicillin (5,000 I.U./mL), and streptomycin
(5,000/~g/mL).
Cell Morphology - Live cells were loaded with calcein AM (5 #mol/L for 1 h)
(Molecular Probes Europe, Leiden, The Netherlands), washed once with Hanks balanced
solution (HBS) and observed within the same solution with an upright epifiuorescence
microscope (Olympus Bx50Wi, Hamburg, Germany) equipped with a 20x water
immersion objective [15].
M T T A s s a y - Confluent cultures were trypsinized according to standard methods. To
124
establish representative growth curves, 1,5x 103 freshly trypsinized cells were seeded into
each well of a 96-well plate either in the presence or absence of the testing materials.
Tests were run for 7 days. At the time points indicated, 20/uL o f a Thiazolyl Blue (MTT,
Sigma, Taufkirchen, Germany) stock solution (0,5 mg MTT per mL HBS) was added to
200 #L culture medium. Cells were allowed to form the reduced formazan dye for 2 h
under cell culture conditions. Thereafter, wells were briefly washed with 200 #L HBS
before the fluid was completely withdrawn. Dried cells were kept on the plate until
formazan formed inside mitochondria was dissolved with an acid dimethyl sulfoxide
containing lysis buffer [16]. A volume of 100 ~tL was used to measure the optical density,
(570/630nm) with a Dynatech MRX plate reader (Dynatech Labs., Billingshirt, UK).
Background corrected OD values were converted into absolute cell numbers by means of
a linear standard curve (not shown) obtained from serially diluted MC3T3-E1 cells.
Controls - For positive control we used cell culture tested polystyrene material
(Falcon, Becton Dickinson Labware, Franklin Lakes, N J, USA). For negative controls
wells received 0,05% NaN 3 which induced cell death within 1 day. The latter procedure
was carried out because copper discs could not be used together with the MTT reagent.
Induction of Alkaline Phosphatase (ALP) with Bone Morphogenetic Protein 2 (BMP2) - Confluent MC3T3-E1 cells (ca. 8 x 1 0 4 cells/well) were incubated with BMP-2
(Biochrom, Berlin, Germany, 0.025-1 #g/ml) for 3 days. FCS concentration was reduced
to 1% (v/v) during the incubation period, such that further cell growth and cellular protein
increase was inhibited [17]. At the end of the experiments, cells were washed twice with
Hanks balanced solution and cell bound ALP was determined as described [18] using 4nitrophenyl-phosphate (Sigma, Taufkirchen, Germany) at a pH of 10.3 as a soluble
chromogen for ALP. Color intensity was measured as optical density 405 nm and used as
a measure for cell bound ALP. All tests were run in triplicates.
Fatigue
Round specimens (Figure 3) with cone ends were machined from the bars. The
surfaces were.ground by hand and afterwards electrolytically polished until the surface
roughness reached values of Ra at 0.3 ~tm. The fatigue tests were carried out under force
control with an R-Value of-1 using
a MTS-BIONIX 858 (MTS
Systems Corp., Minneapolis, MN,
USA) at 0.5, 5, and 20 Hz under
laboratory air (20~ and Ringer
solution (37~ Figure 4). The
statistical analyses of the S-~I
curves according to arcsin-VP
brings about a range for the
average endurance limit [19].
In order to measure the stressFigure 3 - Scheme of the Specimens for Axial Fatigue
Testing
. .I. . . .~. .21

. . . . . - ~i=
125
Figure 4 - Test Set-Up for Axial Fatigue Testing in Ringers Solution

strain curves (Figure 5) continuously at 0.5 Hz a strain gauge MTS Extensometer 10 mm
(MTS Systems Corp., Minneapolis, MN, USA) with a measuring range of 2.5 % and a
resolution 0.001% was fixed to the specimens. This revealed the stress-strain hysteresis
of which the total cyclic strain amplitude (Ca,t) as well as the plastic strain amplitude (Ca,p1)
could be derived according to standard methods [20]. In addition the change of
t3
'
".......
Strain
~a,p = Plastic Strain A m p l i t u d e
-----~~-:4
,_12
;i
2gat
~a,t = Total Strain A m p l i t u d e

~m,p = Mean Plastic Strain
era,t = Total Mean Strain
6a = Stress A m p l i t u d e
6m = M e a n Stress
Figure 5 - Cyclic Stress Strain Hysteresis
126
temperature within the measuring length of the specimen was measured in air.
Corrosion
The current density potential curves of HNS and 316L are shown in Figure 6. It is
obvious that the corrosion and the repassivation behavior of the HNS is distinctly better
Figure 6 - Current Density Potential Curves of HNS (1.4452) and
Standard 316L-Type (1.4441)

than that of the 316L type. This can be attributed to the higher content of Mo and N,
which increases the pitting resistance equivalent [9, 21]. Thus, the pitting potential of the
HNS reaches values, which are similar to that of the 0.4 % Nitrogen containing steel
X5CrNiMoN22-10-3 (ISO5832-9) 1.1 [8] and slightly better than that of CoCr29Mo6
alloy at 0.8 V [22]. It should be mentioned here, that under similar conditions Ti-base
materials like cp-Ti and TiA16V4 have a pitting potential above 2 V [22].
Cytotoxicity
Epifluorescence microscopy of calcein-AM loaded MC3T3-E 1 cells revealed no
127
TIKHOVSKt ET AL. ON NI-FREE AUSTENITIC HIGH-NITROGEN STEEL
hints for a cytotoxic effect of HNS when cells were kept on the cleaned and
electropolished material for up to 7 days. The degree of confluence, cell shape, and
appearance of fine processes were all unchanged as compared to cells on positive control
Figure 7 - Live MC3T3-E1 Cells loaded with Calcein-AM were viewed under Water
Immersion on HNS and on Cell Culture Tested Control Material. Morphological
Differences are Missing. Note that some Image Distortion results e.g. from
Scratches and Irregularities o f the Metal Surface. The lower Contrast on HNS is
due to some Light Back Scattering from the Metal Surface
material which was the bottom o f the cell culture tested plastic material (Figure 7).
Comparison of growth curves (Figure 8A) also revealed no difference between cells on
culture tested polystyrol material and cells on HNS. In these experiments, cells were in
the phase of exponential growth as it is requested by the standard "Biologische
Beurteilung von Medizinprodukten. Teil 5: Priifungen auf in vitro Zytotoxizitaet" (DIN
A
O
20"
- [ 3 - positive control
--O-" negative control
.L
0.7
--El--- positive control
'~" 0.6
C
15
0.5
0.4
10
~9 0.3
~ 0.2
O.
.J
'~ 0.1
0
"~
0.2
0.4
0.6
0.8
BMP-2 (pg/ml)
days in culture
Figure 8 - Growth of MC3T3.E1 Cells (A) and Dose Dependant Stimulation of ALP by
BMP-2 (B). HNS: Cells growing directly on HNS Discs completely covering the Bottom
of the Culture Vessel Positive Control: Cells on Cell Culture Tested Polystyrol.
Negative Control: Cells after Addition of O.05% NaN 3 to the Medium.
128
ISO 10993-5, 1999). Treatment with NaN 3, in contrast, led to a complete destruction of
the cells within one day. Based on these in vitro results HNS should be classified as a
fully biocompatible material.
Since HNS may be used as implant material in hard tissue, a typical feature of
MC3T3-E1 cells, namely their inducible differentiation into osteoblasts, was included into
our tests as well. We performed a stimulation with recombinant BMP-2 to induce an
increase in cellular alkaline phosphatase, which is an early marker of osteoblast
differentiation [ 18]. Also, in vivo BMP-2 stimulation is believed to be of large relevance
for bone cell differentiation and clinical studies are currently trader way to use this peptide
for increased bone formation e.g. during fracture healing. To test whether or not the BMP2 signaling pathway was impaired when cells were in intimate contact with HNS we
applied BMP-2 to MC3T3-E1 cells adhering to a I-INS surface. Figure 8B shows that cells
under these conditions could be successfully stimulated with BMP-2. The efficacy of
BMP-2 stimulation was, however, reduced, such that about a 5-fold higher concentrations
of BMP-2 was required to yield the same amount of ALP. Besides the slope also the
shape of the dose response curve was changed (Figure 8B). Further studies have to
elucidate underlying mechanisms, which may encompass changes in receptor expression,
BMP-2 binding to the receptor and/or subsequent intracellular signaling pathways.
Fatigue
Figure 9 shows the S-N curves at 5 Hz of the solid solution annealed HNS in ambient
air and in Ringer solution. The arcsin-V/P method reveals 50% endurance limits at
346 MPa in ambient air and 302 MPa in Ringer solution.
Fatigue life and endurance limit depends strongly on the parameters under which
600
r
X I 3 C r M n M o N 18-14-3
R 2 -- 0 . 9 0 2
__'2-..
550
R = -1; f = 5 H z
500
--o.941
450
400
<
350
Ez A i r
300
~--*
Ix
3~-..~.~---b.
9
[]--~
lx
2x
5X
2X
a----~ 5X
9 R i n g e r solution
250
10 3
104
105
Number of
106
107~
Cycles N
Figure 9 - S-N curves of Solid Solution Annealed HNS in Air and in Ringers Solution
TIKHOVSKI
ET AL. ON NI-FREE AUSTENITIC
HIGH-NITROGEN
STEEL
129
these values were measured. Thus, any comparison between e.g. axial and rotating
bending results is questionable. This is also true for tests run at different frequencies,
which today might range from 0.5 Hz to 100 kHz. Another factor is the reported notch
sensitivity of materials, which will have a distinct influence on the fatigue life of implants
[23]. Thus, in this paper the fatigue life of the investigated HNS is compared only to
literature data of other unnotched metallic implant materials. The evaluation of S-N
curves is just a first step in the investigation of the fatigue behavior of any material. But
due to the fact that they just represent the numbers of cycles to fracture it is important to
notice that fractography as well as the metallographical analyses of the changes within the
microstructure enlight the findings of mechanical testing. In order to get some information
about the behavior of materials before fracture cyclic stress-strain hystereses have been
measured as well as the temperature of the Specimens. Both bring about information on
the dissipation of energy and, therefore, on the amount of plastic deformation before crack
initiation as well as in addition to the compliance of the specimen after crack initiation
[24].
Fatigue in AmbientAir - Figures 10 and 11 show the plastic strain amplitude ea.pl and
0.5
XI3 CrMnMoN 18-14-3
if, in MPa, R = -1, f = 5 Hz, Air
9~
0.4
------~
--
'
0.3
-:
---
"<
9~ 0.2
.....
4'7 5
O.l
425 425
17. . . . . . . . ib
10 ~
101
.~ .-~'....Z
102
. . . ~. . . . . ,
. ....... i 1
103
104
.... : J . .J. . . . . .
105
1061
Number of Cycles N
Figure 10 - Plastic Strain Amplitude e,,pl vs. Numbers o f cycles N o f Solid Solution
Annealed HNS in Air
the temperature change AT plotted vs. the numbers of cycles for constant stress
amplitudes a a, respectively. For a high constant stress amplitude of 525 MPa obviously
ea,p~increases during the first two cycles and reaches a constant value for about 10 cycles.
Afterwards it increases progressively until the specimen fractures. This characteristic
behavior is more or less true for all %,prN curves, even though the numbers of cycles of
constant %,p~increase with decreasing %. In comparing Figures 10 and 11 one can see that
the temperature change, which is a measure of the dissipated plastic work, shows a similar
130
80
[I
I ........
....
~. 60
50
40
10
0
:0~
101
10 2
10 3
10 4
10 s
10 6
N u m b e r of Cycles N
Figure l 1 - Temperature Change AT vs. Numbers of cycles N of Solid Solution
Annealed HNS in Air

behavior. Due to the fact that during cyclic hardening and softening before crack initiation
as well as during the plastic work dissipated within the plastic zone during stable crack
propagation both plots do not allow a precise definition of the crack initiation phase. In
order to achieve this, the electrical resistance of the specimens should have been
measured as well [25]. But nevertheless it is obvious that there is no distinct cyclic
hardening. The constant values of e~p2are brought about by the compressive residual
stresses, which are introduced during the compression phase of the load cycle. In
Figure 12 the mean plastic strain em.p~is plotted versus the numbers of cycles N.
Depending on the stress amplitudes compressive values Ofgm,pl are introduced for ~a
values between 450 and 525 MPa, bringing about a shift in the effective strain amplitudes
to smaller values. If ~, is equal to 425 MPa or smaller no distinct plastic work is
dissipated neither due to introduction of compressive residual stresses nor by tensile
fraction o f the load cycles. Thus, neither C,~pjnor Cm,p~nor AT change distinctly during
fatigue life at 5 Hz. At no stress amplitude levels any decrease of e~,r could be noticed.
Thus, no cyclic strengthening of these steels could be observed.
Fatigue in Ringer Solution - During the fatigue in Ringer solution the mechanical
behavior is mostly influenced after the first cracks have been initiated. Due to the action
o f H+-Ions at the crack tip, the stable crack propagation is accelerated and the fatigue life
as well as the endurance limit decreases. But, for the solution annealed HNS the
detrimental effect of corrosion is not as pronounced as it is for a standard 316L type of
steel under similar conditions [25]. For the HNS the S-N curve is very close to that in
ambient air and the endurance limit is just about 8 % lower. For e.g. 316L under the same
testing conditions the endurance limit in Ringer solution is lowered by 14 % [25]. The
open corrosion potential, which is a measure for the ongoing corrosion processes and
131
1.5
.2
~.
0.5
o
-o.5
-1
-1.5
10~
101
102
103
104
105
106'
Number of Cycles N
Figure 12 - Mean Plastic Strain em,plvs. Numbers of cycles N of Solid Solution
Annealed HNS in Air

increases markedly with existing cracks, stays at a constant value until a crack has been
initiated. Now the characteristic corrosion mechanisms o f HNS play a dominant role. Due
to the fact that N bring about certain surface reactions [21] the concentration of H+-ions is
lowered and, therefore, the embrittlement within the plastic zone at the crack tip does not
take place or at least is not as pronounced as it is for N-free steels.
Evaluation of the Fatigue Behavior - The investigation o f the specimens surfaces
reveals a crack initiation pattern, which is typical for austenitic steels. Small crack
initiation is governed by the orientation o f the sliding bends within the austenitic grains
Figure 13 - Crack Initiation on the

Surfaces of Fatigue Specimens of a Solid
Solution Annealed HNS tested at a Stress
Amplitude of 500 MPa in Ambient Air
Figure 14 - Crack Inith~tion on the

Surfaces of Fatigue Specimens of a Solid
Solution Annealed HNS tested at a
Stress Amplitude of 500 MPa in Ringers
Solution
132
(Figure 13). It has been reported that for N-free 316L steels in Ringer solution just one
crack initiates becoming the main fatigue crack of the specimens [25]. But for the HNS
many cracks initiated in Ringer solution as well (Figure 14) indicating that the influence
of corrosion on crack initiation and stable crack propagation is by far not as pronounced
Figure 15 - Fatigue Fracture of the Solid

Solution Annealed HNS tested at a Stress
Amplitude of 500 MPa in Ringer Solution
Figure 16 - Forced Fracture of the Solid

Solutions Annealed HNS tested at a Stress
Amplitude of 500 MPa in Ringer Solution
as it is for N-free steels. In addition the fatigue (Figure ! 5) as well as the forced fracture
(Figure 16) surface did not show any marked differences compared to those of the
specimens tested in ambient air.
Even though HNS have a low stacking fault energy and tend to strain harden under
static loads much more pronounced than N-free austenitic steels [ 13], under fatigue loads
no cyclic hardening could be observed. Degallaix et al. [26] showed that cyclic softening
is brought about by the reversibility of planar slip while cyclic hardening is promoted by
subcell structures. Thus, depending on the predeformed state and the cyclic stress or strain
amplitudes both types of cyclic behavior could be observed [13, 25]. N brings about a
short range order pinning effect due to the
formation of strongly bound Mo-N pairs
or clusters within solid solution [27].
Thus any dislocations moving through the
lattice must destroy the Mo-N bonding,
which works as a very strong obstacle to
dislocation motion similar to small
precipitates. Now it is trader discussion
whether this or the low stacking fault
energy has the most pronounced influence
on the planarity of dislocation slip.
Nevertheless it might be concluded that
despite their amount of influence both
mechanisms lead to the same direction.
This results in the fact that cyclic
Figure 17 - Dislocation Structure of HNS
hardening is just observed at high strain Fatigued at a Stress Amplitude of 460 MPa for
19040 Cycles in Air
133
amplitudes and low N contents. At a high N-content - like it is with the investigated steel the dislocation structure is characterized by the absence of cross slip [28]. This supports
the authors' findings, which are characterized by a low density of dislocations and
stacking faults as well as their planar distribution even at high stress amplitudes
(Figure 17).
The endurance limits of the solution annealed HNS as well as of other metallic
materials for implants is listed in Table 3. It should be noted, that the measured endurance
limits strongly depend on all parameters (type of cyclic loading, frequency, grain size,
etc.), which were chosen individually.
Table 3 - Endurance Limits of Solution Annealed Metallic Materials for Implants in
Physiological Media
X13CrMnMoN
18-14-3
316L
ISO5832-9
CoCr29Mo6
cp-Ti
Grade 2
TiAI6V4
302
250-320
400-420
200-280
230-280
400-450
this work
[29]
[30]
[29]
[29]
[29]
The HNS X13CrMnMoN18-14-3 shows fatigue properties in physiological media, which

are in the range of other metallic materials being already applied in daily clinical practice.
Conclusion and Outlook
The austenitic high-Nitrogen steel X 13CrMnMoN 18-14-3 has been evaluated in the
solution annealed state as to its corrosion and fatigue properties. In addition cytotoxicity
was tested in-vitro. This brought about the following results:
1. The HNS investigated can be designated being biocompatible in the presence of
MC3T3 bone like cells.
2. The corrosion behavior in Ringer solution is much better than 316L, slightly better
than that of CoCr29Mo6, and comparable to that of N-containing ISO5832-9.
3. The fatigue properties of the I-INS X13CrMnMoN18-14-3 in Ringer solution are
better than that of 316L.
4. In order to understand the fatigue properties better and derive additional data
rotating bending tests in Ringer solution should be carried out as well as constant strain
amplitude tests. Due to the fact that many implants are not applied in the solution
annealed state the investigations should be extended to different degrees of cold working.
5. Concerning the reported [9, 31] excellent tribological behavior of HNS it can be
designated as a material suitable for implants. Before application further animal tests have
to be carried out.
134
Acknowledgments The authors would like to thank VSG Energie- und Schmiedetechnik
GmbH, Essen, Germany for financial sponsorship. We are in debt to Ms. B. Gteising and
Dr.-Ing, S. WeiB, Werkstofftechnik, Universitaet Essen, Germany together with Prof. Dr.Ing. W. Dudzinski, Institute of Materials Science and Applied Mechanics, TU Wroclaw,
Poland for their TEM investigations. Special thanks goes to Prof. Dr. D. Nast-Kolb and
Dr. G. T~iger, Unfallchirurgie, Universitaetsklinik, Universitaet Essen, Germany for their
important information on the future needs and constraints of implants in daily clinical
practice.
References
[ 1] Puleo, D.A., Nanci, A.; "Understanding and Controlling the Bone-lmplant Interface,"
Biomaterials, Vol. 20, 1999, p.2311-2321.
[2] Hildebrand, H.F., Hornez, J.C.; "Biological Response and Biocompatibility," in
Metals as Biomaterials, J. A.Helsen and H. J.Breme, Eds., Wiley & Sons,
Chichester, UK, 1998, p.265-290.
[3] Hildebrand, H.F., Veron, C., Martin, P.; "Ni, Cr, Co Dental Alloys and Allergic
Reactions: An Overview". Biomateriats, Vol. 10, 1989, p. 545-548.
[4] Goh, C.L.; "Prevalence of Contact Allergy by Sex, Race and Age." Contact
Dermatitis, Vol. 14, 1986, p. 237-240.
[5] M611er, H.; "Nickel dermatitis: Problems Solved and Unsolved." Contact Dermatitis,
Vol. 23, 1990, p.217-220.
[6] Peltonen, L.; "Nickel Sensitivity in the General Population." Contact Dermatitis,
Vol. 5, 1979, p.27-32.
[7] Kiec-Swierczynska, M.; "Allergy to Chromate, Cobalt and Nickel in Lodz 19771988." Contact Dermatitis, Vol. 22, 1990, p.229-231.
[8] N.N.; "REX734 High Nitrogen Austenitic Steel", Brochure, SMP Ltd., Sheffield, UK,
1999.
[9] Thomann, U.I., Uggowitzer, P.J.; "Wear-Corrosion Behaviour of Biocompatible
Austentic Stainless Steels" Wear, Vol. 239, 2000, p.48-58.
[10] Breme, H.J., Helsen, J.A.; "Selection of Materials". in Metals as Biomaterials, J. A.
Helsen and H. J.Breme, Eds., John Wiley & Sons, Chichester, UK, 1998, p.1-36.
[ 11] Steinemann, S.G.; "Corrosion of Titanium and Titanium Alloys for Surgical
Implants" in Titanium, Science and Technology (Proc. Conf.) Ltitjering, G., Zwicker,
U., and Bunk, W., Eds., 5th Internation Conference in Titanium, September, 10th 14th, 1984, Munich, Germany, Deutsche Gesellschaft far Materialkunde e.V.,
Hamburger Allee 26, 60486 Frankfurt, Germany, 1984, p.1373-1379.
[ 12] Oshida, Y., Miyazaki, S.; "Corrosion and Biocompatibility of Shape Memory
Alloys." Corrosion Engineering, Vol. 40, 1991, p. 1009-1025.
[13] Gavriljuk, V.G., Berns, H.; "High Nitrogen Steels" Springer, Berlin, Heidelberg,
New York, 1999, p.135-183.
[14] Sudo, H., Kodama, H.A., Amagai, Y., Yamamoto, S., Kasai, S.; "In vitro
Differentiation in a New Clonal Osteogenic Cell Line Derived from Newborn Mouse
Calvaria." Journal of Cell Biology, Vol. 96, 1983, p. 191-198.
TIKHOVSKI ET AL. ON NI-FREE AUSTENITIC HIGH-NITROGENSTEEL
135
[15] Wiemann, M., Winkler, L., Bingmann, D.; "Light Microscopic Methods to Study
Cells on Non-Transparent Materials." Materialwissenschafi und Werkstoffiechnik,
Vol. 32, 2001, p.976-983.
[ 16] Lindl, T.;"Zell- und Gewebekultur" Spektrum Akademischer Verlag, Heidelberg,
Berlin, 2000.
[17] Wiemarm, M., Rumpf, H.M., Bingmann, D., Jermissen, H.;"The Binding ofrhBMP-2
to the Receptors of Viable MC3T3-E 1 Cells and the Question of Cooperativity."
Materialwissenschaft und Werkstoffiechnik, Vol. 32, 2001, p.931-936.
[18] Hazama, M., Aono, A., Ueno, N., Fujisawa, Y.;"Efficient Expression of a
Heterodimer of Bone Morphogenetic Protein Subunits using a BaculoVirus
Expression System." Biophysical and Biochemical Research Communications, Vol.
209, 1995, p.859-866.
[ 19] Dengel, D.; "Planting und Auswertung von Dauerschwingversuchen bei angestrebter
statitischer Absicherung der Kennwerte." in Verhalten yon Stahl bei schwingender
Beanspruchung, Kontaktstudium WerkstofflcundeEisen und Stahl III, Dahl, W., Ed.,
Verlag Stahleisen, Diisseldorf, Germany, 1978, p.23-46.
[20] Suresh, S., "Fatigue of Materials", Cambridge University Press, Cambridge, UK,
1992, p.30-96.
[21] H~aurinen, H.;"Corrosion Properties of HNS." Materials Science Forum, Vols. 318320, 1999, p.479-488.
[22] Williams, D.F.; "Biocompatibility of Clinical Implant Materials. Vol.I" CRC Press,
Boca Raton, Florida, USA, 1981, p.26.
[23] Bensmann, G., "An Attempt to Assess Material Suitability taking the Example of
Hip Endoprotheses." Materialwissenschaft und Werkstofftechnik, Vol. 30, H. 12,
1999, p.733-745.
[24] Middeldorf, K., Harig, H.; "Thermometrische Bewertung mikrostruktureller
~mderung bei zyklischer Beanspruchung von S~hlen." Zeitschriftfar Metallkunde,
Bd.77, H.9, 1986, p.564-570.
[25] G6bbeler, P.;"Untersuchungen zum Ermtidungsverhalten des kaltumgeformten
austenitischen Implantatwerkstoffes X2CrNiMo 18-15-3 - 1.4441" Ph.D.THesis,
Universitaet Essen, Germany, Werkstofftechnik, 1998, see also Fortschritt-Berichte
Verein Deutscher Ingenieure, VDI-Verlag, Dasseldorf Germany, Reihe 5, Nr. 513,
1998.
[26] Degallaix, S., Foct, J., Hendry, A.; "Mechanical Behavior of High-Nitrogen Steels."
Material Science and Technology, Vol. 2, No. 9, 1986, p.946-950.
[27] Murayama, M., Hono, K., Hirukawa, H., Ohmura, T., Matsuoka, S.; "The Combined
Effect of Molybdenum and Nitrogen on the Fatigued Microstructure of 316 Type
Austenitic Stainless Steels." Scripta Materialia, Vol. 4 I, No. 5, 1999, p.467-473.
[28] Zlateva, G., Kemanova, T., Ilieva, M.; "Dislocation Structure of Austenitic Nitrogen
Steel during High Cycle Fatigue." Praktische Metallographie, Vol. 34, H. 10, 1997,
p.508-518.
[29] Breme, H.J., Helsen, J.A.; "Selection of Materials". in Metals as Biomaterials,
Helsen, J. A. and Breme, H. J., Eds., Wiley & Sons, Chichester, UK, 1998, p.l-71.
[30] Kramer, K.H.; "Metallische Implantatwerkstoffe - ein Oberblick" Biomaterialien,
Vol. 2, H.4, 2001, p.187-197.
136
[31 ] Buescher, R., Krause, A., Fischer, A.; "Sliding Wear Behavior of Wet-Chemically
Modified High-Nitrogen Steels". in Lubricants, Materials, and Lubrication
Engineering, Bartz, W.J., Ed., 13th International Colloquium. January 15 - 17, 2002,
Stuttgart/Ostfildern, Germany Technisehe Akademie, Ostfildern, Germany, 2002,
p.1297-1306.
Suzanne H. Parker, 1 Hsin-Yi Lin, 1 Lyle D. Zardiackas, 2 and Joel D. Bumgardner I

Influence of Macrophage Cells on 316L Stainless Steel Corrosion
Reference: Parker, S.H., Lin, H-.Y., Zardiackas, L.D., and Bumgardner, J.D., "Influence
of Macrophage Cells on 3161, Stainless Steel Corrosion," Stainless Steels for Medical
and Surgical Applications, ASTM STP 1438, G. L. Winters and M. J. Nutt, Eds., ASTM
Abstract: In vitro corrosion tests have not evaluated the role of cells on the corrosion of
implant quality 316L stainless steel. A new cell-culture corrosion cell was used to
simulate the clinical condition of cells attached to and growing on the alloy to evaluate
the effects of cells on alloy corrosion and the effect of corrosion products on cells. The
corrosion potential, charge transfer, and surface composition of the alloy were measured
in the presence and absence of macrophage cells (RAW 264.7) or cells stimulated to
release NO over 72 hours. Whereas there was no statistical difference in the corrosion of
316L stainless steel in the presence of macrophage cells as compared to culture media
alone, there was a trend for higher corrosion to occur in the presence of the cells.
Corrosion was further reduced when cells were stimulated to release NO which may have
oxidized the implant and contributed to an enhancement of its surface oxide. These data
suggest that cells may alter alloy surface oxides and affect alloy corrosion.
Keywords: 316L stainless steel, corrosion, surface analysis, biocompatibility, cell

culture, macrophages
Introduction
Implant quality 316L stainless steel (316L SS) is widely used in orthopaedic,
craniofacial and cardiovascular implant devices. The alloy depends on a chromium oxide
surface layer for corrosion resistance and biocompatibility. Nevertheless, when
implanted, the alloy may corrode, releasing constituents of the alloy into the body. For
example, Walczak et al. [1] observed significant amounts of corrosion in 9 of 11 stainless
steel hip implants retrieved from patients after 9-21 years.
While most implanted 316L stainless steel devices perform well, release of
elements, particularly chromium and nickel, may be of concern due to their potential for
IGraduate Student, Graduate Student, and Associate Professor, respectively, Department of

Agricultural and Biological Engineering, Biomedical Engineering Program, Mississippi State
University, MS State, MS, 39762.
2Director, Biomaterials Division, Universityof Mississippi Medical Center, Jackson, MS, 39216.
137
www.astm.org
138
inducing chronic inflammatory, toxicity, hypersensitivity or other reactions over the

lifetime of the implant. Vieweg et al. [2] recently reported extensive fibrosis, foreign
body reactions, and chronic inflammation in 5 of 13 spinal flxator systems implanted in
patients for 10 months. Evaluating the biocompatibility of materials in a rabbit brain
model, Mofid et al. [3] observed the greatest inflammatory responses for 316L stainless
steel wires as compared to titanium, cobalt-chromium or silicone elastomer. Savarino et
al. [4] observed significant increases in serum chromium and nickel levels, and in sister
chromatid exchanges, an indicator of genotoxicity, in circulating lymphocytes for patients
with stainless steel fracture fixation devices as compared to patients without implants.
Several studies have implicated the development of hypersensitivity reactions to nickelcontaining devices, including stainless steel as a cause for device failure and loss [5-6].
In vitro cell culture studies have indicated that components of 316L stainless steel
released as corrosion products, such as chromium and nickel ions or alloy particulate, are
capable of activating cultured macrophage cells to release pro-inflammatory cytokines
and alter the metabolism of cultured cells at sub-cytotoxic levels [7-11]. These data
suggest that alloy corrosion products may stimulate macrophages and induce an
inflammatory response, which may further aggravate corrosion, thereby setting up a
corrosion-inflammation-corrosion-inflammation feed-back loop.
In vitro and in vivo studies have been used widely to assess the corrosion resistance
of stainless steels [12-21]. In vitro corrosion tests have typically used physiological salt
solutions and have shown that while stainless steels, including 316L, exhibit low
corrosion rates, they are susceptible to accelerated corrosion processes such as pitting and
crevice corrosion [12,13,15,18,19]. For devices like spinal fixation systems and bone
plates and screws, relative motion may occur between components, mechanically
removing the protective oxide film and resulting in fretting corrosion [12,17,19].
Furthermore, when proteins are present in the electrolyte or stresses applied, corrosion
resistance of 316L stainless steel may be reduced and corrosion increased [13-17]. In
vivo corrosion tests have shown qualitative correlations with in vitro corrosion tests [1214], however, differences between are apparent. Brown and Simpson [12] observed more
fretting corrosion in vitro with saline solutions than in vivo, which may have been due in
part to differences between in vitro and in vivo loading. While Bundy et al. [13] reported
that stress-corrosion cracking of 316L stainless steel was more severe in vivo than in
vitro, others have disputed these findings, claiming that biological conditions are not
severe enough [18]. Williams et al. [17] later reported that while proteins increased
passive corrosion rates of stainless steel, they decreased in vitro fretting corrosion.
Additional research by Powell [20] reported that the negative effect of Ringers solution
on 316L stainless steel corrosion fatigue was neutralized when 0.25% of clottable
fibrinogen was added to the solution. Hence, it has been speculated that in vitro corrosion
tests with simple salt solutions may not be a true measure of in vivo corrosion, and that
organics, cells, mechanical loading and bioelectric effects may play an important role [13,
21].
It is important that in vitro corrosion tests closely simulate the in vivo environment
to provide accurate data on the in vivo corrosion behavior of implant alloys. To date, in
vitro corrosion tests have not considered the role of cells in the corrosion behavior of
biomedical alloys. Cells may affect alloy corrosion by limiting oxygen diffusion through
PARKER ET AL. ON MACROPHAGE CELLS
139
the release of reactive metabolites, altering dissolution kinetics, and or changing the
surface oxide. Increases in oxide thickness and incorporation of calcium, phosphorous,
and sulfur into surface layers were reported for stainless steel devices in bony or oral soft
tissues implanted in humans for up to 128 days [22]. Therefore, the aim of this study was
to investigate the hypothesis that the corrosion rate of 316L stainless steel is increased
when macrophage cells are attached and growing on the alloy surface. Macrophages were
chosen since they are often observed adjacent to implants, have been reported to be
stimulated by corrosion or degradation products in vitro, and are capable of releasing a
variety of inflammatory agents including nitric oxide (NO). NO is a molecular mediator
of many physiological processes, such as vasodilation, inflammation, and thrombosis,
which are essential for proper wound healing and immune reactions. NO will
spontaneously oxidize in the physiological environment to form nitrite. Due to its highly
reactive nature, NO may also further aggravate corrosion.
316L stainless steel [ASTM Standard Specification for Wrought-18 Chromium-14

Nickel-2.5 Molybdenum Stainless Steel Sheet and Strip for Surgical Implants (F139)]
plates, 4.0cm X 4.0cm X 0.2cm were prepared by wet grinding through a series of SiC
papers up to 1500 grit. The plates were cleaned ultrasonically for five minutes in acetone,
then rinsed with 70% ethanol, and rinsed again with distilled water. The surfaces were
then passivated according to ASTM (F86) Standard Practice for Surface Preparation and
Marking of Metallic Surgical Implants to simulate clinical conditions, sterilized by
immersion in 70% ethanol for two hours and then air dried under a UV light in a sterile
hood.
TIB-71 murine alveolar macrophages (RAW 264.7) were obtained from American
Type Culture Collection, ATCC, (Manassas, VA) and cultured in a humidified 5% CO2
atmosphere at 37C. Cells were cultured in medium specified by ATCC [Dulbecco's
Modified Eagle's Medium (DMEM) supplemented with 1 mM sodium pyruvate, 4 mM
L-glutamine, 1.5 g/L sodium bicarbonate, 100 units/mL penicillin, 100 ~tg/mL
streptomycin, and 0.25 Ixg/mL amphotericin with a final concentration of 10% FBS],
which will be referred to as complete media. For tests requiring stimulated cells (i.e., cells
induced to release NO as in an inflammatory reaction), growth medium supplemented
either with 50 ~tg/ml lipoplysaccharide (LPS: Sigma, St. Louis, MO, serotype 0127:B8) or
5 ~tg/mL LPS plus 0.05 lag/mL interferon-~ ([FN-y: R&D Systems, Minneapolis, MN)
was added to the cells after a 4-hour attachment period [23,24]. The combination of
LPS+IFN-), used was determined from three-day dual dose-response curves (0 LPS + 0.5
~tg/ml interferon to 50 ~tg/ml LPS + 0 interferon) and shown in pilot studies to be
necessary to induce detectable levels of NO to be produced by the cells growing on the
alloy surfaces [24].
Cell-Culture Corrosion Plate
Corrosion tests were conducted in an electrochemical cell culture corrosion cell

designed to simulate the in vivo condition of ceils growing on implant surfaces (Figure
140
1). The bottom of the box consisted of the test

alloy, the sides were composed of virgin electrical
grade polytetrafluoroethylene, and the lid was
plexiglass. The metal plate was held to the bottom
of the box by a polytetrafluoroethylene plate
screwed to the sides of the box. A stainless steel
screw
was
inserted
through
the
polytetrafluoroethylene bottom of the box to make
an electrical contact with the stainless steel
working electrode. Graphite rods (2) were inserted
through ports on the side of the box and served as
auxiliary electrodes. A glass bridge tube with a
Figure 1-Schematic of
Vycor| tip (Princeton Applied Research, Oak
electrochemical cell culture
Ridge, TN) was inserted through a port in the lid to
corrosion cell.
hold the standard calomel reference electrode
(SCE). All corrosion box components, except the
bridge tube and reference electrode, were sterilized by immersion in 70% ethanol for 24
hours. The bridge tube and SCE were rinsed profusely with 70% ethanol and fresh media
immediately before use. For corrosion tests, the corrosion box was placed inside a 5%
CO2, humidified cell culture incubator at 37~ and the reference, auxiliary, and the
working electrode leads were connected to the potentiometer (Princeton Applied
Research, Model 6310, Oak Ridge, TN) by rtmning the leads between the seal and glass
door of the incubator. Cell density for both controls and the cell culture corrosion box
was 1X105 cells/era 2, with a 1:1 ratio of media (mL): area (cm2).
Corrosion Experiments
A series of corrosion experiments were conducted to determine the open circuit
potential (Ecorr) and corrosion current at 0.0 mV vs. SCE over 24, 48 and 72 hour time
intervals in: 1] complete media only, 2] with macrophages attached and growing on the
test alloy, and 3] with macrophages stimulated with LPS or LPS+INF-7. Using the
Princeton Applied Research M352 Corrosion software (version3.04), Eco~rwas monitored
for 4 hours to allow the potential of the system to stabilize, and then the corrosion rate
0tA/cm2) at 0.0V vs. the SCE reference electrode was recorded for 20 hours. The test
sequence was repeated at 24 hour intervals to provide Eeorr and corrosion current data
over a 72 hour period. Each test was repeated at least five times for each test electrolyte
except for LPS stimulated cells, which was run only three times. The M352 software was
used to calculate the charge transfers from the area under the current vs. time curves. For
tests with stimulated cells, the cells were allowed to attach for 4 hours before stimulation
since early tests indicated that seeding cells with LPS in the media inhibited cell
attachment.
After corrosion testing, plates were cleaned and sonicated in methanol to remove
cells, proteins and associated debris. Photomicrographs of freshly polished and anodized
plates and plates after corrosion testing in media, and with cells were qualitatively
141
compared to assess any surface damage due to corrosion by the cells using a JEOL JSM35CF scanning electron microscope.
Evaluation o f Macrophage Cultures
Viability~Morphology - The viability of the cells on the plate after three days of
testing was qualitatively determined using a Leica laser scanning eonfocal microscope
and the LIVE/DEAD| viability stain (Molecular Probes, Inc., Eugene, OR). Live cells
fluoresced green and dead cells fluoresced red. Viability of the cells was qualitatively
compared to control cells grown on glass coverslip slides in six well culture plates.
Cell morphology at the end of the three-day test period was observed via scanning
electron microscope (SEM). The cells attached to the plate were rinsed twice with PBS
for 10 minutes each, fixed with 2.5% glutaraldehyde in 0.1M potassium phosphate buffer
(KPB) supplemented with 0.1M sucrose at pH=7.0 for 30 minutes, rinsed twice in 0.1M
KPB for 30 minutes each, dehydrated through a series of alcohol washes, washed twice in
hexamethyldisilazane, HMDS (Polysciences, Inc., Warrington, PA), for 10 minutes each,
and then allowed to air dry in a desiccator. The cells were sputter-coated with Au/Pd for
three minutes in a Polaron E5100 sputter coater and viewed from various tilt angles in the
SEM operated at 15 kV beam voltage. The morphology of the cells was qualitatively
compared to control cells grown on glass coverslip slides.
N O Production via Nitrite Quantitation - Stimulation of cells growing on the test
alloy in the electrochemical cell culture corrosion plate was measured at 24-, 48-, and 72hour intervals by the Griess reagent kit (Molecular Probes, Inc, Eugene, OR) using a
taQuant plate reader (Biotek Instruments, Inc.,Winooski,VT). The detection limit for this
test is 0.5 ~tM nitrite. Three 150 ~tl aliquots of the cell supernatant were collected from
the plate on each day of the three-day tests. Each aliquot was combined with 130 txL of
distilled water and 20 ~tL of Griess reagent in a 96-well plate. After incubating for 30
minutes at room temperature, the absorbance of each well was read at 548 nm.
Absorbanee values were converted to ~tM nitrite concentration. Cells grown in glass Petri
dishes with the same growth area as the corrosion plate were used as controls.
Surface Analyses - The chemical composition of the surface oxide of the stainless
steel samples in the polished and anodized condition and after corrosion testing in media,
with cells and with LPS+INF-y stimulated cells was determined via X-ray photoelectron
spectroscopy, XPS. Experiments were conducted with a Physical Electronics Model 1600
surface analysis system (Eden Prairie, MN) with Mg Ks radiation (1253.6 eV) at 200 W.
The analysis area was a circle approximately 800 Jam in diameter. XPS spectra were
referenced to the peak position of adventitious C. Survey spectra were obtained using a
pass energy of 46.95 eV. High resolution spectra were acquired using a pass energy of
23.5 eV. Two spots were evaluated on each sample and two samples per test condition
were evaluated.
142
Statistical Analyses
Factorial analysis o f variance was used to determine if statistical differences existed
in Eeorr and charge transfers for the alloy in media, with cells and with stimulated cells at
the 24-, 48- and 72- hour test periods. When analysis of variance tests detected
differences, least significant difference (LSD) multiple comparison procedures were used
to determine where statistical significances existed. Statistical analyses were conducted
using Statlets v.l.lB at the 95% significance level.
Results
Corrosion Analyses
The Ecorr values of 316L stainless steel in media, with macrophages and with
stimulated macrophages over the three-day test period, are shown in Table 1. Factorial
analysis of variance indicated that there were no significant interactions between the test
conditions and time periods (p=0.9). Differences were detected in Er values over time
(p=0.0001) but not between conditions (p=0.2) in post-hoc individual analyses. LSD
comparisons indicated that for each test condition, there was a significant ennoblement of
the Ecorr potential o f the alloy from the 24 hour period to the 48 hour period (p<0.05), but
not from 48 to 72 hours (p>0.05).
Table 1- Ecorr potential (mV vs. SCE) of 316L stainless steel with and without
macrophage cells.
Condition
24 Hr
48 Hr
72 Hr
Media
Media+cells
Media+ cells stimulated with LPS
Media + cells stimulated with
LPS+INF),
-183.7:t:23.3
-158.6
-161.7
-174.4
-96.3
-118.0
-113.8
-129.4
-93.6
-105.9
-102.7
-121.8
b
c
d
a
b
c
Letter superscripts indicate statistically similar values within test conditions. There were no differences
between test conditions.
Corrosion current versus time curves at 0.0 mV vs. SCE reference in all three test
electrolytes and at each test period were very similar; initial currents rapidly decayed to
very low values (Figure 2). The average calculated charge transfers at each time period in
each test condition and the total cumulative charge transfers calculated are shown in
Table 2. Factorial analysis of the charge transfer data indicated that there were no
significant interactions between the test conditions and time (p=0.3). Analyses detected
statistical differences in the charge transfers over time (p=0.0001) and between test
conditions (p=0.0007). LSD comparisons indicated that for each test electrolyte
condition, there was a significant decrease in the charge transfer values from the 24 hour
period to the 48 hour period (p<0.05), but not from 48 to 72 hours (p>0.05). For each day,
only charge transfers in media with LPS+IFN-7 stimulated cells were different from the
other test conditions (p<0.05). Sum total charge transfers in the presence of cells and LPS
143
stimulated cells were greater than with LPS+IFN-7 stimulated cells (p<0.05), but there
were no differences between the media test condition and those with cells (p>0.05).
current (uA)
2.50
2.00
1,50
1.00
0.50
0.00
. . . .
~11 ,3b,I,
,I, ,[IL'II " ' 1 1 tf, II

50
60
70
103~
Figure 2: Current versus time for macrophage cells
growin on 316L stuface on day 1. Curves were similar
for days 2 and 3 and for the other test conditions.
Time (s X
SEM Analyses
Surfaces o f polished and anodized samples exhibited only fine scratches due to the
polishing procedure. After removal o f non-stimulated and stimulated cells from the test
plates, some protein material remained. However, no evidence of surface deterioration
was observed.
Table 2- Charge Transfer (mC) for 316L stainless steel in the presence and absence of
macrophage cells.
Condition
Media
Media+macrophage cells
Media + macrophage cells
stimulated with LPS
Media + macrophage cells
stimulated w/LPS+IFN-7
Dayl
12.24
]
13.01
[
13.69+2.68 d
Day2
Day3
Total
3.64
] 3.02
] 18.914-5.92 ~^
6.41
b I 5.18
b [ 24.60
5.90-+0.48c J 5.048+0.23 ~ J 24.63_+2.55
5.81
3.40
2.24
10.73+2.12 #
Letter superscripts indicate statistically similar values over time for an individual condition
Brackets,l, indicate statistically similar values between conditions at each time period
Symbols indicate statistically similar values for total charge transferred over the three-day test period
144
Cell Culture Analyses

Visual observations using the confocal microscope indicated that the viability of the
non-stimulated macrophage cells attached and growing on the stainless steel alloy in the
corrosion plate for three days was high and indistinguishable from the viability of control
cells attached and growing on glass coverslips. Nor were differences observed between
the morphology of non-stimulated cells grown in the corrosion plate and on the glass
coverslips seen in the SEM (Figure 3a, b). Non-stimulated cells cultured on glass slips
and 316L SS plates appeared to be large and round with rough exteriors in confluent
layers. However, when cells attached to 316L alloy in the corrosion plate or the glass
coverslips were stimulated with LPS, the numbers and viability of the cells remaining
after three days was obviously less than in the non-stimulated condition. However, the
number and viability of the stimulated cells on the glass coverslip were greater than on
the alloy based on visual observations. The morphology of the stimulated cells on the
alloy and glass coverslips were similar to each other but different from non-stimulated
cells. The stimulated cells exhibited a smooth, ball-shaped morphology with few
attachments in sparsely populated patches (Figure 3c, d). When the cells were stimulated
with LPS+IFN-7, number and viability of cells appeared to be greater than with LPS
alone, particularly on the 316L stainless steel surfaces, though numbers of cells still
appeared to be less than nonstimulated cells. Morphology of the LPS+IFN-7 stimulate
cells were similar to those of the LPS stimulated cells (Figure 3e, f).
NO Production - The results of the tests for the production of NO by the cells are
shown in Table 3. Factorial analysis indicated that there were significant interactions
between the condition of the cells, the growth surfaces, and the time period for the
production of NO by the cells (p=0.0001). For glass surfaces, LPS and LPS+IFN-t,
stimulated cells resulted in higher levels of NO production than the nonstimulated cells.
Levels of NO production were greater on Days 2 and 3 than on Day 1 for the LPS
stimulated cells on glass. For ceils grown on the stainless steel surfaces, there was no
difference in the levels of NO produced by nonstimulated and LPS stimulated cells, and
these levels were significantly lower than for the LPS+IFN-7 stimulated cells. Only for
the LPS stimulated cells were the levels of NO production different between the control
glass and stainless steel surfaces.
Table 3- NO release (#M) by macrophage cells on lass and 316L stainless steel
Condition
Surface Dayl
Day2
Day3
Macrophage cells
Glass^
0.6
0.9~:0.2 0.6
0.4
1.8~:2.2 0.8
316L^
Macrophage cells stimulated w/LPS
Glass
13.7+1.3 36.8
50.9
1.5
1.4
1.4
316L^
Macrophage cells stimulated w/LPS+IFN- 7 Glass~
7.0
12.8
13.6
316L#
5.4
! 0.2
11.3
Symbolsindicate similarities betweenconditionsand letters indicate differencesovertime
Figure 3- Photomicrographs of macrophage cells. Cells exhibit normal

rough~textured morphology with many attachments on (a) 3161, stainless steel
and (b) glass coverslip. Cells stimulated with LPS on (c) 316L stainless steel
and (d) glass or with LPS +IFN-7 on (e) 316L stainless steel and 09 glass
exhibited smooth ball shaped morphology with few attachments. The number,
viability, and production of NO by macrophage cells on 316L stainless steel was
greater when stimulated with LPS+IFN-7 than when stimulated with LPS alone.
145
146
XPS Surface Analyses - Representative spectra of the surfaces of the stainless steel
samples are shown in Figure 4. These spectra indicate that the surface of as-polished and
passivated alloy was composed primarily of Cr and O, as chromium oxide/hydroxide with
minor levels of Fe. After 3 days in media, sufficient protein deposition occurred to mask
alloy surface elements as indicated by the large C and N peaks. After 3 days with
macrophage cells or cells stimulated with LPS+IFN-% chromium oxide/hydroxides, as
indicated by the large shoulders at 533-530 eV for O and 580-577 eV for Cr, were
observed to be enhanced on the alloy's surface as well as other organic compounds.
While the N detected on the surfaces was attributed primarily to proteinacious material
for the cell culture media and cells, contributions to this peak from the nitric acid
passivation treatments and from production of NO from the cells is possible.
C(A)
Cr2o
Media
0(,~ Fe
01s
Cls
....
3
"
lw
1000
800
600
400
200
BindingEnergy(eV)
Figure 4,4."XPS analyses of 316L stainless steel after three days in culture media,
with TIB macrophage cells and with TIB macrophage cells stimulated with
LPS +IFN-7. As-polished and nitric acid passivated samples are also shown.
Representative survey spectra [.4] shows major peaks. The '(A) ' designation
indicates an Auger peak in the spectra.
I
|
/
c t oxJdes
hydroxides
and n~rates
Cr oakJes
' e $ Cr-carbonate '
hydroxid
Cr-hy~docarbons
and s a b
'
I
I
I
147
carDor~tcs,
and hydrox/~s
organics
_~td
oxides
water
Media
TIB
AS polished
595
590
585
580
575
570
BE(eV)
540
535
53O
525
BE(eV)
protelnac~ous
404
402
400
398
3~5
BE(eV)
Figure 4B-C: XPS analyses of316L stainless steel after three days in culture
media, with TIB macrophage cells, and with TIB macrophage cells stimulated
with LPS +IFN-7. As-polished and nitric acid passivated samples are also shown.
Representative high resolution spectra are shown for chromium [B]; oxygen IC],9
and nitrogen [1)].
148
Discussion
In this study, the Eeorr o f the alloy became more electropositive over the 3-day test
period though there were no statistical differences between tests with or without cells.
Similarly, corrosion rate measurements indicated that there was little difference in the
amount of charge transferred when macrophage cells or LPS stimulated cells were
attached and growing on the alloy's surface as compared to the media only test condition
and that corrosion significantly decreased after the first 24 hours for all test conditions.
These data suggest, at least initially, that when exposed to a cell culture environment,
changes in the alloy's corrosion behavior are due primarily to interactions of the media
with the alloy and not to the presence of cells. The ennoblement of stainless steel's Eeorr
in an aqueous environment has been attributed to the development o f a chromium rich
oxide layer that thickens with time [ 16,25]. The XPS data in this study also demonstrated
the presence of a chromium rich surface oxide particularly in the presence of the cells and
LPS+IFN-y stimulated cells. Thickening and decreasing crystallinity of surface oxides
and incorporation of calcium have been observed for austenitic stainless steel devices
retrieved after 20-136 days, which, at least for titanium implant alloys, have been reported
to reduce corrosion [14,22,26,27]. On the other hand, loss of surface oxide and significant
corrosion were demonstrated for austenitic stainless steel femoral components retrieved
after 9-21 years [1].
While there was a trend for increased corrosion in the presence of non-stimulated
cells as compared to media only, corrosion was still low, indicating that the alloy
remained in the passive state and was well tolerated by the cells. The non-stimulated cells
exhibited normal morphology on the stainless steel, and were observed to have a high
viability similar to controls. No evidence of surface damage or corrosion was observed.
These observations agree with the clinical data indicating that 316L stainless steel devices
are generally well tolerated during short-term implantation. The interpretation, though, on
the long-term corrosion behavior of the alloy due to the cells is uncertain because these
data were collected only from short test periods. The increase in corrosion may not be of
significance within the three-day period used in these tests, but even small increases in
corrosion and subsequent metal ion release may quickly become clinically relevant over
the long term. Increases in Ni blood levels were reported in patients with stainless steel
devices, which were associated with increased genotoxicity [4] and have been implicated
in the induction of metal hypersensitivity reactions [5,6]. Small increases in the release of
Ni ions from stainless steel also were observed in the presence of cultured cells as
compared to media without cells [24]. Thus, the small increases in corrosion in the
presence of the macrophage cells may lead to clinically relevant levels of released metal
ion corrosion products. Hence, hypotheses regarding the role of cells on corrosion of
316L stainless steel still need to be investigated.
What was initially surprising was that corrosion of the alloy was not more affected
when the macrophage cells were stimulated with LPS as compared to non-stimulated
cells. The 50 lag/ml concentration of LPS used to stimulate the cells was determined from
previous reports in the literature and dose-response curves generated in our lab with the
cells in tissue culture plastic dishes [23]. This concentration gave a maximal NO
production by the cells. However, it was noted that the number and viability of the
149
macrophages on the test samples, and the levels of NO produced were greatly reduced
when LPS was added to the medium to stimulate them. The reduced number of cells
remaining on the alloy surface then would have limited any effect on the alloy's corrosion
behavior, and thus appeared to have affected corrosion in a manner similar to nonstimulated cells. For the control LPS stimulated cells, part of the reduction in cell number
and viability may be attributed to the cumulative toxic effects of NO produced and
released during the three day cultures. However, the LPS stimulated cells on the alloy
produced little if any NO during the tests, suggesting that NO may not account for the
toxic effects. It is not clear why the stimulation of cells growing on the alloy surface lead
to decreases in viability and inhibition of NO production as compared to stimulated
controls. One hypothesis may be that macrophages attach to the alloy surface differently
than to tissue culture plastic or glass and therefore do not receive the contact stimulus
necessary for full stimulation. It has been demonstrated that there are differences in
attachment of cells, e.g. osteoblasts, to different surfaces [28]. Differences in the
absorption of proteins and their conformations between the metal and cell culture dish
surfaces may have further affected cellular responsiveness. It has been reported that
fibrinogen coated surfaces caused macrophage cells preferentially to release superoxide at
the basal location of cells at the cell-material interface while albumin coated surfaces did
not [29]. If the macrophage cells in this study preferentially released NO at the basal
location of the cell at the cell-alloy interface, NO may have become incorporated into the
alloy surface film instead of accumulating in the media. Unfortunately, due to the large
amount of proteinacious material on the surface, we have not been able to determine if
.any of the N in the alloy surface was due to NO from the cells. Nevertheless, changes in
the attachment of the cells due to the alloy surface and or absorbed proteins may have
altered their responsiveness to LPS stimulation. Changes in macrophage responsiveness
to antigen alone at the cell-material interface may have negative implications regarding
implant associated bacterial infections.
In vivo, activated T-cells produce INF-y to aid in macrophage activation [30].
Pruett and Higginbotham showed that a combination of LPS+IFN-y was required to
stimulate the RAW 264.7 macrophages to produce maximal amounts of NO [31]. INF-y
was not initially included in this study because ceils in culture wells when stimulated with
LPS alone produced NO. Titration tests in our lab determined that a combination of 5
~tg/mL LPS+ 0.05~tg/mL IFN-y stimulated the cells to release detectable levels of NO
without causing significant cytotoxicity [24]. When cells grown in cell culture dishes and
on stainless steel samples were stimulated with LPS+IFN-y, they produced NO in
measurable quantities. The number and viability of the cells on the stainless steel samples
also appeared greater than when stimulated only with LPS. The production of NO by the
cells on the stainless steel though, resulted in statistically lower corrosion as compared to
the non-stimulated cells and media only conditions. It is not clear as to the mechanism or
how the corrosion would be reduced. It may be that NO and other reactive oxygen species
produced by the cells such as H202 and 0 2 may have contributed to an enhancement of
the alloy's surface oxide, as observed in the XPS analyses. This effect may be responsible
for the in vivo observations that corrosion and metal ion release rates of alloys are high
upon initial implantation, but then quickly subside soon after resolution of the initial
inflammatory/wound healing process [32]. The reactive oxygen species released by
150
macrophage cells close to the implant surface may facilitate the development of the
implant's surface oxide and lower corrosion. The magnitude of this effect may depend in
part on the type of the tissues and time in which the device is implanted. Sundgren et al.
[22] noted differences in the thickness of the surface oxide of 316L stainless steel after
short implantation periods in bone, bone marrow and soft tissues. For implants in bone
marrow and soft tissues, oxide thickness increased 1.5 to 4 times that of the un-implanted
controls and those in bone. They attributed the increase in surface oxide in the marrow
and soft tissues to increased metabolic activity and inflammatory processes. These results
support their hypothesis.
While this study supports the hypothesis that macrophage cells play a role in alloy
corrosion, it was limited due to the short three-day test time frame. This time frame was
initially chosen based on the ability to grow and maintain the cells in a monolayer under
standard culture conditions. It is possible, and experiments are being planned, to increase
from three days to several weeks the time frame the cells remain in culture on the test
alloy samples through adjustment of culture conditions. Thus, additional data on the role
of cells on the corrosion behavior of the alloy may be gained and used to further assess
the long term cell-implant alloy interaction. Furthermore, it may be possible to include
other factors such as mechanical loading, different surface treatments and roughness with
and in addition to other cell/tissue types to more closely simulate the clinical condition.
This may provide additional insights into alloy biocorrosion processes in vivo.
Conclusion
The results of this study indicated that the corrosion of 316L remained low in the
presence of macrophage cells and was further reduced when cells were stimulated to
release NO. NO and other reactive oxygen species can oxidize the implant and may have
contributed to an enhancement of its surface oxide. These data suggest that the
macrophage cells may change alloy surface oxides and affect alloy corrosion behavior.
Given the trend for increased corrosion in the presence of the cells, and the potential for
clinically relevant levels of metal ion corrosion products to accumulate, the investigation
of hypotheses regarding the role of cells on corrosion of 316L stainless steel should
continue.
Acknowledgments
Synthes (USA), Paoli, PA, for providing 316L samples. This work was supported by a
grant from The Whitaker Foundation.
References
[l]
[2]
Walczak, J., Shahgaldi, F., and Heatley, F.,
"In Vivo Corrosion of 316L

Stainless-Steel Hip Implants: Morphology and Elemental Compositions of
Corrosion Products," Biomaterials, Vol. 19, 1998, pp. 229-237.
Vieweg, U., van Roost, D., Wolf, H.K., Schyma, C.A., and Schramm J.,
"Corrosion on Internal Spinal Fixator System," Spine, Vol. 24, 1999, pp. 946-951.
[3]
151
Mofid, M.M., Thompson, R.C., Pardo, C.A., Manson, P.N., and Wander-Kolk,
C.A., "Biocompatibility of Fixation Materials in the Brain," Plastic
Reconstructive Surgery, Vol. 100, 1997, pp. 14-21.
[4]
Savarino, L., Stea, S., Granchi, D., Visentin, M., Ciapetti, G., Donati, M.E., Rollo,
G., Zinghi, G., Pissoferrato, A., Montanaro, L., and Toni, A., "Sister Chromatid
Exchanges and Ion Release in Patients Wearing Fracture Fixation Devices,"
Journal of Biomedical Materials Research, Vol. 50, 2000, pp. 21-26.
[5]
Christiansen,K., Holmes, K., and Zilko, P.J., "Metal Sensitivity Causing
Loosened Joint Prostheses," Annals of Rheumatic Diseases, Vol. 38, 1979, pp.
476-480.
[6]
Kerosuo, H., Kullaa, A., Kerosuo, E., Kanerva, L., and Hensten-Pettersen, A.,
"Nickel Allergy in Adolescents in Relation to Orthodontic Treatment and Piercing
of Ears," American Journal of Orthodontic Dentofacial Orthopaedics, Vol. 109,
1996, pp. 148-154.
[7]
Wang, J.Y., Wicklund, B.H., Gustilo, R.B., and Tsukayama, D.T., "Titanium,
Chromium and Cobalt Ions Modulate the Release of Bone-associated Cytokines
by Human Monocytes/Maerophages In Vitro," Biomaterials, Vol. 17, 1996, pp.
2233-2240.
[8]
Haynes, D.R., Boyle, S.J., Rogers, S.D., Howie, D.W., and Vernon-Roberts, B.,
"Variation in Cytokines Induced by Particles from Different Prosthetic Materials,"
Clinical Orthopaedics, Vol. 352, 1998, pp. 223-230.
[9]
Charissoux, J.L., Najid, A., Moreau, J.C., Setton, D., and Rigaud, M.,
"Development of In Vitro Biocompatibility Assays for Surgical Materials,"
Clinical Orthopaedics, Vol. 326, 1996, pp. 259-269.
[10] Bumgardner, J.D., Doeller, J., and Lucas, L.C., "Effects of Nickel-based Dental
Casting Alloys on Fibroblast Metabolism and Ultrastructural Organization,"
[11] Schedle, A., Samorapoompichit, P., Fureder, W., Rausch-Fan, X.H., Franz, A.,
Sperr, W.R., Sperr, W., Slavicek, R., Simak, S., Klepetko, W., Ellinger, A.,
Ghannadan, M., Baghestanian, M., and Valent, P., "Metal Ion-induced Toxic
Histamine Release from Human Basophils and Mast Cells," Journal of
Biomedical Materials Research, Vol. 39, 1998, pp. 560-567.
[12] Brown, S.A., and Simpson, J.P., "Crevice and Fretting Corrosion of Stainless
Steel Plates and Screws," Journal of Biomedical Materials Research, Vol. 15,
1981, pp. 867-878.
[13] Bundy, K.J., Marek, M., and Hochman, R.F., "In Vivo and In Vitro Studies of
the Stress-corrosion Cracking Behavior of Surgical Implant Alloys," Journal of
[14] Shih, C-.C., Lin, S-.J., Chung, K-.H., Chen, Y-.L., and Su, Y-.Y., "Increased
Corrosion Resistance of Stent Materials by Converting Current Surface Film of
Polycrystalline Oxide into Amorphous Oxide," Journal of Biomedical Materials
Research, Vol. 52, 2000, pp. 323-332.
152
[ 15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
Milosev, I., and Strehblow, H-.H., "The Behavior of Stainless Steel in

Physiological Solution Containing Complexing Agent Studied by X-ray
Photoelectron Spectroscopy," Journal of Biomedical Materials Research, Vol. 52,
2000, pp. 404-412.
Omanovie, S., and Roscoe, S.G., "Interracial Behavior ofbeta-Lactoglobulin at a
Stainless Steel Surface: An Electrochemical Impedance Spectroscopy Study,"
Journal of Colloidal Interface Science, Vol. 227, 2000, pp. 452-460.
Williams, R.L., Brown, S.A., and Merritt, K., "Electrochemical Studies on the
Influence of Proteins on the Corrosion of Implant Alloys," Biomaterials, Vol. 9,
1988, pp. 181-t86.
Sheehan, J.P., Morin, C.R., and Packer, K.F., "Study of Stress Corrosion Cracking
Susceptibility of Type 316L Stainless Steel In Vitro", Corrosion and degradation
of implant materials STP 859, A.C. Fraker, and C.D. Griffin, Eds., American
Society for Testing and Materials, West Conshohocken, PA, 1985, pp. 57-72.
Xulin, S., Ito, A., Tateishi, T., and Hoshino, A., "Fretting Corrosion Resistance
and Fretting Corrosion Product Cytocompatibility of Ferritic Stainless Steel,"
Powell, J. Effects of Fibrinogen on 316L Corrosion Fatigue, MS Thesis, Ohio
State University, 1979.
Clark, G.C.F., and Williams, D.F., "The Effect of Proteins on Metallic
Corrosion," Journal of Biomedical Materials Research, Vol. 16., 1982, pp. 125134.
Sundgren, J.E., Bodo, P., Lundstrom, I., Berggren, A., and Hellem S., "Auger
Electron Spectroscopic Studies of Stainless Steel Implants," Journal of
Shanbhag, A.S., Macaulay, W., Stefanovic-Racic, M., and Rubash, H.E., "Nitric
Oxide Release by Macrophages in Response to Particulate Debris," Journal of
Parker, S.H., "Evaluation of the Effects of Murine Macrophage Cells on the
Biocorrosion of Two Implant Alloys", MS Thesis, Mississippi State University,
2001.
Pan, J., Theirry, D., and Leygrak, C., "Electrochemical and XPS Studies of
Titanium for Biomaterial Applications with Respect to the Effect of Hydrogen
Peroxide," Journal of Biomedical Materials Research, Vol. 28, 1994, pp. 113-22.
Montague, A., Merritt, K., Brown, S., and Payer, J., "Effects of Ca and H2Oz
Added to RPMI on the Fretting Corrosion of Ti6A14V," Journal of Biomedical
Hazan, R., Brener, R., and Oron, U., "Bone Growth to Metal Implants is
Regulated by their Surface Chemical Properties," Biomaterials, Vol. 14, 1993, pp.
571-74.
Schmidt, C., Ignatius, A.A., and Claes, L.E., "Proliferation and Differentiation
Parameters of Human Osteoblasts on Titanium and Steel Surfaces," Journal of
[29]
153
Tang, L., and Jiang, W-.W., "The Role of Adsorbed Proteins on the Persistence of
Implant-associated Bacteria," Transactions of the Annual Meeting of the Society
for Biomaterials, 2001, p. 1.
[30] Janeway, C.A., and Travers, P., "Macrophage Activation by Armed CD4 TH1
Cells," lmmuno-Biology: the immune system in health and disease. Garland
Publishing, New York: Garland, 1997, p. 737.
[31 ] Pruett, S.B., Higgenbotham, J.N., and Lin, T.L., "Effect of Macrophage
Activation on Killing of Listeria monocytogenes. Roles of Reactive Oxygen or
Nitrogen Intermediates, Rate of Phagocytosis, and Retention of Bacteria in
Endosomes," Clinical Experimental Immunology, Vol. 88, 1992, pp. 492-498.
[32] Black, J., "Host response: Biological Effects of Implants," Biological
Performance of Materials - Fundamentals of Biocompatibility, 2ndedition,
Marcel Dekker, Inc., New York, 1992, pp. 125-147.
Lyle D. Zardiackas l, Michael Roach 1, Scott Williamson 1, and Jay-Anthony Bogan l

Comparison of Notch Sensitivity and Stress Corrosion Cracking of a Low-Nickel
Stainless Steel to 316LS and 22Cr-13Ni-5Mn Stainless Steels
Reference: Zardiackas, L. D., Roach, M., Williamson, S., and Bogan, J. -A.,
"Comparison of Notch Sensitivity and Stress Corrosion Cracking of a Low-Nickel
Stainless Steel to 316LS and 22Cr-13Ni-SMn Stainless Steels," Stainless Steels for
Abstract: Recently, low-nickel stainless steels have been developed, in part due to
concerns over patient hypersensitivity reactions. These alloys have been investigated and
found to provide excellent mechanical and corrosion properties. This study compares the
mechanical properties, notch sensitivity, and stress corrosion cracking (SCC)
susceptibility of one such alloy, BioDur | 108, to 316LS and 22Cr-13Ni-5Mn stainless
steels which have a long successful implant history. BioDur | 108 was found to have a
tensile strength, yield strength, and notch tensile strength similar to 22Cr-13Ni-5Mn, and
both had higher values of these properties as compared to 316LS. In addition, BioDur|
108 had the largest percentage elongation during tensile testing of the three alloys but
lower reduction of area than 316LS. No evidence of SCC mechanisms was revealed in
either the test results or fracture analysis on any of the materials tested. With these
material properties and its very low nickel content, BioDur | 108 may be a useful
alternative for certain medical implant applications.
Keywords: stainless steel, stress corrosion cracking, notch sensitivity
Introduction
High strength and good corrosion resistance have proven austenitic stainless steels to
be useful for biomedical implant applications. Significant amounts of nickel are often
added to these steels to stabilize the austenitic microstructure. Although less than 5% of
the patient population have experienced metal sensitivity reactions due to implants, nickel
does account for approximately 90% of the clinically observed sensitivity reactions of
metals [1]. In part as a response to concerns over such sensitivity reactions, several low
~Professor and Coordinator of Biomaterials and Professor of Orthopaedic Surgery,
materials engineer, senior materials engineer and senior materials engineer,
respectively, University of Mississippi Medical Center, School of Dentistry/
Biomaterials, 2500 North State Street, Jackson, MS, 39216-4505.
154
www.astm.org
ZARDIACKAS ET AL. ON NOTCH SENSITIVITY AND STRESS CORROSION
155
nickel stainless steels have been developed. One such alloy, BioDur | 108 from Carpenter
Specialty Alloys, contains very low nickel (<0.05%) and uses a high nitrogen content
(>0.90%) to stabilize its austenitic microstructure. This high nitrogen content has also
been shown to strengthen the alloy as well as contribute to its corrosion resistance [2 -4].
Most studies that have documented susceptibility of austenitic stainless steels to stress
corrosion cracking (SCC) have been performed in high-chloride content, caustic, or acid
environments and/or at elevated temperatures. However, some studies have hypothesized
chloride SCC to be at least a partial failure mechanism in stainless steel medical devices
in vivo [5-7]. These hypotheses have led to a great deal of controversy among
researchers over the potential of SCC to occur at the chloride concentrations,
temperatures, and pH of physiological solutions. Several investigators [5,8] have noted
that minimum temperatures ranging from 60~ to 80~ are needed in chloride solutions
to induce SCC susceptibility ofaustenitic stainless steels, while others [9] have shown
that a critical acidity and chloride ion combination may induce SCC even at ambient
temperatures. Bundy et al. [8] also hypothesized that SCC may occur at lower
temperatures when 300 series austenitic stainless steels are stressed beyond their yield
point, which can happen in certain biomedical applications such as over-torqued bone
screws. Sheehan et al. [10] reported no effects on the strength and ductility of smooth
tensile samples subjected to slow strain rate SCC and no signs of SCC mechanisms on
the fractured surfaces of 316L in Ringer's solution at 37 ~ in vitro. Bundy and other
researchers [8,11] have hypothesized that due to possible bioelectric and biochemical
effects, the human body may be more conducive to stress corrosion cracking than a
simple chloride environment at 37 ~ From their observations, they suggest the
possibility for SCC of 316L in physiological solutions under certain extreme conditions.
It has been shown by Sedriks [12] that low nickel containing stainless steels may be
beneficial in resisting chloride SCC, but he also cautions that SCC susceptibility is
dependent on compositional, structural, and environmental factors that should not be
overlooked.
The purpose of the current study was to characterize the notch sensitivity and SCC
susceptibility of BioDur | 108 (alloy A), as compared to 316LS (alloy B) and 22Cr- 13 Ni5Mn (alloy C), which have a long and successful history as implant materials.
Samples of each alloy (A, B, C) were prepared from single lots of centerless ground 8
mm round bar stock supplied by Carpenter Specialty Alloys. Quantitative compositional
analyses was performed on a SPECTRO l inductively coupled plasma spectrometer (ICP)
with LISA spark attachment to confirm that the alloys met ASTM or Carpenter internal
melt limits for major elements. Compositional values, as well as, metallurgy and
hardness results have been reported elsewhere for these alloys [13,14].
Mechanical testing was performed using an MTS 2 servo hydraulic test system. Five
smooth and five notched tensile samples of each alloy were machined by Low Stress
ISPECTRO Analytical Instruments, Fitchburg, MA, USA

2MTS, Eden Prairie, MN, USA
156
Grind 3 (LSG) to a maximum surface roughness of 16 micro-inches (Ra=16) in the gauge.

Smooth samples were prepared with a 36 mm gauge length and a 6 mm gauge diameter
as shown in Figure la. Testing was performed at a 0.3 mm/min stroke rate to yield, and a
3.0 mm/min stoke rate from yield to failure and strain was measured using an
extensometer with a 25 mm gauge length. The ultimate tensile strength (UTS), 0.2%
yield strength (YS), elastic modulus (MOD), percentage elongation (%El.) to fracture,
and reduction of area (%ROA) were calculated. Notched tensile samples were prepared
with a K t factor of 3.2 and a notch diameter of 6 mm as shown in Figure lb. For many
years, the ratio of the notch tensile strength to the smooth tensile strength was considered
the most valid method for evaluating the sensitivity to a notch and the toughness of a
material. As specified in the ASTM "Standard Test Method for Sharp-Notch Tension
Testing with Cylindrical Specimens" (E602-91 [Reapproved 1997]), it is widely
recognized that since the onset of plastic deformation occurs at the yield strength, the
ratio of the notch tensile strength to the 0.2% yield strength may be a more useful
predictor of toughness. Notch tensile testing was performed according to ASTM E602
such that the load rate did not exceed the limit o f 690 MPa/min. The notch tensile
strength (NTS), the ratio of NTS to UTS, and the ratio of NTS to 0.2%YS were
determined. Scanning electron microscopy (SEM) was used to characterize the fractured
surfaces.
There are two generally accepted methodologies that may be employed when
evaluating SCC [15, 16]. The first method uses a constant strain rate. In order to perform
this test accurately, a strain gauge must be attached to the sample and remain in solution
for the duration of the test. This method presents a number of potentially significant
problems, including retention of the gauge on the sample, corrosion of the gauge, and
galvanic coupling o f the gauge and the sample. The second method uses a constant
extension rate as described in the ASTM "Standard Practice for Slow Strain Rate Testing
to Evaluate the Susceptibility of Metallic Materials to Environmentally Assisted
Cracking" (G129-00). In this method, only the sample and fixtures are in the corrosive
environment. By making the fixtures out of the same material as the samples, the
potential galvanic coupling effect can be eliminated. While the strain rate is not constant
with this method, if it is sufficiently slow, the small change in strain rate over the test
duration will not adversely affect the results. Due to the potential problems involved with
the strain gauge, testing was performed by the slow extension rate method as outlined in
ASTM G129, In order to use this method, a sufficiently slow or critical stain rate must be
determined. Previous research in our laboratories and several investigations by other
researchers as cited in the Metals Handbook [17] have determined that the appropriate
strain rate for anstenitic stainless steels is in the range of 10.5 to 10.6 sec 1.
Smooth and notched SCC samples (Kt=3.2, Ra=l 6 gin) of each alloy were machined
by LSG. Smooth SCC samples were machined to a gauge length of 10 mm and a gauge
diameter of 4 mm as shown in Figure 2a, and notched samples were machined with a 1
mm notch width and a 4 mm notch diameter as shown in Figure 2b. The mechanical
testing setup for SCC testing is shown in Figure 3. Triplicate smooth and notched SCC
3Low Stress Grind, Cincinnati, OH, USA
157
were tested in distilled/de-ionized water (DI) and Ringer's solution at 37 ~ using an

MTS servo hydraulic test system. The samples in this study were tested at a constant
stroke rate of 10 .5 mm/sec, to obtain the suggested initial strain rate o f samples
10 -6 mm/rmrdsec. In addition, one sample of each alloy in each solution was tested at
strain rates o f 10 -5sec 4 and 10-7sec-1 to validate the critical strain rate o f 10-6mm/mm/sec.
Critical dimensions o f samples were measured prior to and after testing to an accuracy of
10 -3 mm to calculate the %El, and %ROA. The percent elongation ratio (PER) and the
reduction of area ratio (ROAR) were calculated by dividing the mean values in Ringer's
solution by the mean values in distilled water. SEM was used to examine the fractured
surfaces.
A)
Ra = 16 I.tin Gauge
I[ "q'-"36'
1 6O~Omm
. 0 0"-'1~
m m[
~8.00mm
15Z5mm
D,
NOTCHDETAIL
B)
NotchRootDiameter6.00ram---fl ~".---NotchRadius.239mm
Ra = 16 pin Gauge
V
I II
,t
II
ii(i A
8.00mm
Ilu
nlllll
II III
152.5mm
Figure 1 - A) Sample drawingfor smooth tensile samples, B) Sample drawingfor notch
tensile samples.
Results
Tensile testing results for smooth and notched samples are summarized in Table 1.
Alloys A and C were found to have similar ultimate tensile strengths and yield strengths,
which were greater than those for alloy B. Alloy B had the greatest reduction of area,
followed by alloy A, and then alloy C. Alloy A was found to have the largest percentage
1 58
A)
Ra = 16 pin Gauge
~
9.5mm ~ [ q
,Li._~o.o~,~.;i, ,.,:m ,[, ,.,mm,(
9.5mm -
F/////////////A
B)
l ,OOmm I ............. ,F/////////////A
:~&00mm
NOTCHDETAIL
NolchRootDiameter4,00ram_ J "X---Notch Radius.187mm
Ra = 16 pin Gauge
Figure 2 - A) Sample drawingfor smooth SCC samples, B) Sample drawingfor notch

SCC Samples.
Figure 3 - Testing apparatusfor stress

corrosion cracking.
159
of elongation, followed by alloy B, and then alloy C. Figure 4 shows areas of the gauge
length of representative samples of each alloy after tensile testing. The effect of coldworking on the surface during the tensile test is clearly observable. The smooth tensile
samples for all three alloys demonstrated typical ductile cup and cone fractures (Figure
5a). Examination at higher magnifications showed a dimpled morphology normally seen
in ductile metals (Figure 5b).
Table 1 - Smooth and Notch Mechanical Testing Results (n =5)
Alloy
Alloy A
Alloy B
Alloy C
Elastic Modulus (GPa)
1884-3
167i5
186:52
UTS ~(MPa)
1344:57
10144-0
1351:k13
0.2% YS2(MPa)
1179:521
7934-7
1082:~13
% ROA3
68:50
714-0
60il
% Elongation
354-1
25+1
19:51
NTS 4 (MPa)
2255~-41
14554-7
2206:k41
1.91
1.83
2.04
1.68
1.43
1.63
( K t = 3.2)
NTS/
0.2% YS Ratio
NTS/UTS
Ratio
1UTS = ultimate tensile strength
2YS -- 0.2% yield strength
3ROA = reduction of area

4NTS = notch tensile strength
Testing on notched samples revealed alloys A and C to have similar notch tensile
strengths, which were greater than the values for alloy B. Both the NTS/UTS and
NTS/YS ratios for all three alloys were significantly greater than 1.00, indicating that
none of the materials tested were notch sensitive. SEM fracture analysis of notch tensile
samples of alloys B and C showed ductile cup and cone fracture with a dimpled
morphology at higher magnification (Figure 6a). The notched tensile fracture surfaces on
these two alloys were very similar to that of the smooth tensile fractures. On alloy A,
however, the notch tensile fracture revealed a complex morphology with areas of ductile
overload dimples due to microvoid coalescence and other areas with faceted appearance
of brittle cleavage as shown in Figure 6b (500X). It should be noted, however, that even
with what appears to be a partial brittle cleavage fracture mechanism, alloy A had the
highest notched tensile strength, percentage elongation, and NTS/UTS ratio.
Materials are considered to be susceptible to SCC if the percentage of elongation and
reduction of area in an aggressive corrosion medium are significantly inferior to those
properties in a non-aggressive solution. An elongation percentage ratio (PER) or
reduction of area ratio (ROAR) below 0.90 is deemed indicative that SCC
160
Representative samples of each alloy revealing the effect

of cold-working on the surface during smooth tensile testing. Note
the high level of deformation on the surface of alloy A.
Figure 4 -
F i g u r e 5 a - Representative scanning electron micrograph showing cup

and cone fracture of alloy A.
Representative scanning electron micrograph showing

ductile dimples on the fracture surface of smooth tensile samples of
alloy A.
Figure 5b -
F i g u r e 6 a - Representative scanning electron micrograph of the ductile

dimpled fracture mechanism on the notched tensile samples of alloy C.
161
162
Figure 6b - Representative scanning electron micrograph showing mixed
mode fracture mechanism on the notched tensile samples of alloy A.

mechanism may be taking place. Smooth and notched SCC results are provided in Tables
2 and 3, respectively. Percentage elongation and reduction of area values on the smooth
and notched samples for all of the alloys tested revealed only minor if any differences in
these values for each alloy between the aggressive Ringer's solution and non-aggressive
distilled/de-ionized (DI) water. The closer the PER and ROAR values are to 1.00, the
less susceptible the material is to SCC in the particular corrosive media in which testing
has been performed. Because all of the PER and ROAR ratios approach 1.00, SCC of
each alloy appears negligible. It is generally considered that values of PER and ROAR
above .90 constitute no SCC effects.
SEM analysis on the smooth SCC samples revealed ductile cup and cone fractures
with dimples evident at higher magnifications, These fracture surfaces had an essentially
identical morphology to that seen on the samples tested at higher strain rates in ambient
temperature air (Figures 5a and 5b). In addition, no differences in fracture morphology
were seen on the smooth fracture surfaces for SCC samples pulled at the slower and faster
strain rates in both solutions.
The appearance of the fracture surfaces on notched SCC samples of alloy B and alloy
C were essentially identical to those of the smooth samples indicating no visual signs of
SCC (Figure 7a). On notched SCC samples of alloy A, SEM analysis (Figure 7b)
revealed a similar morphology to that seen for notched tensile samples tested at a higher
strain rate in air (Figure 6b). These results, which once again show no detectible
differences in the fracture surface morphology regardless of testing media, indicate that
SCC mechanisms are not occurring regardless of the testing solution or strain rate.
163
Table 2 - SCC Results of Smooth Samples (n =3)

Alloy
Solution
% ROA 1
Alloy A
DI 4
56.8+2.2
26.3
58.0-2-1.1
26.5
71.5
18.5+0.6
Ringer's
71.3+1.4
18.5+0.2
DI 4
55.84-0.8
14.0+0.2
Ringer's
55.94-1.8
Ringer's
Alloy B
Alloy C
DI 4
~ROA = reduction o f a r e a
2PER = percentage elongation ratio
% Elongation
PER 2
ROAR 3
1.01
1.02
1.00
1.00
0.97
1.00
13.64-0.5
3ROAR = reduction o f area ratio
4DI = distilled/de-ionized water
Table 3 - SCC Results of Notched Samples (n=3)

Alloy
Solution
Alloy A
DI 4
Alloy C
% Notch
Elongation
16.2
43.9
15.4+1.7
40.3+1.9
DI 4
34.6
68.8+2.1
Ringer's
35.6
65.7+9.2
13.14-0.0
26.1
Ringer's
Alloy B
% ROA l
DI 4
Ringer's
12.8+0.3
~ROA = reduction o f a r e a
2pER = percentage elongation ratio
PER 2
ROAR 3
0.92
0.95
0.95
1.03
0.94
0.98
24.5
3ROAR -- reduction of area ratio
4DI = distilled/de-ionized water

Two commonly accepted measures for ductility in a material are elongation and
reduction o f area. Normally these variables are proportional, and thus a material with
more elongation will also have a greater reduction o f area. In the present study,
mechanical test results to determine notch sensitivity as well as SCC indicated alloy B
and alloy C followed these general principles. Alloy A, which is the low Ni stainless
steel, was found to have a substantially higher (-30%) percent elongation and yet 3-20%
less reduction of area in the smooth tensile and SCC samples than alloy B. This large
percent elongation with a somewhat lower reduction of area is believed to be a function
164
Representative scanning electron micrograph showing the

ductile fracture mechanism on the notched stress corrosion cracking
samples of alloy B.
Figure 7a -
Representative scanning electron micrograph revealing

mixed mode fracture mechanism on the notched stress corrosion
cracking samples of alloy A.
Figure 7b -
165
of the higher degree of work hardening experienced by alloy A as compared to the other
two alloys. This observation is evidenced by the appearance of the surfaces o f the
samples as seen in Figure 4. These observations may have a direct effect on processing
parameters during the fabrication of devices.
The SCC results obtained for alloy B agreed with the previous studies o f Sheehan et
al. [10] and Jones et al. [18], as the material exhibited typical ductile cup and cone
fractures by microvoid coalescence with no evidence of stress corrosion cracking
mechanisms on samples tested in physiological saline solutions. Alloy C revealed
similar fractures to those of alloy B, but this alloy showed significantly higher smooth
and notched tensile strengths, while testing resulted in lower percent elongation and
reduction o f area values. These are the results expected based upon data reported by
others [19-21]. It is interesting to note that alloy A had the highest elongation
percentages on the smooth tensile samples of the alloys tested, showed no signs of notch
sensitivity by the NTS/UTS and NTS/YS ratios, and yet SEM fracture analysis revealed a
mixed fracture mode. Areas of smooth terraces indicative of brittle cleavage intermingled
with areas of dimpling were noted. This fracture mechanism is similar to some forms of
SCC for other types of alloys described by others [12,22]. However, the percentage
elongation and reduction of area ratios between the more corrosively aggressive Ringer's
solution and distilled/de-ionized water were well above 0.90. In addition, the same
complex fracture morphologies, including ductile overload and brittle cleavage, were
observed on the notched samples of alloy A regardless of testing environments including
air, distilled/de-ionized water, and Ringer's solution. This combined evidence indicates
that SCC mechanisms are not taking place in this alloy or alloys B and C.
Results do indicate that the mixed fracture modes experienced by alloy A may be
occurring as a function of localized grain orientation, and/or greater work hardening in
specifically oriented grains as related to the direction o f applied stress. Alloy A has a
very high work hardening rate which is reportedly due to its high nitrogen content [2-4].
This mixed mechanism of fracture on the notched samples of alloy A may be a function
this high work hardening rate. If localized areas o f high tensile residual stress are created
at the root of the notch or crack tip during testing, areas that may become highly work
hardened can fracture by a brittle cleavage mechanism due to the presence of a tri-axial
state of stress. Once this localized brittle fracture has occurred, these internal residual
stresses are relieved. The total stress is the sum of the applied stress and these internal
residual stresses. Because the residual component of the total stress is now relieved, the
areas surrounding these brittle cleavage terraces are subjected to only the applied stress.
Ductile fracture may then occur. On the smooth samples, the applied stress is spread over
the entire gauge section and thus not concentrated at the notch. With this larger area of
stress distribution, these local areas do not become as highly cold worked and thus the
entire sample experiences ductile failure.
Overall, the mechanical properties for alloy A were similar to those of alloy C and
superior to alloy B. Based upon these results, and with the low-nickel content of alloy A,
this alloy appears to be a possible alternative for biomedical implant applications.
Acknowledgment
This research was supported by a grant from Carpenter Technology Corporation.
166
References
[ 1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[ 10]
[11]
[12]
[ 13]
[14]
[ 15]
[16]
[ 17]
[18]
Hierholzer, S. and Hierholzer, G., Internal Fixation and Metal Allergic, Thieme
Medical Publishers, 1992.
Disegi, J. A. and Eschbach, L., "Stainless Steel in Bone Surgery," Injury, Vol. 31,
Suppl. 4, 2000, pp. D2-D6.
Gebeau, R C. and Brown, R.S., "Tech Spotlight - Biomedical Implant Alloy,"
Advanced Materials and Processes, Vol. 159, No. 9, 2001, pp. 46-48.
Carpenter Technology Corporation Alloy Data, BioDur| 108 Alloy, Carpenter
Technology Corp., Reading, PA, 2000.
Bombara, G. and Cavallini, M., "Stress Corrosion Cracking of Bone Implants,"
Corrosion Science, Vol. 17, 1977, pp. 77-85.
White, W.E. Postlethwaite, J., May, I.L., "On the fracture of Orthopaedic Implants,"
Microstruetural Science, Vol. 4, 1976, pp. 145-158.
Gray, R.J., "Metallographic Examinations of Retrieved Intramedullary Bone Pins and
Bone Screws from the Human Body," Journal of Biomedical Materials Research
Symposium, No. 5, 1974, pp. 27-38.
Bundy, K.J., Marek M., and Hochrnan R.F., "In vivo and in vitro studies of the stresscorrosion cracking behavior of surgical implant alloys," Journal of Biomedical
Harston, J. D. and Scully, J. C., "Stress Corrosion of Type 304 Steel in HzSO4/NaC1
Environments at Room Temperature," Corrosion, Vol. 25, No. 12, 1969, pp. 496-501.
Sheenan, J.P., Morin C.R., and Packer K.F., "Study of Stress Corrosion Cracking
Susceptibility of Type 316L Stainless Steel in Vitro," Corrosion and Degradation of
Implant Materials, ASTM STP 859, 1985, pp 57-72.
Bundy, K. J. and Desai, V.H., "Studies of Stress-Corrosion Cracking Behaviour of
Surgical Implant Materials Using a Fracture Mechanics Approach," Corrosion and
Degradation of Implant Materials, ASTM STP 859, 1985, pp 73-90.
Sedriks, J. A., "Stress Corrosion Cracking," Corrosion of Stainless Steels, John Wiley
and Sons, 1992, pp. 267-325.
Roach, M.D. et al., "Physical, Metallurgical, and Mechanical Comparison of a LowNickel StainIess Steel," Proceedings of the 27th Annual Meeting of the Society of
Biomaterials, 2001 p. 343.
Zardiackas, L.D. et al., "Comparison of AnodicPolarization, Galvanic Corrosion, and
Fretting Corrosion of a Low-Nickel Stainless Steel to 316L and 22Cr-13Ni-5Mn,"
ASTM F4 Symposium on Stainless Steels for Medical and Surgical Applications, In
Review 2002.
Parkins, R.N., "Stress Corrosion Cracking - The Slow Strain Rate Technique," ASTM
STP 665, 1979, pp. 5- .
Lisagor, W.B., "Enviromaaental Cracking - Stress Corrosion," Corrosion Tests and
Standards, 1995, pp. 240-252.
Sprowls, D.O., "Evaluation of Stress-Corrosion Cracking," Metals Handbook, Ninth
Edition, t987, Vol. 13, pp. 245-282.
Jones R.L., Wing, S.S., and Syrettl B.C., "Stress Corrosion Cracking and Corrosion
Fatigue of some Surgical Implant Materials in a Physiological Saline Environment,"
Corrosion, Vol. 34, No. 7, 1978, pp. 226-236.
ZARDIACKAS ET AL. ON NOTCH SENSITIVITYAND STRESS CORROSION
167
[19] Black, J., Biological Performance of Materials: Fundamentals of Biocompatibility,

[20] Carpenter Technology Corporation Alloy Data, Carpenter 22Cr- 13Ni-5Mn, Carpenter
Technology Corp., Reading, PA, 1987.
[21] Carpenter Technology Corporation Alloy Data, Carpenter Stainless Type 316L,
Carpenter Technology Corp., Reading, PA, 1986.
[22] Kerlins, V. and Phillips, A, "Fractography: Modes of Fracture" ASMHandbook, Vol.
12, 1999, pp.12-71.
Ruth F. V. Villamil,1 Amaldo H. P. de Andrade,2 Celso A. Barbosa, 3 Alexandre

Sokolowski, 3 and Silvia M L. Agostinho4
Comparative Electrochemical Studies of F 1586-95 and F 138-92 Stainless Steels
in Sodium Chloride, pH = 4.0 Medium
Reference: Villamil, R. F. V., de Andmde, A. H. P., Barbosa, C. A., Sokolowski, A.,

and Agostinho, S. M. L., "Comparative Electrochemical Studies of F 1586-95 and
F 138-92 Stainless Steels in Sodium Chloride, pH = 4.0 Medium," Stainless Steels
for Medical and Surgical Applications, ASTM STP 1438, G. L. Winters and M. J.
Nutt, Eds., ASTM International, West Conshohocken, PA, 2003.
Abstract: In this work, ASTM F 1586-95 stainless steel used for surgical implants is
studied in 0.9 % NaCI aqueous solution, pH = 4.0 at 40 ~ using potentiodynamic
polarization curves and optical surface analysis techniques and its performance is
compared to F 138 stainless steel.
F 1586-95 remains passivated until
transpassivation potential above, which shows only generalized corrosion. F 138-92
presents breakdown potentials of 370 mV / SCE, with the presence of pitting
corrosion. The presence of 0.40 % Nb in F 1586-95 promotes passivation currents
one order of magnitude higher than that observed for F 1586-95 containing 0.28 %
Nb.
Keywords: breakdown potentials, austenitic stainless steel, ASTM F 1586-95,
ASTM F 138-92, orthopedic implants.
Stainless steels are extensively used in orthopedic implants [1-10]. These alloys are
less expensive than cobalt and titanium alloys [1-6]. Pitting corrosion of these ferrous
materials in chloride media has been one of the main reasons for the development of
new stainless steel compositions for implant applications [11-23]. The nucleation and
growth of pitting corrosion are related to the characteristics of oxide film formed
on the metallic surface [2-5]. Two types of stainless steels have been used in surgical
implants in Brazil: ASTM F 138 since 1982 and ASTM F 1586-95, more recently.
Postdoctoral Student, Instituto de Quimica, Universidade de Sgo Paulo, Av. Prof.
Lineu Prestes 748. 05508-900, Sgo Paulo, SP, Brazil.
2 Research Scientist, Instituto de Pesquisas Energ6ticas e Nucleares (IPEN) Trav. R
400 - Cidade Universitfiria, S.P - Brazil.
3 Manager and Research Engineer, respectively, Villares Metals S. A., Av. Alfredo
Dumont Villares, 155. 13177-900, Sumar6, SP, Brazil.
4Professor, Instituto de Quimica, Universidade de Sgo Paulo, Av. Prof. Lineu Prestes,
748. 05508-900, SgoPaulo, SP, Brazil.
168
Copyright9 2003 by ASTM International
www.astm.org
VILLAMIL ET AL. ON COMPARATIVE ELECTROCHEMICAL STUDIES
This last stainless steel presents about 0.5 weight percent of niobium in order to
reduce the chromium precipitation as chromium carbide, and also higher chromium,
molybdenum and nitrogen contents to promote a higher PRE number (Pitting
Resistance Equivalent). This composition also improves the mechanical properties in
such material.
The purpose of the present paper is to characterize the pitting corrosion resistance
of ASTM F 1586-95 samples in dearated 0.9 % NaC1 aqueous solution, pH = 4.0 at
40 ~ with two different niobium contents: 0.28 % (ISO A) and 0.40 % (ISO B).
The results are compared to those observed in the steel ASTM F 138-92 (F 138).
A pH = 4.0 of this solution was chosen for two reasons: its more aggressive
character and establish a better comparison than the normal pH = 7.0, which is
observed in human body fluids. This pH value can be attained when an inflammation
process is observed [8, 22].
Experimental Procedure
The chemical compositions of the stainless steel samples studied as well as the
PRE number of each alloy (Pitting Resistance Equivalent) are presented in (Table 1).
One can note the higher PRE number of the steel ASTM F 1586-95. The F 1586-95
was produced with the niobium content in the lower range of the ASTM specification
in order to have the minimum amount of niobium carbonitrides particles. The F 13892 was produced using an industrial 3.5 t vacuum induction furnace plus ESR
refining. The F 1586-95 material was produced starting from a 50 kg ingot in a
vacuum induction furnace of the Villares R&D Center. All the ingots were
homogenized at high temperatures and rolled to 30 mm diameter bars. The bars were
solution annealed at 1050 ~ for one hour followed by water cooling.
The stainless steel disk working electrodes were taken out from the core regions of
the bars and had an area of 0.363 cm 2. A cylindrical potytetrafluoroethylene (PTFE)
sleeve was fitted on the steel disk. A concentric brass rod was coupled to the steel +
PTFE. The disks surfaces were grounded and polished through no 320, 400 and 600
mesh wet silicon carbide sandpaper, rinsed with distilled water and ethanol and air
dried. The metallographic samples were polished through 1 ~tm diamond paste, rinsed
with water and ethanol and air dried. A platinum foil with a large area was used as
counter electrode and the saturated calomel electrode was the reference electrode.
The electrolytic cell was a boronsilicate glass with 150 mL capacity. The cell top
had three entries to pass the three electrodes. The experiments were conducted at
40 ~
Analytical grade sodium chloride and twice-distilled water were used to prepare
the solutions. The electrolyte was a 0.9 % NaC1 aqueous solution, pH = 4.0 adding
HC1.
The solutions were dearated with previously purified N2 for 1 hour. The electrodes
were immersed in the solution and after 45 minutes at open circuit potential an anodic
galvanostatic polarization was applied for 3 minutes at 100 mA cm 2 before the
electrochemical measurements. Anodic polarization curves were obtained using 0.17
mVs 1 scan rate. A computer controlled EG&G PAR 273A potentiostat/galvanostat
was used in the electrochemical measurements. This procedure is according to ASTM
G5 standard.
Optical microscopy of the surfaces were conduced for the materials after
polishment and anodic polarization curves. The anodic potentiodynamic polarization
was followed by 15 min etching at a value 50 mV higher than the potential
169
F 1586 - 95 e
(S 31675)
F 138- 92 ~
Grade 2
(S 31673)
F 1586 - 95
(S 31675)
Steel B
5832 -1A
F 138 - 92
Grade 2
(S 31673)
z
c,D
E
F
0.08
max.
0.030
max.
0.013
0.015
0.012
2.00
4.25
2,00
m~
3.77
4.09
1.94
Mn
0.010
max.
0.005
0.005
0.002
max.
max.
0,025 0.010
0,025
max.
0.015
0,014
0.023
0.75
max.
0.75
m~
0.32
0.33
0.26
Si
19.50
22.00
17.00
19.00
20,30
20.70
17.60
Cr
9.00
11.00
13.00
15.50
9.91
9.94
14.20
Ni
In ISO 5832 - 1 (1987) standard, molybdenum content is 2.25 to 3.50 %.

PRE ~ % Cr + 3,3 %Mo.
PRE= % C r + 3 . 3 % M o + 1 6 % N .
Maximum, unless range or minimum is indicated. Copper content is 0.50 % max.
Maximum, unless range or minimum is indicated. Copper content is 0.25 % max.
5832 - 9
5832 - 1 a
5832 - 9
5832 - 9
Designation
(UNS Designation)
F 1586 - 95
(S 31675)
ISO
ASTM
2.00
3.00
2.00
3.00
2.49
2,50
2,08
Mo
0.25
0.50
0.10
ma~
0.320
0.320
0.021
0,25
0.80
0.40
0.28
---
Nb
29.0 ~
29.0 c
24.5 ~
PRE
Chemical Composition of the Studied Materials and the Composition Specification According ASTM Standards, (wt %),
Steel A
F 138-92
Sample
T a b l e 1.
z
o3
f--
>
"u
r-
I'=
C)
>
m
~3
o)
-11
0
2o
I-"
m
o3
(/)
O3
--I
m
m
r-
>_
o)
-4
C)
corresponding to breakdown potential (Eb), for F138-92 and to transpassivation

potential (1100 mV) for ASTM F 1586-95.
The scanning electron microscopy (SEM) of the samples performed after 15
minutes attack at 50 mV above the potential corresponding to the current increase and
in different surface regions were analyzed using energy dispersive spectroscopy
(EDS)
The anodic polarization curves are shown in (Figure 1). All the materials are
naturally passivated in the aqueous solution studied but there is a significant
difference in the potential values corresponding to the current increase. F 138-92
presents breakdown potential (Eb) meanwhile for Steel A and Steel B the current
increases only at the transpassivation potential values. The potential value (El) of
current increase (breakdown potential (Eb) for F 138-92) and pass177
current
density values are presented in (Table 2).
0.72
,?
0.52
'J_ /
0.32
Steel B
0.12
F 138 - 9 2
Steel A
-0.08
-300
-100
100
300
500
700
900
1100
E vs SCE / mV
Figure 1. Potentiodynamic polarization curves o f stainless steels studied in 0.9 %

NaCl, p H = 4.0 medium at 40 ~ using O.17 m V s -1 scan rate.
The Steel A shows lower pass177
current values than F 138-92 and steel B. The
results reveal that Steel A (with 0.28 % niobium) is the best alloy. It can be seen that
Steel A and Steel B present a 600 mV potential range above F138-92 Eb, where they
are passivated.
Table 2. Potential values o f current increase (El) and pass177
current densities
O'pass) values for the three steels studied in 0.9 % NaCI aqueous solution, p H = 4.0 at
40~
Stainless steel
Samples
F 138-92
Ei vs. ECS / mV
370 +
Jpass //.tA cm 2 (100mV)
4 (6)
8.26 + 4(9)
SteelA
1150 + 12 (7)
0.82 1.3 (7)
Steel B
1050 32 (7)
6.61 1.8 (5)
The numbers in brackets denote the number of samples tested.
171
172
Optical microscopy of the samples polished through 1 ~m diamond showed the

presence of globular oxide inclusions for all materials studied, according to (Figure
2). The oxide inclusion rating was at the upper acceptable limit according ASTM F
138-92 and F 1586-92
F 138-92
Steel A
Steel B
Figure 2 (a) to (c). Optical microscopy of stainless steels samples polished through 1
ltm diamond paste, Inclusion rating according to ASTM E45, method A, plate 1-r.
Magnification: IOOX.
Figure 3 (a) to (e). Optical microscopy of stainless steels samples after 15 min
etching at 50 mV higher than the potential corresponding to the current increase (Et).
The optical microscopy of the surfaces of the three steel after etching at 50 mV
higher than the potential corresponding to the current increase (El) are shown in
(Figure 3). F 138-92 presents a significative localized corrosion while Steel A and
Steel B show surfaces similar to those observed when they are only polished, showing
only the non metallic inclusions. In order to show the absence of localized corrosion
in steel A and steel B different surface regions were analyzed using energy dispersive
spectroscopy (EDS).
The scanning electron microscopy (SEM) of the samples performed after 15
minutes attack at 50 mV above the potential corresponding to the current increase are
presented in (Figure 4). The delimited areas for EDS analysis and the results are
shown in (Table 3). The presence of localized corrosion for F138-92 SS is confirmed
from (Figure 4). The presence of more silicium in region II can be attributed to
silicate microinclusions show in (Table 3). These results suggest that the pitting
corrosion can begin around these inclusions, while the region I presents a composition
similar to the bulk.
It can be seen for steel A and steel B that regions I and II present nearly the same
chemical composition when compared to the alloy chemical composition, except in
region II in relation to niobium content. These results suggest that the attach at
potentials higher than the transpassivation potential is generalized. The higher
niobium content can be attributed to niobium carbonitride particles present in region
Figure 4. Scanning electron microscopy-SEM of stainless steel samples performed

after attack 15 minutes at 50 mV above the potential corresponding to the current
increase.
Table 3. Different surface regions analyzed using energy dispersive spectroscopy
(EDS).
ASTM F 138-92
F 138-92 (bulk)
Region I
Region II
Mn
1,94
1.7
2.05
Cr
16.60
17,3
16.29
Ni
14,20
10.64
12.09
Mo
2,08
2,55
2,02
Si
0.26
0.65
10.72
ASTM F 1586-95
Mn
Cr
Ni
Mo
Nb
Steel A (bulk)
4.09
20.70
9.94
2.50
0.28
Region I
4.14
20.22
9.24
2.31
0.45
Region II
3.95
20.14
9.14
2.31
3.4
ASTM F 1586-95
Mn
Cr
Ni
Mo
Nb
Steel B (bulk)
3.77
20.30
9.91
2.49
0.40
Region I
4.39
20.7
10.04
2.56
0.42
Region [I
4.75
20.35
10.1
2.59
0.7
Conclusions
Based on the tests conducted on the three Steels, the following conclusions are
pertinent:
ASTM F 138-92 and F 1586-95 stainless steel show different electrochemical
behavior in 0.9% NaC1 aqueous solution, pH = 4.0 at 40~
F 138-92 presents
breakdown potential at 370 mV/SCE and pitting corrosion while F 1586-95 is
passivated until the transpassivation potential.
173
174
F 1586-95 shows only generalized corrosion at 50 mV above the transpassivation

potential.
The Steel A sample, with lower niobium content (0.28 %) presents a better
performance in this solution when compared to Steel B sample, containing 0.40 %
niobium.
Acknowledgments
The authors acknowledge support from the Fundag~o de Amparo h Pesquisa do
Estado de S~o Paulo (FAPESP) for supporting part of this work. R.F.V. Villamil
also thanks FAPESP for the grant of a postdoctoral fellowship.
References
[1] Fonseca, C. Barbosa, M.A., "Corrosion behaviour of titanium in biofluids
containin H202 studied by electrochemical impedance spectroscopy", Corrosion
Science, Vol. 43, 2001, p. 547.
[2] Gonzhles, J. E. G., Mirza-Rosca, J. C., "Study of The Corrosion Behavior of
Titanium and some of its Alloys for Biomedical and Dental Implant
Applications", Journal
Electroanalytical Chemestry, Vol. 471, 1999, pp. 109-115.
13] Yu, S. Y., Scully, J. R., "Corrosion and Passivity of Ti-13% Nb-13% Zr in
Comparison to Other Biomedical Implant Alloys", Corrosion, Vol. 53, No. 12,
1997, pp. 965-976.
[4] Lausmaa J., "Surface spectroscopic characterization of titanium implant materials.
Journal of electron spectroscopy and related phenomena", Vol. 81, 1996, p.
343.
[5] Pankuch, M., Bell, R., Melendrest, C. A., "Composition and structure of the
anodic films on titanium in aqueous solutions", Electrochemical Acta, Vol. 38,
No.18, 1993, pp. 2777-2779.
[6] Tomashov, N.D., Chemova, G. P., Ruscol, Yu, S., Ayuyan, "The passivation of
alloys on titanium bases", Electrochimica Acta, Vol. 19, 1974, pp. 159-172.
[7] Lei M.K., Zhu X.M., "In vitro corrosion resistance of plasma source ion nitrided
austenitic stainless steels", Biomaterials, Vol. 22, 2001, p. 641.
[8] Mears, D.C., "Metals in medicine and surgery", Review 218, International Metals
Reviews, 1977, pp. 119-155.
[91 Fraker, A.C., Ruff, A.W., "Metallic surgical implants: state of art", Journal of
Metals, 1977, pp. 22-28.
[101 Gotman I., "Characteristics of Metals Used in Implants", Journal of
Endourology, Vol. 11, No. 6, 1997, Pp. 383-389.
VILLAMtL ET AL. ON COMPARATIVE ELECTROCHEMICAL STUDIES
[11] Walczak J. Shahgaldi F. Heatley F., "In vivo corrosion of 316 L stainless-steel
hip implants: morphology and elemental compositions of corrosion
products", Biomaterials, Vol.19, No. 1-3, 1998, pp. 229-237.
[12] Taira M, Lautenschlager EP, "In vitro Corrosion Fatigue of 316 L Cold-Worked
Stainless-Steel", Journal of Biomedical Materials Research, Vol. 26, No. 9,
1992, pp. 1131-1139.
I131 Sobral A.V.C., Ristow Jr. W., Azambuja D.S., Costa I., Franco C.V.,
"Potentiodynamic test and electrochemical impedance spectroscopy of
injection model 316 L steel in NaC1 solution", Corrosion Science, Vol. 43,
2001, p. 1019.
[141 Cakir A., Tuncell S., Aydin A., "AE response of 316 L SS during SSR test under
potentiostatic control", Corrosion Science, Vol. 41, 1999, p. 1175.
[15] Guemmaz M., Mosser A., Grob J-J, Stuck R., "Sub-surface modifications
induced by nitrogen ion implantation in stainless steel (SS 316 L). Correlation
between microstructure and nanoindentation results", Surface and Coating
Technology, Vol. 100, i998, p. 353.
1161 Rogero S.O., Higa O.Z., Saiki M., Correa O.V. and Costa I., "Cytotoxicity due to
corrosion of ear piercing studs", Toxicology in vitro, Vol. 14, 2000, p. 497.
[17] Reclaru L., Lerrf R., Eschler P.-Y., Meyer J.-M., "Corrosion behavior of a
welded stainless-steel orthopedic implant", Biomaterials, Vol. 22, 2001, pp.
269.
[181 Basridas J. M., Polo J. L., Torres C. L., Cano E., "A study on the stability of
AISI
316 L stainless steel pitting corrosion through its transfer
function", Corrosion
Science, Vol. 43, 2001, p. 269.
[19] Laycock N.J., Newman R.C., "Temperature dependence of pitting potentials for
austenitic stainless steels above their critical pitting temperature", Corrosion
Science, Vol. 40, No. 6, 1998, p. 887.
[20] Carranza R.M., Atvarez M.G., "The effect of temperature on the passive film
properties and pitting behaviour of a Fe-Cr-Ni alloy", Corrosion Science, Vol.
38, No. 6, 1996, p. 909.
121] Disegi J.A., Eschbach L., "Stainless steel in bone surgery", International Journal
of the Care of the Injured, Vol. 31, 2000, S-D2-6.
[22] Raimondi M.T., Pietrabissa R., "Modelling evaluation of the testing condition
influence on the maximum stress induced in a hip prosthesis during ISO 7006
fatigue testing", MedicalEngineering & Physics, Vol. 21, 1999, p.353.
[23] Paszenda Z., Marciniak J., "The influence of base structure and carbon coating
on the corrosion resistance of Co-Cr-Mo alloy", Journal of Materials
Processing Technology, Vol. 78, 1998, p. 143.
175
Bernard S. Covino, Jr, 1 Charles H. Craig, 2 Stephen D. Cramer, 3 Sophie J. Bullard, I

Margaret Ziomek-Moroz, 1 Paul D. Jablonski, 4 Paul C. Turner, 4 Herbert R. Radisch,
Jr., 2 Nev A. Gokcen, s Clifford M. Friend, 6 and Michael R. Edwards 6
Corrosion Behavior of Platinum-Enhanced Radiopaque Stainless Steel (PERSS ~)

for Dilation-Balloon Expandable Coronary Stents
Reference: Covino, B. S., Jr., Craig, C. H., Cramer, S. D., Bullard, S. J., Ziomek-Moroz,
M., Jablonski, P. D., Turner, P. C., Radisch, H. R., Jr., Gokcen, N. A., Friend, C. M., and
Edwards, M. R., "Corrosion Behavior of Platinum-Enhanced Radiopaque Stainless
Steel (PERSS e) for Dilation-Balloon Expandable Coronary Steuts," Stainless Steels
for Medical and SurgicalApplications, ASTMSTP 1438, G. L. Winters and M. J. Nutt,
Eds., A S T M International, West Conshohocken, PA, 2003
Abstract
Dilation-balloon expandable coronary stents are made o f implant grade stainless steels,
UNS $31673, e.g., BioDur | 316LS. Boston Scientific / Interventional Technologies
(BS/IVT) determined that addition o f platinum to UNS $31673 could produce a stainless
steel with enhanced radiopacity, which made such stents more visible radiographically.
A goal o f the program was to ensure the platinum additions would not adversely affect
the corrosion resistance o f the resulting 5-6 wt % PERSS | alloys. Corrosion resistance
o f PERSS and BioDur 316LS was determined using electrochemical tests for general,
pitting, crevice and intergranular corrosion. Experimental methods included A262E,
F746, F2129, and potentiodynamic polarization. The ~ 6 wt % PERSS alloy (IVT 78)
had a resistance to pitting, crevice and intergranular corrosion similar to base materials.
IVT 78 was a single-phase austenitic PERSS alloy with no evidence o f inclusions or
precipitates; it was more resistant to pitting corrosion than the - 5 wt % PERSS alloys.
PERSS performance was not a function o f oxygen content in the range 0.01 to 0.03 wt %.
J Research Chemist, Materials ConservationDivision, Albany Research Center, U.S. Department of
Energy, 1450 Queen Ave., SW, Albany, OR 97321.
2 Principal Engineer and Vice President for Research & Development,respectively, Boston Scientific
Corporation/InterventionalTechnologies, 3574 Ruffin Rd, San Diego, CA 92123.
3 Chemical Engineer, Materials Conservation Division, Albany Research Center, U.S. Department of
Energy, 1450 Queen Avenue, SW, Albany, OR 97321.
4 Metallurgist and Division Chief, respectively, Thermal Treatment Technology Division, Albany Research
Center, U.S. Department of Energy, 1450 Queen Avenue, SW, Albany, OR 97321.
5 Consultant, Thermodynamicsand Inorganic Materials, Palos Verdes Estates, CA 90274.
Cranfield Postgraduate Medical School, Shrivenham, Swindon SN6 8LA, UK.
176
Copyright9 2003 by ASTM lntcrnational
www.astm.org
COVINO ET AL. ON CORROSION BEHAVIOR OF PERSS
177
Keywords: coronary stents, corrosion, stainless steel, platinum, alloy, Ringers solution,
pitting corrosion, breakdown potential, radiopacity, x-rays
Introduction
Stainless steels have long been used to fabricate medical devices for the human body.
These devices include artificial joints and, especially during the past decade, vascular
stents. It is estimated that nearly a million patients worldwide undergo interventional
procedures each year with 50 % of these patients receiving coronary vascular stents [1].
The commercial, implant grade alloy BioDur | 316LS 1 (UNS $31673) has been used to
fabricate such stents. The stents are inserted into diseased arteries and are designed to
have adequate mechanical and corrosion properties in blood plasma environments.
Stainless steel stents as used today can often be almost radiolucent to x-rays [2] when
placed in the human body. This in turn makes it difficult to locate their position with the
required precision. Gold, platinum, or tantalum marker bands are sometimes affixed to
catheters or coatings applied to stents to increase radiopacity for precise location of stents
[2]. A new method is described in this volume for increasing stent radiopacity by
modifying the composition of the stainless steel alloy used in fabricating coronary stents.
This method consists of alloying precious metals, such as platinum, with a stainless steel.
Such platinum additions enhance the radiopacity of stainless steel, leading to a new class
ofradiopaque stainless steels, PERSS | 2, for applications within the human body [3-4].
However, it must be shown that such alloying can be accomplished without significant
reduction in the mechanical properties or corrosion resistance of the deployed stent.
Several studies have been conducted on the effect of precious metal additions on the
corrosion performance of stainless steels. A summary of the research conducted between
1949 and 1980 is found in the reviews of McGill [5-6]. These papers report on the
effects of palladium, ruthenium, iridium, osmium, and platinum additions on the
corrosion resistance of stainless steels. Test environments ranged from oxidizing to
reducing solutions and included solutions that cause either pitting or stress corrosion
cracking. A summary of the reviews suggests: (1) a critical noble metal concentration
must be reached for passivity of the steel; (2) noble metal modified steels corrode rapidly
before becoming passive; and (3) the improvement in corrosion resistance with noble
metal additions is less for the higher alloyed steels that contain, for example,
molybdenum. There is also some evidence that noble metal modified steels are more
susceptible to stress corrosion cracking. Unfortunately the results presented by McGill
[5-6] are for solutions more typical of the chemical process industry than of the human
body. To assess the corrosion resistance of alloys considered as biomaterials, it is
necessary to work in an environment comparable to that encountered in the body. In
vitro corrosion experiments do not normally use whole blood or blood plasma, but rather
a synthetic environment that simulates the properties of blood. Corrosion studies have
been conducted in a variety of simulated blood environments. Synthetic body fluid
Registered trademark, CarpenterTechnologyCorporation,P.O. Box 14662, ReadingPA 19612-4662.

2Registeredtrademark, Boston Scientific Corporation/ScimedLife Systems,Maple Grove,MN 55311.
178
(SBF) is one such solution [7]. It is compared to the composition o f a synthetic blood
plasma [8] in Table 1.
Table 1 - Comparison of synthetic

bodyfluid and blood plasma solutions
Solution
ions
Composition, mM
SBF [71
PLasma [8]
Na
142
CI
194
K
5.0
Mg 2
1.5
Ca 2
2.5
HCO3
4.2
HPO4 2
1.0
SO42NA
NA = Not applicable
142
148
5.0
1.5
2.5
4.2
1.0
0.5
Another environment is a synthetic blood plasma that is in conformity with the

recommendations of Association Frangaise de Normalisation (AFNOR)[9], but no details
are given. ASTM Test Method for Conducting Cyclic Potentiodynamic Polarization
Measurements to Determine the Corrosion Susceptibility of Small Implant Devices
(F2129) offers three solutions for testing implant devices such as stents - Tyrodes,
Ringers, and Hanks; the compositions of these solutions are given in Table 2.
Table 2 - Simulatedphysiological solutions

suggested in ASTM F2129
Compound
NaC1
CaC12
KC1
MgC12 9 6H20
MgSO4 9 7H20
T~,rodes Rin[ers Hanks

Concentration, ~/L
8.0
0.20
0.2
0.10
NA
NaHCO3
1.00
Na2H2PO4
0.05
Na2HPO4 92H20
NA
Glucose
1.00
Solution [oH
7.4
NA = Not applicable
9.0
0.24
0.42
NA
NA
0.20
NA
NA
NA
7.4
8.0
0.14
0.4
0.10
0.06
0.35
0.10
0.06
1.00
7.4
Leclerc reviewed both in vivo and in vitro research on surgical implant materials,
including UNS $31603 (AISI 316L), conducted between 1966 and 1985 [ 10]. The
research included measurements of corrosion potential and corrosion rate under pitting,
179
stress corrosion cracking, corrosion fatigue, fretting and crevice corrosion, galvanic
corrosion, and sensitization conditions. He concluded that the basic surgical implant
materials, when in their correct metallurgical condition, have adequate strength and
corrosion resistance for most orthopedic applications
The primary purpose of the research reported here is to determine the effects of
platinum additions on the corrosion resistance of UNS $31673-based PERSS alloys in
deaerated Ringers solution representative of blood or blood plasma. A secondary
purpose is to determine the effect of dissolved oxygen, introduced into alloys in the
melting process, on the corrosion resistance of the UNS $31673-based PERSS alloys.
Experimental Methods
Alloy Fabrication
The materials used in this study were BioDur 316LS, the base material used to make
these PERSS alloys, and PERSS alloys made by the addition of 5 to 6 wt % platinum to
the base material. The ~ 6 wt % PERSS alloy (IVT 78) has more optimized processing
parameters and more closely resembles a commercial product. The ~ 5 wt % alloys (IVT
30-series and 50 series), made earlier, are exploratory alloys; their processing parameters
were somewhat less refined than either the BioDur 316LS or IVT 78. Corrective
additions of Cr and Mo were made to all PERSS alloys to account for dilution resulting
from the addition of platinum [3]. The corrections were made so the pitting resistance
equivalent (PRE) of the PERSS alloys was equal to a PRE of 26 or greater; this was done
by computing the corrections using the equation PRE = [Cr] + 3.3 * [Mo], where the
chromium [Cr] and molybdenum [Mo] concentrations of the alloy are given in weight %.
Alloys IVT 37, 38, 50 and 78 were consolidated by vacuum induction melting BioDur
316LS base with either a 5 wt% Pt addition (IVT 37, 38 and 50) or 6 wt % Pt addition
(IVT 78) and enough additional Cr and Mo to return the overall composition to a PRE >
26. IVT 50 and 78 were remelted in a vacuum arc furnace to further refine the structure
and provide sound material for further processing. Both IVT 54 and 56 were derived
from IVT 50; they were produced to yield alloys of different oxygen content. IVT 54 and
IVT 56 were vacuum induction melted under a partial pressure of argon, but they were
not subsequently vacuum arc remelted. IVT 54 consisted of 1 kg of IVT 50 inductively
melted in a new alumina (A1203) crucible and poured into a new conical mold. IVT 56
consisted of I kg IVT 50 plus 250 ppm aluminum (A1) plus 750 ppm calcium oxide
(CaO) inductively melted in the same crucible as IVT 54 and poured into a conical mold.
Alloy Description and Analysis
The results of wet chemistry and inductively-coupled plasma atomic absorption

spectroscopy (ICP AA) chemical analyses of the base material and the PERSS alloys are
180
listed in T a b l e 3. I V T 50-series alloys h a d h i g h e r o x y g e n c o n t e n t s t h a n the b a s e material.

Cr, Ni, and M o contents o f the P E R S S alloys w e r e similar to t h o s e o f the b a s e material.
All o f the alloys (Table 4) were tested in their a n n e a l e d condition. T h e b a s e material and
I V T 78 h a d similar h a r d n e s s values. The I V T 78 grain size was refined b y cold rolling,
as c o m p a r e d to the grain size o f the base material, w h i c h was o b t a i n e d in bar form. B o t h
alloys were single-phase austenitic alloys w i t h n o e v i d e n c e o f i n c l u s i o n s or precipitates.
Table 3 Element
Chemical Analysis of alloys, wt %
BioDur
i 316LS
Heat 1
BioDur J
316LS
Heat 2
IVT 78
Carbon
0.018
0.018
0.015
NA
NA
Silicon
Manganese
0.45
1.80
0.38
1.78
0.41
1.18
0.45
1.54
0.45
1.54
Sulfur
Phosphorus
Chromium
0.001
0.015
17.6
0.001
0.017
17.4
NA
NA
NA
NA
NA
NA
18.2
18.7
Nickel
14.8
14.7
Molybdenum
2.81
2.79
13.4
2.90
13.3
2.94
Copper
0.09
0.11
0.10
Cobalt
0.07
0.05
>0.01
Aluminum
0.009
0.006
Nitrogen
Titanium
0.025
0.002
0.029
<0.002
>0.01
NA
0.007
NA
NA
NA
NA
NA
Niobium
0.013
0.020
0.022
0.014
0.014
0.014
0.014
0.015
Vanadium
Platinum
0.07
0.00
0.06
0.00
0.0069
NA
0.06
6.01
NA
0.03
5.32
0.0205
0.03
5.32
0.0305
0.03
5.32
0.0100
0.07
4.95
NA
0.06
4.78
0.0400
Oxygen
UNS $31673.
IVT 50
IVT 54
IVT 56
IVT 37
IVT 38
NA
NA
0.027
0.45
1.54
0.48
1.71
0.47
0.96
NA
NA
NA
NA
0.0025
NA
18.7
18. 7
17.5
17.5
13.3
2.94
13.3
2.94
13.6
2.87
14.2
2.89
0.10
0.10
0.10
0.08
0.07
NA
NA
NA
NA
NA
0.005
NA
0.005
NA
0.013
NA
0.006
NA
0.009
0.056
NA = not available.
Table 4 -
Alloy hardness and grain size
Material
BioDur 316LS, Heat 2,
annealed and electropolished
IVT 78, vacuum annealed and
electro~aolished
Hardness
, VHN ~1 k~ load t
155
ASTM grain
size number
6.5
161
10.0
Corrosion Tests
T h e p r i m a r y test m e t h o d used for evaluating corrosion susceptibility was F2129; it

was used to m e a s u r e pitting corrosion resistance in n i t r o g e n d e a e r a t e d Ringers solution o f
181
the alloys in Table 3, with the exception o f l V T 30-series alloys. For comparison with
the results from F2129 tests, potentiodynamic polarization measurements were made on
selected alloys. Additional measurements were made on BioDur 316LS, Heat 1, and on
IVT 37 and IVT 38, using Practice E of ASTM Practices for Detecting Susceptibility to
Intergranular Attack in Austenitic Stainless Steels (A262), and ASTM Test Method for
Pitting or Crevice Corrosion of Metallic Surgical Implant Materials (F746).
F2129 - This test method is for assessing corrosion susceptibility of specified metallic
medical devices or components using cyclic forward and reverse potentiodynamic
polarization. Examples of specified devices include coronary stents. The test method is
intended to evaluate the medical device in its final form and finish, i.e., as it would be
implanted. While it was not the aim of the present research to evaluate finished devices,
the test method provided a valuable basis for assessing pitting corrosion resistance of
PERSS alloys and comparing it to that of their base material. Most of the alloys were
surface ground to a 120-grit aluminum oxide finish prior to testing. The BioDur 316LS,
Heat 2, and IVT 78 alloys were prepared for testing in a manner closely approximating
that used for coronary stents, i.e., polished to a 1 pm finish and then electropolished. All
testing was in Ringers solution (Table 2), because this solution most closely simulates the
composition of blood plasma. Samples were immersed in Ringers solution at 37 C after
deaerating with high purity nitrogen. All potential measurements were made with respect
to the saturated calomel electrode (SCE). The open circuit corrosion potential (Ecor~)was
measured for one hour. After one hour, the cyclic potentiodynarnic scan was started in
the positive (noble) direction at 10 mV/min, beginning at a potential 100 mV negative to
the last measured Eco,r value. The potential scan was reversed when current density
reached a value approximately two decades greater than current density at the breakdown
potential (Eb), sometimes called the pit nucleation potential, Enp [11]. The scan was
halted when the potential reached 100 mV negative of the Ecor~or when current density
dropped below the passive current density and a protection potential (Eprot) was observed.
All tests were conducted in a flat cell modified to simulate the standard Avesta cell
[12]. High purity water flowed through a fiber washer at 0.6 ml/min to maintain crevicefree conditions at the test sample surface. Figure 1 shows the experimental apparatus,
Figure 1 - F2129 and potentiodynamic polarization experimental equipment showing the

modified flat cell with an Avesta head, solution reservoir, water reservoir, and pumps
182
including the flat cell fitted with an Avesta head, the solution reservoir, and the high
purity water reservoir. Two or more tests were performed on each alloy.
Potentiodynamic polarization measurements were made on the two electropolished
alloys, BioDur 316LS, Heat 2, and IVT 78, to show the full polarization curve, including
both anodic and cathodic branches. All procedures were the same as in F2129 except that
the cyclic scan was started in the positive (noble) direction at 10 mV/min, beginning at a
potential of-300 mV vs SCE. The scan was reversed when current density reached a
value approximately two decades greater than the current density at the breakdown
potential (Eb). The scan was halted at the potential where the current density dropped
below the passive current density, i.e., the protection potential (Eprot).
A 2 6 2 E - T h i s practice is a requirement of ASTM Specification for Wrought:l 8
Chromium-14 Nickel-2.5 Molybdenum Stainless Steel Bar and Wire for Surgical
Implants (F138) and ASTM Specification for Wrought 18 Chromium-14 Nickel-2.5
Molybdenum Stainless Steel Sheet and Strip for Surgical Implants (F139). It is intended
to determine susceptibility of austenitic stainless steels to intergranular corrosion attack
associated with precipitation of chromium-rich carbides.
Duplicate samples of BioDur 316LS, Heat 1, and IVT 37 and IVT 38 were tested in
annealed and sensitized heat-treat conditions. The sensitized samples were heat-treated at
675 C for one hour and air cooled. All samples were ground to a 120-grit finish. They
were then embedded in copper granules and exposed for 24 hours to a boiling solution of
100g/L hydrated copper sulfate (CuSO4" H20) and 100 mL/L of concentrated sulfuric acid
(H2SO4), Figure 2. After exposure, the samples were bent through a 180 arc over a
mandrel with a diameter equal to the thickness of the samples. The bent samples were
examined at 20X magnification for cracks indicative of a sensitized material.
Figure 2 - Apparatus for conducting ASTM A262, practice E

F 7 4 6 - This procedure is not a requirement of F138 and F139 for UNS $31673 alloys.
It is designed solely for determining comparative laboratory indices of performance. The
results are used for ranking alloys in order of increasing resistance to pitting and crevice
corrosion under the specific conditions of the test method. The method is intentionally
designed to produce conditions sufficiently severe to cause breakdown of UNS $31603,
an alloy similar to UNS $31673 and also considered acceptable for surgical implant use.
183
Alloys that suffer pitting and crevice corrosion during the more severe portion of the test
do not necessarily suffer localized corrosion when placed in the human body as implants.
Three samples each of BioDur 316LS, Heat 1, and IVT 38 were evaluated in the
annealed condition. The surface of each cylindrical sample was first ground to a 600-grit
finish. Samples were fitted with an inert tapered collar and immersed in a saline solution
(consisting of 9 g/L sodium chloride (NaC1) in distilled water) at 37~ Ecor~was
established after immersion for one hour. Localized corrosion was then stimulated by
potentiostatically polarizing the sample to a potential of+800 mV vs SCE. Localized
corrosion was marked by a large and generally increasing anodic current. The potential
was then decreased rapidly to the original Er
If the alloy was susceptible to localized
corrosion, the current remained at a relatively high value and fluctuated with time. If the
pits or crevices repassivated and localized attack was halted, the current dropped to a
value typical of a passive surface and continued to decrease. If the sample repassivated,
it was repolarized to +800 mV vs SCE and the potential decreased to a value more noble
(more positive) than Ecor~and the current response observed. This was repeated stepwise
until a potential was reached where the sample did not repassivate. The critical potential
for localized attack is the most noble potential at which localized corrosion repassivated,
i.e., similar to Eprot.
Results
F2129
I V T 78 P E R S S alloy - Results of the F2129 test on the Biodur 316LS, Heat 2, and the
PERSS alloy IVT 78 alloy are shown in Figure 3 and summarized in Table 5. Parameters
measured in the ASTM F2129 tests were Eeoc, Eb, F-wrotand I~o~;derived values included
(Eb - Ecorr) and (Eb - Eprot). The heavy curves in Figure 3 are for Biodur 316LS, Heat 2.
They bracket the results for IVT 78. Both the base material and IVT 78 were passive
until Eb was achieved. Measured Eb values for IVT 78 were grouped between those
measured for Biodur 316LS, Heat 2, Figure 3 and Table 5. No Eprotwas measured for
Biodur 316LS, Heat 2, or IVT 78 because the ASTM F2129 test did not allow potentials
more negative than -100 mV below the last measured E~o~rto be included in the scans.
Full potentiodynamic polarization curves, including both the anodic and cathodic
branches, are shown in Figure 4 for Biodur 316LS, Heat 2, and IVT 78. They are typical
for stainless steel-based alloys in contact with chloride solutions at moderate pH values
[13] as opposed to the curves generated by ASTM F2129, Figure 3. The curves in Figure
4 show extended regions of passivity, a breakdown of the passive film due to the
initiation and growth of pits, and a well-developed hysteresis loop. The presence of the
hysteresis loop is an indication that the alloys are susceptible to localized corrosion.
These results are summarized in Table 5. The polarization curves include both Er and
Eprot. They show the curves for Biodur 316LS, Heat 2, and IVT 78 were similar except
for the values of E~o~. E~o~for IVT 78 was higher than that for Biodur 316LS, Heat 2.
Both Biodur 316LS, Heat 2, and IVT 78 were passive at potentials up to Eb. The Eb
184
values for the two alloys were almost the same and Eprot values for the two alloys were
similar. However, IVT 78 repassivated at potentials below (cathodic to) Ecorrwhile
Biodur 316LS, Heat 2, repassivated at potentials above (anodic to) Eeo~r. Corrosion
currents for the Biodur 316LS, Heat 2, and IVT 78 were low and similar.
Photomicrographs are shown in Figure 5 of the BioDur 316LS, Heat 2, and IVT 78
surfaces following the ASTM F2129 test and potentiodynamic polarization scans (the
staining is not significant). In each case pitting occurred at only one or two locations on
the 1 cm 2 exposed surface. There was no other population of pits distributed over the
surface and there was no evidence o f crevice corrosion. Results were the same for both
BioDur 316LS, Heat 2, and IVT 78.
1.0
Biodur 316LS, Heat 2
0.8
0.6
0.4
O
0.2
0~
1.E-09
1.E-07
1.E-05
1.E-03
Current density, A/em 2

Figure 3 - F2129 results f o r Biodur 316LS, Heat 2, and IVT
IVT 50-series PERSS alloys - Figure 6 shows cyclic potentiodynamic polarization
curves for BioDur 316LS, Heat 1, and IVY 56 in de-aerated Ringers solution. The curve
for IVT 56 shown in Figure 6 is qualitatively similar to that for each o f the IVT 50-series
alloys. Pits were observed on the exposed surface of the IVT 50 series alloys, but there
was no indication of crevice corrosion in the area where the sample was sealed to the test
cell. Both BioDur 316LS, Heat 1, and the IVT 50-series alloys exhibited breakdown
potentials more noble than their corrosion potentials. Eb for BioDur 316LS, Heat 1, was
more noble than that for the IVT 50-series alloys, indicating a greater resistance to pitting
corrosion for BioDur 316LS than the IVT 50-series alloys.
185
Table 5 - Results of the ASTM F2129 Test

Sample and finish
BioDur 316LS,
Heat 1, 120-grit
finish
BioDur 316LS, Heat
2, electropolished
Ecorr
Eb
Eprot
Icorr
Eb-Ecorr
wt%
VvsSCE
VvsSCE
VvsSCE
~tA/cm2
Eb-Eprot
V
0.007
0.150
0.742
0.154
NA
0.592
0.588
NA
0.008 I
0.860 1
0.128 j
0.011 t
0.852 I
0.732 I
0.042 I
0.722 i
0.918 I
0.856
0.535
NA
IVT 78,
electropolished
0.217 t
0.939 1
0.021 i
0.636
0.608
IVT 56, 120-grit

finish
0.0100
IVT 50, 120-grit

finish
0.0205
IVT 54, 120-grit

finish
0.0305
-0.098
0.340
0.103
0.378
0.438
0.237
-0.079
0.319
0.100
0.138
0.398
0.219
-0.212
0.272
0.157
0.051
0.484
0.429
-0.185
0.515
0.117
NA
0.700
0.632
-0.223
0.204
-0.009
0.192
0.427
0.213
0.014
0.452
0.158
0.022
0.466
0.610
-0.183
0.339
0.165
0.141
0.522
0.174
0.008
0.326
0.195
0.180
0.334
0.131
1 potentiodynamic polarization. NA = not available
1.2
Biodur 316LS, Heat 2

IVT 7
0.8
m
0.4
0.0
-0.4
1.E-10
1.E-08
1.E-06
Current density,
Figure 4 -
1.E-04
1.E-02
A/era 2
Potentiodynamicpolarization resultsfor Biodur 316LS, Heat 2, and IVT 78
186
Figure 5 - Alloy surface following polarization tests in Ringers solution showing pits
that formed following breakdown of the passive film (staining inconsequential to tesO.
Effect of Alloy Oxygen Content on Localized Corrosion of lVT 50-series PERSS alloys
In general, local imperfections in passive films, such as inclusions, increase the
susceptibility of an alloy to localized corrosion [14]. Oxygen incorporated into an alloy
during melting and fabrication can result in the formation of oxide inclusions. Oxide
inclusions at the Surface o f stainless steels can affect the stability of the passive film
formed on stainless steels when evaluated by F2129. Inclusions can become sites for pit
initiation, thereby reducing the alloy resistance to pitting corrosion. The IVT 50-series
alloys were made with different oxygen contents and had second phases present in their
microstructure. Results for the alloys are plotted in Figure 7 and show that Eco=, Eb, or
Eprot were not a function of oxygen content in the range 0.01 to 0.03 wt % oxygen.
However, IVT 50-series alloys were more susceptible to pitting corrosion than IVT 78.
187
1.00
0.75
Biodur 316LS
Heat 1
0.50
oH
0.25
@
0.00
-0.25
1.E-10
I V T 56
1.E-08
1.E-06
1.E-04
1.E-02
C u r r e n t density, A/cm 2
Cyclic potentiodynamic polarization curves for Biodur
316LS, Heat 1, and IV]" 56 conducted according to ASTM ['2129
Figure 6 -
0.8
9 Ecorr
9
9 Eprot
Breakd'own Potential - Biodur 316LS

0.6
O
cJ
.i.i
0.2
c-
o
tL
Breakdown Potential - IVT 50, 54, 56
0.4
I
BE
m
n
Protection Potential - kVT 50, 54, 56
Corrosion Potential - IVT 50 v 547 56
-0.2
-0.4
0.01
0.02
0.03
Alloy oxygen content, wt %

F i g u r e 7 - Effect of alloy oxygen content on electrochemical parameters
for Biodur 316LS, Heat 1, and IVT 50-series alloys
188
The behavior of alloys IVT 37 and IVT 38 was identical to that of BioDur
316LS, Heat 1, in the A262E test. Figure 8 shows the surface o f the BioDur 316LS, Heat
1, and IVT 37 alloys after 24 hour immersion in the copper sulfate solution and then
bending through a 180~ arc. None of the samples exhibited cracks or fissures on the bend
radius. This indicates the alloys were not susceptible to intergranular corrosion attack.
As a consequence of these results it can be surmised that annealing did not precipitate
chromium-rich carbides and the alloys were not susceptible to sensitization at the
annealing conditions. While not tested by A262E, metallographic examination of BioDur
316LS, Heat 2, and IVT 78 alloy microstructures showed them to be single-phase
austenitic materials with no evidence o f inclusions or precipitates.
A262E
Figure 8. AHoy surface after ASTM A262E test followed by bending through
a 180 ~ arc (20X).
ASTM F746
The results of F746 are shown in Table 6 for BioDur 316LS, Heat 1, and 1VT 38. The
critical potential for pit repassivation, Eprot, o f the alloy in the saline solution for BioDur
316LS, 0.200 to 0.250 VscE, was slightly more noble than that for IVT 38, 0.100 to 0.150
VSCE. In terms of ability to repassivate, BioDur 316LS, Heat 1, had greater resistance to
pitting and crevice attack than IVT 38. Examination of the samples following the F746
test revealed no evidence o f localized corrosion attack, either crevice corrosion attack
around the tapered collar or pitting corrosion attack on the sample face.
189
Table 6 - Results of ASTM F746 Experiments
Sample
BioDur 316LS
Heat 1
IVT 38
Exposed
Area
Area
Under
Collar
cm 2
Initial
Ecorr
VSCF~
Final
Ecort
VscE
VscE
3.62
0.61
-0.177
-0.133
0.200
3.62
0.61
-0.163
-0.124
0.250
3.62
0.61
-0.177
-0.117
0.200
3,62
0.61
-0,171
j -0.093
0.150
3.62
0.61
-0.164
-0.102
0.100
3.62
0.61
-0,221
-0.164
0.150
cm 2
Eprot
Discussion
Composition of alloys
The PERSS alloy IVT 78 exhibited substantially greater resistance to pitting corrosion
than the IVT 50-series alloys. The value of Ecor~was substantially more noble and more
clearly in the passive region than that for the IVT 50-series alloys. The value Of Eb for
IVT 78 was more noble than that for the IVT 50-series alloys and comparable to that of
BioDur 316LS. The value of Ieo~ for IVT 78 was more than a factor of 5 lower than that
of the IVT 50-series alloys and comparable to that ofBioDur 316LS. The value of(EbEeoc) for IVT 78 was substantially higher than most values for the IVT 50-series alloys
and comparable to that for BioDur 316LS.
Results in Table 5 show that Eb values for 1VT 78 were similar to those for BioDur
316LS. Values for the IVT 50-series alloys were less noble than those for BioDur
316LS. A reduction in Eb similar to that seen for the IVT 50-series alloys was suggested
in McGill's review [5-6] of the literature for two ferritic stainless steels, although no data
were reported for austenitic stainless steels such as UNS $31673. Leclerc [10] has
reported Eb values of 0.160-0.650 VSCEfor 316 stainless steel in Ringers solution. Eb
values slightly more noble to these are reported in Table 5 for BioDur 316LS and for IVT
78. Values comparable to them are reported in Table 5 for the IVT series-50 alloys.
Another way to assess pitting corrosion susceptibility is to consider the derived
parameters given in the last two columns of Table 5. A welt-defined hysteresis loop
indicates susceptibility to localized corrosion, and the width of the hysteresis loop (EbEprot) characterizes that susceptibility (see for example Figure 4). A larger Eb-Eprotvalue
generally indicates greater susceptibility to localized corrosion [17]. (Eb-Eprot)values in
Table 5 for both BioDur 316LS mad the PERSS alloys are similar in magnitude,
indicating they have a similar susceptibility to localized corrosion. However,
susceptibility to localized corrosion does not mean pitting corrosion will initiate.
190
Susceptibility to pit initiation is indicated by the second parameter, (Eb-Er

The greater
the value o f (Eb-Eco,~), the lower the probability that pitting corrosion will initiate on an
alloy surface [17]. Values of(Eb-Eeorr) given in Table 5 are large (0.334 to 0.852 V) for
all of the alloys, indicating low susceptibility to pit initiation. Furthermore, both BioDur
316LS and IVT 78 had values on the high end o f this range, indicating even greater
resistance to pit initiation.
None o f the measured corrosion parameters for the PERSS alloys in deaerated Ringers
solution were a function of alloy oxygen content between 0.01 and 0.03 wt %. Results in
other environments may be different. However, the fabrication practice for the PERSS
alloys is to keep the oxygen content as low as possible.
Surface finish and possible effects

Surface preparation o f the BioDur 316LS alloy and the IVT 30-series and 50-series alloys
was a 120-grit finish, which is much rougher than that of finished stents. Surface
preparation of the BioDur 316LS, Heat 2, and IVT 78 was similar to that o f
electropolished (finished) stents. Surface finish can greatly affect the resistance of an
alloy to localized corrosion. For example, the F2129 results for BioDur 316LS were
quite different for Heat 1 ground to a 120-grit surface finish and Heat 2 polished to a 1
~tm finish and then electropolished. A surface that is more homogeneous, both
chemically and physically, will have a more noble value for Eb and a better resistance to
localized corrosion [ 16]. Physical homogeneity of an alloy surface refers to roughness
caused by the surface preparation process. Eb has been shown to be more noble as peakto-peak surface roughness decreased from 100 to 0.02 ~tm [16]. While chemical
homogeneity of an alloy surface can mean many things, one of the most critical factors is
the amount of Cr in the passive surface film. Studies have shown that the amount of Cr
in the passive film, and value of Ecor~and Eb, increased as follows: dry polishing < wet
polishing < HF etching < 10% HNO3 etching < 30% HNO3 passivation [16]. Clearly
surface finish is an important factor and should be considered when evaluating alloy
resistance to localized corrosion. The optimum surface finish is that o f the finished stent,
as was done with the BioDur 316LS, Heat 2, and the PERSS alloy IVT 78, and is
reflected in the higher Ebvalues for these two materials, Table 5.
Potentiodynamic polarization scan rate

Scan rate can affect measured values Of Eb and Eprot [17]. Faster scan rates tend to
produce more noble values Of Eb and Eprot. Scan rate also can affect different alloys
differently. The scan rate used in the present tests was 10 mV/min, well within the range
where Eb for 18Cr-9Ni stainless steels is reported to be unaffected by scan rate [17].
191
Potentiodynamic polarization scan versus F2129 scan
The potentiodynamic polarization and F2129 scans for BioDur 316LS, Heat 2, are given
in Figure 9. The F2129 scan is seen to be identical to that of the potentiodynamic
polarization scan over the potential range covered by the F2129 test criteria (the lower
value of current after the scan direction was reversed is not considered to be significant).
However, for materials like BioDur 316LS, Heat 2, and IVT 78 with an electropolished
surface finish, the F2129 test did not permit a determination ofEco=, Icorr, or Eprot. The
difficulty lies in the way Ecorr and the scan start potential are determined in the test for a
material that is strongly passive, as both BioDur 316LS and the IVT 78 alloys were.
With a strongly passive material the "last measured Eeorr"carl vary widely over the
passive region. The F2129 test limits the cyclic potential scan from a point 100 mV more
negative than "the last measured Eco~" to Eb and back. Clearly from the results in Figure
8, this range is insufficient to measure some of the important parameters used to rate the
performance of alloys in body fluids.
1.2
Potentiodynamic
0.8
r~
0.4
m
o
0.0
-0.4
1.E-10
1.E-08
1.E-06
1.E-04
Current density, A/cm
1.E-02
Figure 9 - Comparison ofF2129 and potentiodynamic polarization
results for Biodur 316LS, Heat 2
Conclusions
The PERSS alloy IVT 78 alloy had a resistance to pitting corrosion in deaerated
Ringers solution that was similar to that of the BioDur 316LS base (UNS $31673).
192
9
9
9
*
9
9
Metallographic examination showed annealed IVT 78 and annealed BioDur 316LS to

be single-phase alloys with no evidence of inclusions or precipitates.
Values of (Eb-Eeor~), which characterize the probability that localized corrosion will
initiate, were fairly large and were similar for both BioDur 316LS and IVT 78,
suggesting the probability was low that pits would initiate in deaerated Ringers
solution.
The IVT 78 alloy was more resistant to pitting corrosion in deaerated Ringers
solution than were the IVT 50-series alloys.
The corrosion performance of the IVT 50-series alloys, as measured in deaerated
Ringers solution by the F2129 test procedure, was not a function of alloy oxygen
contents between 0.01 and 0.03 wt %.
There was no evidence of intergranular attack due to sensitization of either BioDur
316LS or the IVT 30-series alloys as measured by the A262E test procedure.
There was no evidence of pitting or crevice attack on the BioDur 316LS and IVT 38
alloys as measured by the F746 test procedure. The IVT 38 results were similar to
those obtained by F2129 for IVT 50-series alloys.
Potentiodynamic polarization scans provide more information on the corrosion
performance of alloys that are strongly passive in body fluids than does F2129. This
arises primarily because of the way Ecor~and the scan start potential is established in
F2129. Potentiodynamic polarization measurements provided values for Er
Ir
and Eprot for electropolished BioDur 316LS and IVT 78. These values were not
available from F2129.
References
[1]
Pepine, C.J. and Holmes, D.R, "Coronary Artery Stems," Joumal of the American
College of Cardiology, Vol. 28, No. 3, 1996, pp. 782-94.
[2]
Sahagian, R., "Critical Insight: Marking Devices with Radiopaque Coatings,"
Medical Device and Diagnostic Industry, May, 1999, 6 pp.
[3]
Craig, C. H., Radisch, H. R:, Jr., Trozera, T. A., Turner, P. C., Govier, R. D.,
Vesely, E. J., Jr., Gokcen, N. A., Friend, C. M., and Edwards, M. R., "Development of a
Platinum-Enhanced Radiopaque Stainless Steel (PERSS| '' Stainless Steels for Medical
and Surgical Applications, ASTM STP 1438, G. L. Winters and M. J. Nutt, Eds., ASTM
International, West Conshohocken PA, 2003.
[4]
Dennis, J. Z., Craig, C. H., Radisch, H. R., Jr., Pannek, E. J., Turner, P. C., Hicks,
A. G., Jenusaitis, M., Gokcen, N. A., Friend, C. M., and Edwards, M. R., "Processing
Platinum Enhanced Radiopaque Stainless Steel (PERSS | for Use as Ballon-Expandable
Coronary Stents," Stainless Steels for Medical and Surgical Applications, ASTM STP
1438, G. L. Winters and M. J. Nutt, Eds., ASTM International, West Conshohocken PA,
2003.
[5]
McGill, I. R., "Platinum Metals in Stainless Steels - A Review of Corrosion and
Mechanical Properties," Platinum Metals Review, Vol. 34, No. 2, 1990, pp. 85-97.
[6]
McGill, I. R., "Platinum Metals in Stainless Steels - Part II: Further Corrosion
and Mechanical Properties," Platinum Metals Review, Vol. 34, No. 3, 1990, pp. 144-154.
193
[7]
Andersson, O.H. and Kangasniemi, I., "Calcium Phosphate Formation at the
Surface of Bioactive glass In Vitro," Journal of Biomedical Materials Research, Vol. 25,
1991, pp. 1019-1030.
[8]
Canham, L.T., Reeves, C.L., Newey, J,P., Houlton, M.R., Cox, T.I., Buriak, J.M.,
and Stewart, M.P., "Derivatized Mesoporous Silicon with Dramatically Improved
Stability in Simulated Human Blood Plasma," Advanced Materials, Vol. 11, No. 18,
1999, pp. 1505-1507.
[9]
Sella, C., Martin, J.C., Lecoeur, J., Bellier, J.P., and Davidas, J.P, "Corrosion and
Protection of Surgical and Dental Metallic Implants," Biomaterials and Clinical
Applications, Elsevier Science Publishers, Amsterdam, The Netherlands, pp. 119-124,
1986.
[10] Leclerc, M.F., "Surgical Implants," Corrosion - Volume 1: Metal~Environment
Interactions, Shreir, L.L., Jarman, R.A., and Burstein, G.T., eds., 3rd edition,
Butterworth-Heinemann Ltd., Oxford, UK, 1994, pp. 2:164-2:180.
[11] Szklarska-Smialowska, Z., Pitting Corrosion of Metals, National Association of
Corrosion Engineers, Houston, TX, 1986, pp. 39-40.
[12] Qvarfort R., "New Electrochemical Cell for Pitting Corrosion Testing," Corrosion
Science, vol. 28, No. 2, 1988, pp 135-140.
[13] Szklarska-Smialowska, Z., op.cit., pp. 3-4.
[14] ibid., p. 69
[15] Baboian, R. and Haynes, G.S., "Cyclic Polarization Measurements Experimental Procedure and Evaluation of Test Data," Electrochemical Corrosion
Testing, ASTM STP 727, American Society for Testing and Materials, West
Conshohocken, PA, 1981, pp. 274-282.
[16] Szklarska-Smialowska, Z., op.cit., pp. 240-246.
[17] ibid., pp. 50-57.
Lyle D. Zardiackas, 1 Michael Roach,~ Scott Williamson, ~and Jay-Anthony Bogan ~

Comparison of Corrosion Fatigue of BioDur | 108 to 316L S.S. And 22Cr-13Ni-$Mn
S.S.
Reference: Zardiackas, L. D., Roach, M., Williamson, S., and Bogan, J. -A.,
"Comparison of Corrosion Fatigue of BioDur | 108 to 316L S.S. And 22Cr-13Ni5Mn S.S.," Stainless Steelsfor Medical and Surgical Applications, ASTM STP 1438,
G.L. Winters and M.J. Nutt, Eds., ASTM International, West Conshohocken, PA 2003.
Abstract: Fatigue testing was performed on three austenitic stainless steels in tensiontension in distilled water and in Ringer's solution at 37~ using an MTS servo-hydraulic
testing system following the guidelines of ASTM F1801. The materials chosen were
BioDur | 108, BioDur | 316LS and BioDur | 22Cr- 13Ni-5Mn stainless steel. Samples
with a 2.5 mm diameter in the gauge length of 10 mm were prepared from 8 mm diameter
bar stock using low stress grinding techniques. A minimum o f three samples at each of
five load levels were cycled to failure or to a maximum of 106 cycles. S/N curves for
each material under each condition were plotted and fracture surfaces were examined
using scanning electron microscopy (SEM). Comparison between the fatigue response of
each material in distilled water and Ringer's solution as well as between alloys was
evaluated. Results showed typical S/N curve response comparison of stainless steels in
distilled water vs. salt solutions with curves slightly shifted in a positive direction (higher
numbers o f cycles to failure at equal stress values) in distilled water as compared to
Ringer's solution. Additionally, although not statistically evaluated, the fatigue curves
were shifted toward higher numbers of cycles to failure for the two nitrogen strengthened
alloys with the number of cycles to fracture for 22Cr-13Ni-5Mn being slightly higher in
distilled water at lower stress values and the fatigue curves for 22Cr-13Ni-5Mn and
BioDur | 108 in Ringer's solution were essentially equivalent. Evaluation of fracture
surfaces by SEM showed variations in fatigue striation spacing as a function of loading in
all alloys. A difference in fracture morphology of the BioDur | 108 was noted near the
area of crack initiation as compared to the other two alloys. Flutes and terraces, not
normally seen in fatigue of austenitic stainless steels at in vivo temperature, were seen.
Keywords: Stainless steel, fatigue, corrosion fatigue, implant materials
~Professor and Coordinator o f Biomaterials and Professor of Orthopaedic Surgery, materials

engineer, senior materials engineer and senior materials engineer, respectively,
University o f Mississippi Medical Center, 2500 North State Street, Jackson, MS,
39216.
194
www.astm.org
ZARDIACKAS ETAL. ON BIODUR 108
195
Introduction
Austenitic stainless steels have enjoyed a wide variety of applications as biomedical
implants. A spectrum of mechanical properties achieved by annealing, cold working and
forging, good corrosion resistance and biocompatibility have contributed to their long
term use. Retrieval studies [1-8] and fundamental research [9-11] conducted on
orthopaedic implants as well as stainless steel alloys in the1960s and 1970s, indicated the
need for compositional and processing changes to produce an implant quality 316L
stainless steel with greater corrosion resistance and a lower concentration of inclusions.
Since that time, significant changes in the composition requirements of 316L stainless
steel, as reflected in the ASTM standards for "Wrought 18 Chromium-14 Nickel-2.5
Molybdenum Stainless Steel Bar and Wire for Surgical Implants (UNS $31673) TM (F13800) and "Wrought-18 Chromium-14 Nickel-2.5 Molybdenum Stainless Sheet and Strip
for Surgical Implants (UNS $31673) 1'' (F139-00), have occurred. The most significant
of these changes include increases in the lower and upper limits of Ni, a decrease in the
upper limit of Cr and a decrease in the upper limit of Mo. In order to meet the PRE
requirement of 26 or greater and to reduce the possibility of 6-ferrite formation, alloy
manufacturers raised the Mo concentration to near the maximum (3.0) and kept the Cr
concentration near the middle of the allowed range. These changes have resulted in
superior mechanical and corrosion properties as compared to their predecessors.
Additionally, 22Cr-13Ni-5Mn stainless steel which is covered by ASTM standard for
"Wrought Nitrogen Strengthened-22 Chromium-12.5 Nickel-5 Manganese-2.5
Molybdenum Stainless Steel Bar and Wire for Surgical Implants TM (F1314-95), which is a
nitrogen strengthened austenitic stainless steel, has been added as another implant quality
stainless steel for use especially in the area of orthopaedic fracture fixation.
As changes in composition and properties of 316L stainless steel have occurred,
additional studies have reported on the results of implant retrieval [12-17]. These studies
have shown that there continues to be some crevice and fretting corrosion which occurs in
retrieved implanted devices of 316L stainless steel but that fracture of these devices, in
essentially all cases, is due to fatigue. Basic research studies on the fatigue and corrosion
fatigue of 316L stainless steel have shown that the fatigue resistance of this alloy is
reduced when subjected to fatigue in NaC1 solutions [11, 18-20] as compared to air or
distilled water. Results of fatigue in air, generally using rotating beam samples or
samples fatigued under bending loads, have indicated that the fatigue strength of 316L
and 22Cr-13Ni-5Mn stainless steel is proportional to the yield strength and that higher
fatigue endurance limits are achieved by increased cold working [21-22]. Studies have
also shown that the percentage of increase in fatigue resistance of 22Cr- 13Ni-5Mn
stainless steel as a function of cold working is not linear but drops off as the percentage
of cold working increases. Failure analysis of both retrieved implants and in vitro fatigue
fractures have shown no difference in failure mechanism or fracture surface morphology
as a function of test media in these alloys [17-18].
While providing excellent properties, implant quality austenitic stainless steels such as
316L and 22Cr-13Ni-5Mn stainless steel have been questioned by some as relates to their
biocompatibility. Research has shown that the release of elements in vivo, particularly
nickel and chromium, may lead to inflamation, toxicity or hypersensitivity over time [23-
196
27]. Questions related to biocompatibility have been one of the primary factors initiating
research on the development of low Ni containing austenitic stainless steels.
The purpose of this research was to characterize and compare the corrosion fatigue of
a new low Ni, austenitic stainless steel (BioDur| 108) to implant quality 316L stainless
steel and implant quality 22Cr-13Ni-5Mn stainless steel using a tension-tension fatigue
model as described by ASTM standard for "Corrosion Fatigue Testing of Metallic
Implant Materials b' (F 1801-97) [28]. Comparison was performed by the generation of
S/N curves and through analysis of fracture surface morphology by scanning electron
microscopy.
The low nickel stainless steel chosen for the study was BioDur | 108 (alloy A)
supplied by Carpenter Specialty Alloys. The materials chosen for comparison were
BioDur | 316LS (alloy B) and BioDur | 22Cr-13Ni-5Mn (alloy C) since they were
available from the same source and represented the other two commonly used austenitic
stainless steel alloys in compliance with ASTM F 138 and ASTM F 1314, respectively.
Alloys were supplied as 8mm diameter centerless ground bar in the cold worked
condition. In order to verify information supplied on primary certification documents and
to establish baseline mechanical properties for fatigue testing, each alloy was
characterized for composition, microstructure and tensile properties. Quantitative
compositional evaluation of the major alloying elements of each alloy was determined
using an inductively coupled plasma spectrometer (ICP) with a spark attachment for
analysis of solid samples (SPECTRO z ICP with LISA). The microstructure of each alloy
material was evaluated according to ASTM standard for "Determining Average Grain
Size h' (El 12-96) with all samples prepared in duplicate. The inclusion content was
determined according to ASTM standard for "Determining the Inclusion Content of
Steel 1'' (E45-97) Method A using Plate I-r. Prior to fatigue testing, smooth tensile
samples with a gauge length of 36 mm, a diameter in the gauge length of 6 mm and a
maximum surface roughness of 16 micro-inches (Ra=16) in the gauge were prepared by
Low Stress Grindk Testing was performed using an MTS 4 servo-hydraulic test system
operating at 0.3 mm/min to yield and 3.0 ram/rain from yield to fracture according to
ASTM standard for "Tension Testing of Metallic Materials 1'' (E8-00b). Ultimate tensile
strength, yield strength, elastic modulus, % elongation to fracture, and reduction of area
were determined in order to establish baseline data for fatigue testing.
Corrosion fatigue was performed in tension-tension in aerated distilled/de-ionized
water and in aerated Ringer's solution at 37~ using an MTS servo-hydraulic testing
system. Testing followed the guidelines of ASTM F 1801. Samples were prepared in an
identical manner to the smooth tensile samples previously described except that the
diameter in the gauge length of 10 mm was 2.5 mm (Figure 1). Samples were mounted
vertically to facilitate complete submersion of the sample through the gauge section, and
all fixturing in contact with the solution was either teflon or the same alloy as the samples
to eliminate any corrosion couples. A tension-tension load was applied as a sinnsoidal
2SPECTRO Analytical Instruments, Fitchburg, MA, USA
3LowStress Grind, Cincinnati, OH USA

4MTS, Eden Prairie, MN, USA
ZARDIACKAS ET AL. ON BIODUR 108
197
wave function in load control at 1 Hz. A minimum of three samples was tested at each of
five stress levels to failure or to a maximum of 10 6 cycles. Data was plotted as stress v.
number of cycles to fracture to generate typical S/N curves for each alloy under each of
the two test conditions.
Ra: 16 pinGauge
~-- 9.5mm.~, ~4,- 9.5mm--,~
lO,OOrnm--.~ ~-.,9.5mm,-~, ~~,-,-,9.5mm-,-~
8,oomm
Figure 1 - Sample drawingfor corrosionfatigue samples
After fatigue testing was completed, analysis of the fracture surfaces from at least one
sample of each alloy in each test solution which fractured at less than 105 cycles and
one sample of each alloy in each test solution that fractured at greater than 105 cycles was
examined by scanning electron microscopy (SEM). Comparison was made between
alloys to determine the area of crack initiation, fracture morphology during Stage II
fatigue crack propagation, and the final fracture mode (Stage III fatigue). Additionally,
comparison was made to identify differences in fracture surface morphology between
fatigue in distilled/de-ionized water and Ringer's solution for each alloy.
Results
Results of the baseline evaluation of each of the three alloys including composition,
microstructure and mechanical testing are reported elsewhere [29-30] and given for
informational purposes in Table 1. S/N plots of the results of the fatigue testing to
compare the fatigue behavior of each of the three stainless steel alloys evaluated in each
of the two test media are seen in Figures 2 and 3. It must be remembered that the
endurance limit for these alloys has not been statistically determined, and therefore, the
values interpreted from the S/N curves are not statistically precise. As seen in Figures 2
and 3, the results indicate that in distilled/de-ionized water the fatigue curves cross
with alloy A showing the greatest number of cycles to failure at the higher loads and alloy
C showing the greatest number of cycles to failure at the lower loads. However, in
Ringer's solution the curves are essentially superimposed. The fatigue strength of both
alloy A and alloy C are greater than alloy B regardless of the solution used during testing.
These results are essentially consistent with the data reported for fatigue testing in air
regardless of testing methodology, although the actual fatigue strength numbers are not as
great as those reported in most of the literature where testing has been performed in
reverse bending [21,22].
198
Table I - Mechanical Testing Results (n =5)

Alloy
Alloy A
Alloy B
Alloy C
Elastic Modulus (GPa)
188
167
186
UTS* (MPa)
1344
1014
1351
0.2% YS** (MPa)
1179
793
1082
% ROA***
68
71
60 1
% Elongation
35
25
19
*UTS = ultimate tensile strength

**YS = 0.2% yield strength
***ROA = reduction of area
Fracture analysis revealed a classical macroscopic fatigue pattem for samples fatigued
in tension-tension with no stress concentration [31 ]. This pattern was consistent
regardless of the alloy and the percentage of the fractured sample which experienced
Stage II fatigue increased with decreasing load (Figure 4). The remainder of each
fracture surface showed a Stage III fatigue fracture pattern (ductile overload). SEM
fracture analysis of alloy B and alloy C showed crack initiation at the surface of the
samples with the fracture morphology being typically fatigue striations with secondary
cracking (Figures 5a and 5b). The striations were less clearly defined in some areas of
the fracture surfaces in the samples tested in Ringer's solution presumably due to
corrosion effects. This type of morphology is similar to that seen in retrieved fracture
fixation devices [14, 17]. Final fracture of each sample was due to tensile overload as
exhibited by ductile dimples (Figure 5c). While this type of morphology is typically
seen in test samples, dimpling, due to ductile tensile or shear overloading, is very often
not seen in retrieved devices where there has been some degree of bone healing which
results in load sharing.
SEM fracture analysis of the fatigued samples of alloy A showed a more complex
pattern with the morphology deviating from that typically seen in the other two FCC
austenitic stainless steels evaluated. While fatigue striations were noted over much of the
fracture surfaces regardless of testing media or stress level (Figure 6a), fluting, terraces
with feather marks, and furrowing (Figures 6b and 6c) were noted especially in the areas
near crack initiation.
ASTM F1801 requires that a minimum of three samples be tested at each of five load
levels to describe the fatigue curves. This same specification recommends that a
minimum of six samples be tested at each load level to evaluate statistical significance. It
has been our experience that ten samples are often required to perform statistical tests.
Figure 2 -
Figure 3 -
S/N curves of alloys A, B, and C in distilled~de-ionized water
S/N curves of alloys A, B, and C in Ringer's solution
199
200
Figure 4 -
SEM micrograph showing the fracture surface of alloy A in

distilled~de-ionized water (30X, 713,015 cycles)
Figure 5a -
SEM micrograph of the fracture surface of alloy B in Ringer's

solution showingfatigue striations with secondary cracking.
Figure 5b -
SEM micrograph of the fracture surface of alloy C in Ringer's

solution showingfatigue striations with secondary cracking.
Figure 5c -
Representative SEM micrograph of ductile dimples due to tensile

overload in the final fracture region of alloy B in Ringer's
solution.
201
202
Figure 6a -
Figure 6b -
SEM micrograph showing the fatigue striations on alloy A in

Ringer's solution.
SEM micrograph showing fluting, terraces, and furrowing on the

fracture surface of alloy A near crack initiation (Fatigue in
distilled~de-ionized water).
203
Figure 6c - SEM micrograph showing terraces with feather marks on
the fracture surface of alloy A near crack initiation

(Fatigue in Ringer's solution).
Due to the large number of samples required and thus the overwhelming machine time to
statistically compare the fatigue of the three alloys, especially when testing at the
recommended 1 hz, the initial testing regimen using three samples at each load level as
described by ASTM Ft801 was followed. Of significant interest is the comparison of the
S/N curves between alloys in each of the two test solutions (Figures 2 and 3). These
results show that in both Ringer's solution and distilled/de-ionized water the number of
cycles to fracture at a given stress value are greater for the two nitrogen strengthened
alloys as compared to alloy B. However, whilethere is a difference, it is not as great as
anticipated in light of the much greater yield strength of alloy A (1179 MPa) and alloy C
(1082 MPa) as compared to that of alloy B (793 MPa) and results reported elsewhere
[21,22]. Because the differences seen between the fatigue curves of alloy A and alloy C
are slight, they may or may not be seen if the testing were repeated using the same alloys
from different lots with essentially identical properties. Information taken from the
Carpenter data sheets as well as other sources indicates that an increase in fatigue strength
is not necessarily proportional to the increase in tensile and yield strength especially in
these nitrogen strengthened stainless steels which have been cold worked to achieve even
higher strength values. It has been our observation that even though alloy A has high
elongation to fracture, especially for material that has been highly cold worked to achieve
high strength, the work hardening rate during deformation is very high and the amount of
work hardening to fracture is also high. The work hardening rate and increase in tensile
strength properties for the same percentage of cold working, including both yield strength
and ultimate tensile strength, of the two nitrogen strengthened alloys is significantly
greater than that of alloy B [32]. Through discussions with LSG and machinists'
204
fabricating fixtures, as well as Carpenter, we have learned that alloy A cold works even
more rapidly than does alloy C and much more rapidly than does alloy B. This
phenomenon significantly increases the care required during machining of the material.
However, these observations are even more significant in explaining the results of the
fatigue data. Since fatigue cracks initiate and propagate due to applied or resolved tensile
stresses, and the tensile stress is additive to any residual tensile stresses, any test regimen
that continues to apply tensile stress and cold work the material in tension should result in
increased fatigue damage (crack initiation and propagation). Because the nitrogen
strengthened alloys are especially susceptible to cold working, cycling in a tensiontension mode could accelerate fatigue as compared to alternating tension and
compression. If the material were not so highly cold worked, the relative fatigue strength
to the yield strength might be increased. Even though the samples that were evaluated
during the research reported here were prepared using low stress grinding procedures, the
susceptibility of alloy A to very rapid cold working could, at least in part, contribute to
the slightly lower than anticipated fatigue strength. If the surfaces were prepared with
compressive residual stresses, the addition of applied tensile stresses should not result in
as great a total tensile stress during loading. Previous unreported research in our
laboratories comparing the bending fatigue of 316LS and 22Cr- 13Ni-5Mn hip screw
plates with a variety of surface finishes including shot peening and electro-polishing
indicates improved fatigue properties of plates that are shot peened. Additionally,
preliminary results of ongoing research related to fatigue testing of fracture fixation
devices of alloy A indicates that surface treatment such as shot peening, which increases
compressive residual stress, results in significantly greater fatigue strengths than expected
from the results obtained during this study. Further research is also underway to
examine the fatigue behavior and endurance limit of alloy A which has only been cold
worked to achieve strength values in the range of those of alloy B evaluated during this
project.
The results, comparing the fatigue of all three alloys in the two media (Figures 2 and
3), corroborate the results reported by others concerning the effect of NaC1 containing
solutions on the reduction of the number of cycles to fracture at a given stress value. The
effect of NaC1 containing solutions on the fatigue of the two nitrogen strengthened alloys
might be even more clearly delineated if the endurance limits were statistically evaluated
using the 'two point method' [33]. The use of NaC1 solutions results in a negative
displacement of the entire fatigue curve of austenitic stainless steels [ 11,22]. The
consistency of the trend in the affect on the fatigue performance in the two different test
solutions on each of the alloys is to be expected because the crystalline structure and
mechanism of resistance to corrosion (passivation by Cr-O production) is the same for all
three alloys. Additionally, little difference in the corrosion characteristics of the three
alloys but especially the two nitrogen strengthened alloys (alloys A and C) occurs within
the testing parameters of this experiment and as documented by other research [29].
It is well understood that fatigue is a brittle rather than ductile fracture mode, and that
Stage II fatigue crack propagation of austenitic stainless steels generally exhibits a
striation type of morphology when fatigue occurs at strain rates and temperatures in the
range of those imposed here. Deformation that includes terraces, flutes and furrowing is
often attributed to a more brittle fracture or fatigue of nitrogen strengthened austenitic
stainless steels [34] at low temperatures or fracture of metals with far fewer slip systems
205
such as titanium, especially when a complex state of stress is present [ 17]. Because this
type of morphology is only present in the samples of alloy A and not present iri the
fatigued samples of alloy C, which is also nitrogen strengthened, it is hypothesized that
this type of fatigue fracture morphology may be as a result of the cold working behavior
of alloy A as previously described. Cold working during manufacturing as well as during
the testing could lead to pilming of dislocations, especially due to the presence of a high
concentration of nitrogen interstitials in the FCC lattice. This phenomenon could lead to
a more brittle type of fracture, especially in areas where the material has experienced
significant cold working creating localized high intensity stress fields or where the tensile
stresses have been built up prior to crack initiation. These observations on the fracture
surfaces reinforce the previous conclusions made concerning the trends observed in the
S/N curves.
In conclusion, the results reported here indicate that based upon the fatigue
performance, alloy A is a candidate for applications where 316LS or 22Cr-13Ni-5Mn
stainless steel have been used previously. This alloy could be especially useful as an
implant material for devices subjected to higher loading, especially if the hypothesis
concerning the effects of compressive residual stress and the ability of altering the surface
to increase the fatigue strength are accurate. The effects of surface treatments such as
shot peening and electro-polishing, which removes areas of high surface tensile stress
concentration, should serve to improve the fatigue characteristics of devices produced
from this alloy.
Acknowledgment
This research supported by a grant from Carpenter Technology Corporation.
References
[1 ]
[2]
[3]
[4]
[5]
[6]
Cahoon, J.R., Paxton, H.W., "Metallurgical Analyses of Failed Orthopedic Implants,"

Journal Of Biomedical Materials Research, Vol. 2, 1968, pp. 1-22.
Colangelo, V.J., "Corrosion Fatigue in Surgical Implants," Journal of Basic
Engineering, December 1969, pp. 581-586.
Dumbleton, John H., Miller, Edward H., "Failures of Metallic Orthopaedic Implants,"
Metals Handbook 8th Edition, Howard E. Boyer, Ed., American Society for Metals,
Vol. 10, 1975.
Gray, R.J. "Metallographic Examinations of Retrieved Intramedullary Bone Pins and
Bone Screws from the Human Body," Journal of Biomedical Materials Research
Symposium, No. 5, John Wiley & Sons, Inc., New York, 1974, pp. 27-38.
Gray, R.J., "Metallographic Examinations of Retrieved Intramedullary Bone Pins and
Bone Screws from the Human Body," Report No. ORNLTM4068, Oak Ridge National
Laboratory, Tennessee, U.S. Department of Commerce, National Technioal
Information Service, February 1973.
Hughes, A.N., Jordan, B.A., "Metallurgical Observations on Some Metallic Surgical
Implants Which Failed in Vivo," Journal of Biomedical Materials, Vol. 6, 1972, pp.
33-48.
206
[7]
[8]
[9]
[ 10]
[11]
[12]
[13]
[14]
[ 15]
[16]
[17]
[18]
[19]
[20]
[21]
Sloter, L.E., Piehler, H.R., "Corrosion-Fatigue Performance of Stainless Steel Hip

Nails - Jewett Type," Corrosion and Degradation of Implant Materials, ASTM STP
684, B.C. Syrett and A. Acharya, Eds., American Society for Testing and Materials,
1979, pp. 173-195.
Weinstein, A., Amstutz, H., Pavon, G., Franceschini, V., "Orthopedic Implants - A
Clinical and Metallurgical Analysis," Journal of Biomedical Materials Research
Symposium, No. 4, John Wiley & Sons, Inc., 1973, pp. 297-325.
Cahoon, J.R., Paxton, H.W., "A Metallurgical Survey of Current Orthopedic
Implants," Journal Of Biomedical Materials Research, Vol. 4, 1970, pp. 223-244.
Imam, M.A., Fraker, A.C., Gilmore, C.M., "Corrosion Fatigue of 316L Stainless Steel,
Co-Cr-Mo Alloy, and ELI Ti-6AI-4V," Corrosion and Degradation of Implant
Materials, ASTM STP 684, B.C. Syrett and A. Acharya, Eds., American Society for
Testing and Materials, 1979, pp. 128-143.
Pohler, O.E.M., Straumarm, F., "Fatigue and Corrosion Fatigue Studies on StainlessSteel Implant Material," Evaluation ofBiomaterials, G.D. Winter, J.L. Leray, K. de
Groot, Eds., John Wiley & Sons, Inc., 1980, pp. 89-113.
Brantina, W.J., Young, S.B., Yue, S., Morgan, M.J., Pilliar, R.M., Wallace, A.C.,
"Fatigue Deformation and Fractographic Analysis of Surgical Implants and Implant
Materials," Proceedings of the International Conference and Exhibits on Failure
Analysis, Montreal, Quebec, Canada, July 8-11, 1991, pp. 299-310.
Brunner, H., Simpson, J.P., 'Tatigue Fracture of Bone Plates," Injury: the British
Journal of Accident Surgery Vol. 11, No.3, pp. 203-207.
Pohler, O.E.M, Straumann, I., "Failures of Metallic Orthopedic Implants,"Metals
Handbook 9thEdition, American Society for Metals, Vol. 11, 1986.
Sivakumar, M., Dhanadurai,K., Rajeswari, S., Thulasiraman, V.,"Failures in Stainless
Steel Orthopaedic Implant Devices: A Survey," Journal of Materials Science Letters,
Vol. 4, Chapman & Hall, 1995, pp. 351-354.
Zardiackas, Lyle D., Black, R. John, Hughes, James L., Reeves, R. Bradley, Jun, J.N.,
"Metallurgical Evaluation of Retrieved Implants and Correlation of Failures to Patient
Record Data," Orthopedics, Vol. 12, No. 1, January 1989, pp. 85-92.
Zardiackas, Lyle D., Dillon, Lance D., "Failure Analysis of Metallic Orthopedic
Devices," Encyclopedic Handbook of Biomaterials and Bioengineering, Part B:
Applications, Vol. 1, Donald L. Wise, Debra J. Trantolo, David E. Altobelli, Michael
J. Yaszemski, Joseph D. Gresser, Edith R. Schwartz, Eds., 1995, pp. 123-170.
Nakajima, M., Akatsuka, Y., Tokaji, K., "Fatigue Crack Initiation and Growth of
Stainless Steels in 3% NACL Solution," Advances in Fracture Research, Vols. 1-6,
1997, pp. 1621-1628.
Nakajima, M., Shimizu, T., Kanamori, T., Tokaji, K., "Fatigue Crack Growth
Behaviour of Metallic Biomaterials in a Physiological Environment," Fatigue &
Fracture of Engineering Materials & Structures, Vol. 21, Blackwell Science Ltd.,
1998, pp. 35-45.
Yu, J., Zhao, Z.J., Li, L.X., "Corrosion Fatigue Resistance of Surgical Implant
Stainless Steels and Titanium Alloy," Corrosion Science, Vol, 35, Pergamon Press
Ltd., Great Britain, 1993, pp. 587-597.
Black, Jonathan, Orthopaedic Biomaterials In Research and Practice, 1988.
ZARDIACKAS ETAL. ON BIODUR 108
207
[22] Shetty, Ravi H. and Ottersberg, Walter H., "Metals in Orthopedic Surgery,"
Encyclopedic Handbook of Biomaterials and Bioengineering, Part B: Applications,
Vol. 1, Donald L. Wise, Debra J. Trantolo, David E. Altobelli, Michael J. Yaszemski,
Joseph D. Gresser, Edith R. Schwartz, Eds., 1995, pp. 509-540.
[23] Vieweg U., van Roost D., Wolf H.K., Schyma, C.A., Schramm, J. "Corrosion On
Internal Spinal Fixator System," Spine, Vol. 24, 1999, pp. 946-951.
[24] Mofid M.N., Thompson, R.C., Pardo. C.A., Manson, P.N., Wander-Kolk, C.A.,
"Biocompatibilityof Fixation Materials In the Brain,"Plastic Reconstruction Surgery,
Vol. 100, 1997, pp.14-21.
[25] Savarino, L., Stea, S., Granchi, D., Visentin, M., Ciapetti, G., Donati, M.E., Rollo, G.,
Zinghi, G., Pissoferrato, A., Montanaro, L., Toni, A., "Sister Chromatid Exchanges
And Ion Release In Patients Wearing Fracture Fixation Devices," Journal of
Biomedical Material Research, Vol. 50, 2000, pp. 21-26.
[26] Hensten-Pettersen, A., "Replacement of Restorations Based on Material Allergies,"
Quality Evaluation of Dental Restorations: Criteria for Placement and Replacement,
Anusavice K.J., Ed., Quintessence Pub. Co., 1989, pp. 357-372.
[27] Kerosuo, H., Kullaa, A., Kerosuo, E., Kanerva, L., Hensten-Pettersen A., "Nickel
Allergy In Adolescents In Relation To Orthodontic Treatment And Piercing Of Ears,"
American Journal of Orthodontics and Dentofacial Orthopedic, Vol. 109, 1996, pp.
148-154.
[28] Dobbs, H.S., Scales, J.T., "Corrosion Fatigue Testing ofMetallic lmplant Materials,"
Medical Devices and Services, ASfM Vol. 13.01, B.C. Syrett and A. Acharya, Eds.,
American Society for Testing and Materials, 1999, pp. 1289-1294.
[29] Zardiackas, L.D., Roach, M.D., Williamson, S., Bogan, J.A., "Comparison of Anodic
Polarization And Galvanic Corrosion of a Low-Nickel Stainless Steel to 316L S.S. and
22Cr- 13Ni-5Mn S.S.," Stainless Steelsfor Medical and Surgical Applications, ASTM
STP 1438, G.L. Winters and M.J. Nutt, Eds., ASTM International, West
Conshohocken, PA 2003.
[30] Zardiackas, L.D. Roach, M.D., Williamson, S., Bogan, J.A., "Comparison of Notch
Sensitivity and SCC of a Low-Nickel Stainless Steel to 316L S.S. and 22Cr-13Ni-5Mn
S.S.," Stainless Steelsfor Medical and Surgical Applications, ASTM STP 1438, G.L.
Winters and M.J. Nutt, Eds., ASTM International, West Conshohocken, PA 2003.
[31 ] Powell, G.W., "Identification of Types of Failures," Failure Analysis an Prevention,
ASM Handbook, Formerly 9th Edition, Metals Handbook, G.W. Powell and S.E.
Mahmoud, Coordinators, 1986, pp. 75-81.
[32] Carpenter Technology "BioDur| and Other Specialty Alloys for Medical
Applications," Reading, PA.
[33] Little, R.E., "Manual on Statistical Planning And Analysis For Fatigue Experiments,"
ASTMSTP 588, American Society for Testing and Materials, Philadelphia, PA, 1975.
[34] Vogt, J.B., "Fatigue Properties of High Nitrogen Steels," Journal of Materials
Processing Technology, Vol. 117, 2001, pp. 364-369.
Wear and Corrosion-Related Wear
Markus Windier] Julie E. MacDougall, 2 and Rolf Schenk2
Investigation Into Wear-Induced Corrosion of Orthopaedic Implant Materials
Reference: Windier, M., MacDougall J. E., and Schenk, R., "Investigation Into WearInduced Corrosion of Orthopaedic Implant Materials," Stainless Steels for Medical
and Surgical Applications, ASTMSTP 1438, G. L. Winters and M. J. Nutt, Eds., ASTM
Intemational, West Conshohocken, PA, 2003.
Abstract: The purpose of this study was to compare the corrosion behaviour under
sliding wear of four materials commonly used for surgical implants: Ti-6AI-7Nb,
Co-28Cr-6Mo, and two stainless steels, high-nitrogen stainless steel (Fe22Crl 0NiN) and
AISI 316L. A machine was constructed to simulate sliding wear, and the current density
was measured to compare the response of each material to wear and to study the effect of
pH (varied from 1 to 6). The results indicate that titanium exhibits higher current densities
during wear than other materials tested over the pH range of 2.5 to 6. Additionally, pH
has a greater effect on the corrosion behaviour of the two stainless steels than on the
titanium or cobalt alloys.
Keywords: high nitrogen stainless steel, stainless steel, CoCr alloy, titanium alloy,
corrosion behaviour, sliding wear, repassivation
Introduction
Currently, a major issue in orthopaedic implant design (e.g. hip or knee prosthesis)
is wear of metallic components. Primarily, wear occurs on articulating surfaces, such as
on the femoral head of total hip implant. In modular designs, the combination of different
materials and components can promote micromotion between parts, which can result in
1 Director materials research, Material Research Department, Sulzer Orthopedics Ltd.,

P.O. Box 65, 8404 Winterthur, Switzerland.
2 Corrosion scientist and material engineer, respectively, Sulzer Innotec, Sulzer Markets
and Technology AG, P.O. Box 414, 8401 Winterthur, Switzerland.
211
www.astm.org
212
fretting wear and lead to crevice corrosion [1-4]. With cemented hip prostheses, loosening
may occur at the cement/implant interface and can lead to wear of the stem component
[5-7]. In addition, in total hip replacements, wear between articulating components causes
scratching and polishing of the metal surfaces [8]. This phenomenon is well-known in hip
prostheses especially in metal-on-metal articulations where both components are made of
CoCr alloys [9]. Fretting corrosion can also take place in the connection between plates
and screws [ 10].
In all such situations, the passive layer of the metal component is subject to
degradation. If the passive layer is destroyed and bare metal exposed, leading to a rise in
current density, the metal surface must repassivate to prevent corrosion of the material. If
repassivation is inhibited in any way, severe corrosion can result, which can lead to the
failure of the implant. Therefore, due to the importance of maintaining corrosion
resistance in the body, characterizing wear of implant materials or, more specifically,
testing for wear-induced corrosion of such materials is crucial. This is an area of implant
material characterisation that has not been extensively researched to date.
The purpose of this study was to compare the corrosion behaviour under sliding
wear conditions of four materials commonly used for surgical applications: Ti6AI7Nb,
Co28Cr6Mo, and two stainless steels, Fe22Crl 0NiN and 316L stainless steel. The
corrosion behaviour was evaluated at different pH valued from -1.0 to -6.0 as low pH
levels are present during crevice corrosion.
Materials and Procedures
Potentiodynamic testing (without wear) and potentiostatic testing (with wear)

were used to characterize the corrosion behaviour of four implant materials: 316L
stainless steel (ISO 5832-1E), Fe22Cr10NiN high-nitrogen stainless steel (ISO 5832-9),
Co28Cr6Mo (ISO 5832-12), and Ti6AI7Nb (ISO 5832-11). All bar materials were in
wrought and annealed conditions and had diameters from 20 to 22 ram. Each material
was cut into cylindrical coupons. Each coupon was connected to an insulated copper wire
and encased in a cold-curing resin (Technovit 4071, Kulzer GmbH, Wehrheim,
Germany). Its surface was ground to 1200 grit specification. The reference electrode used
in all experiments was the Standard Calomel Electrode (SCE) and all testing was
conducted with Nz gas bubbling.
Potentiodynamic Corrosion Tests (no wear)

In order to characterize the general corrosion behaviour of each material initially
(breakthrough potential, passive current density, etc.), potentiodynamic testing was
performed in the absence of wear (scan rate: 0.167 mV/s). A standard potentiodynamic
corrosion cell (ASTM-G5) was used to measure the effects ofpH and temperature
(Potentiostat: Schmitzo AG, Switzerland, Software: CMS 100, Version 2.40a, Gamry
Instruments Inc., Warrninster, PA, USA). Electrolyte pH was varied from 1.2 to 6 at a
constant temperature of 24~ 0.15mol/L HC1 at pH 1.2; 0.9% NaC1 at pH 4 (pH adjusted
with diluted HC1); and 0.9% NaC1 at pH 6. The temperature was varied from 24~ to
WINDLER ET AL. ON WEAR-INDUCED CORROSION
213
60~ (Julabo HC5/2 water bath, Julabo Inc., Allentown, PA, USA) at a constant pH of 4.
All the potentiodynamic corrosion experiments were performed using one specimen. The
materials were also analyzed for correct chemical composition, physical properties, and
microstructure before they were tested.
Potentiostatic Tests under Sliding Wear Condition

Potentiostatic testing was used to measure the passive current density produced
during sliding wear. The samples were fixed in a PVC frame and placed in a basin filled
with the same electrolytic solutions used in the potentiodynamic tests (Figure 1). A
1.5mm diameter alumina ceramic head was lowered onto the flat sample surface and the
downward stress adjusted to 2.78 MPa using a specially designed spring-controlled
loading system. The counter electrode was platinum wire. Potentiostatic scans were taken
for each sample over the potential range of the passive region. Initially, the material was
allowed to reach a constant passive current density (steady state). Sliding wear was then
simulated by running the flat ceramic head back and forth over the material surface for an
interval, allowing for repassivation between trials. Current density data was recorded
every 10 seconds during the whole experiment. The horizontal movement and vertical
position of the ceramic head were controlled by a motor-driven precision cross-table
(Jacob AG, Switzerland). The length of the scratch was fixed at 7.5ram with linear
potentiometers (LP-200F-5K, Midori Precisions, Japan) and the scratching speed set to
1.5 mm/s (DC-motor GNM 2145, Engel GmbH, Germany). The current-time data were
recorded using CMS-100 software (Gamry Instruments Inc., USA).
Figure 1 - Custom made construction to study the corrosion behaviour during sliding
wear condition
214
The following tests were performed under sliding wear conditions:
Influence of Scratching Time

Only Ti6A17Nb was tested at an applied potential of+100 mV and pH 4. Six
different scratch times were introduced: I0 s, 30 s, 60 s, 300 s, 600 s, 1800 s. All tests
were repeated twice.
Influence of Speed
All materials were tested at an applied potential of 0 mV, pH 4 for a scratch time
of 2 minutes. Three different speed levels were used: 1.1 mrrds, 1.9 mm/s and 2.5 mm/s.
All test were repeated twice.
Influence ofpH
All materials were tested at three applied potentials of-100, 0 and +100 mV, each
for a scratch time of 2 minutes. Four different pH levels: 0.15mol/L HC1 at pH 1 and
0.9% NaC1 at pH 2.5, pH 4, and pH 6 were tested, all tests were repeated twice.
NOTE: In living tissue, at pH-7.4, the typical redox potential is - 0 mV[sczl [11].
Results
Potentiodynamic Corrosion Tests (no wear)

From the potentiodynamic testing, it was found that Ti6AI7Nb exhibited the best
overall corrosion resistance in terms of the breakdown potential (>1.5 V) at any pH level
or temperature, followed by Fe22Cr10NiN, Co28Cr6Mo, and finally 316L stainless steel
(Table 1 and 2). The values of passive current density showed a trend of increasing with a
decreasing pH value. This trend was especially pronounced when the pH varied from 4.0
to 1.2 (the only exception was for Fe22Cr10NiN at pH 1.2).
It was found that all materials, particularly 316L stainless steel, were more
sensitive to changes in pH than to changes in temperature. In fact, temperature had no
significant effect on the corrosion behaviour of any material over the tested range (24~ 60~
It has therefore been proposed that data collected from experiments carried out at
room temperature are reasonable approximations of data generated at slightly higher
femperatures (e.g. at 37~ - the average human body temperature).
The effect of pH was only predominant at very low pH values and the corrosion
behaviour of the test materials did not vary significantly over the pH range of 4 to 6.
215
Table 1 - Influence of Temperature on the Breakdown Potential (Eb) and Passive

Current Density (PCD), measured at pH 4
316L
(ISO 5832-1E)
Fe22Cr 10NiN
(ISO 5832-9)
Co28Cr6Mo
(ISO 5832-12)
Ti6A17Nb
(ISO 5832-11)
Temperature [~
24~
40~
60~
24~
40~
60~
24~
40~
60~
24~
40~
60~
Eb [mV(scE)]
100
-40
-100
1000
1050
1050
620
600
550
1500
1700
1650
PCD [pA/cm2]
0.63
0.44
0.63
0.40
0.79
0.89
0.63
0.63
0.67
0.79
0.84
0.99
Table 2 - Influence of p H on the Breakdown Potential (Eb) and Passive Current Density
(PCD), measured at 24~
316L
(ISO 5832-1E)
Fe22CrlONiN
(ISO 5832-9)
Co28Cr6Mo
(ISO 5832-12)
Ti6A17Nb
(ISO 5832-11)
pH
1.2
4.0
6.3
1.2
4.0
6.3
1.2
4.0
6.3
1.2
4.0
6.3
Eb [mV(scE)]
-60
100
160
840
1000
1000
820
620
600
2000
1500
1700
PCD [pA/cm2]
19.95
0.40
0.38
0.45
0.63
0.56
0.79
0.63
0.56
1.26
0.79
0.79
Potentiostatic Tests under Sliding Wear Condition

Figure 2 shows a typical potentiostatic scan taken of the material Fe22Crl 0NiN
(superimposed) at pH 4 and -200 mV. From this figure three different current densities
can be identified: (1) the passive current density (PCD), which was the current density
before the material was scratched, (2) I peak,which was the maximum average current
density, and (3) delta current density (A CD), which was the difference between I peakand
passive current density. In this study the focus was on the delta current density (A CD),
which was normalized on the total scratched area (T.S.A.) 7.5 mm x 1.5 mm.
216
100
Fe22CrlON~, pH 4, -200mV
<
60
t~
40
Passiv Current ]
Density ~
t
(w/o wear) ~ 1
"I
A CD
,,1peak
20
o ~=~" .~:~....;_=t__~ . . . . _~a_'-0
200
.-~_:-_
400
600
Time [s]
....
800
1000
Figure 2 - Typical potentiostatic scan o f Fe22Crl ONiN at p H 4 and +100m V.

NOTE." The current density is normalized to the total scratched area ~.S.A.) which is
7.5mm x 1.5ram = ll.25mm 2
Influence o f Scratching Time
Figure 3 shows the influence of scratching time on Ti6AI7Nb. No difference could

be demonstrated when varying the scratching time in steps from 10 seconds up to 30
minutes (1800 s). Only the standard deviation increased slightly from 14% (10 s) to 39%
(1800 s).
Influence o f Speed
Increased scratch speed resulted in an increase of delta current density (Table 3).
This phenomenon could be observed with all materials except 316L stainless steel. It is
hypothesized that, as the scratching speed increases, a larger amount of surface area of
bare material is exposed at any given time. The data collected for 316L stainless steel did
not follow the trend observed with the other materials. Other factors (e.g., pitting
corrosion) might have influenced this result.
217
350
300
"7.
<.
Ti6AI7Nb, pH 4, +100 mV [
250
C/3
200
<
150
100
,0
50
0
10
30
60
120
300
600
1800
Scratch Thne [s]

Figure 3 - Influence of the scratching time on Ti6AI7Nb at pH 4 and applied potential of
+100 mV
Table 3 - Influence of speed on the delta current density of
all tested materials at pH 4 and applied potential of 0 m V

Materials
316L
(ISO 5832-1E)
Fe22Cr10NiN
(ISO 5832-9)
Co28Cr6Mo
(ISO 5832-12)
Ti6A17Nb
(ISO 5832-11)
Speed [mm/s]
1.1
1.9
2.5
1.1
1.9
2.5
1.1
1.9
2.5
1.1
1.9
2.5
A CD [~tA / T.S.A.]
194.4
143.6
158.1 19.9
54.6~15.0
84.0~19.2
132.4 + 33.6
77.9 :k 15.1
143.6 15.7
164.8 24.6
111.6+36.5
184.2 + 48.2
217.3 :k 106.1
Influence of pH
Figure 4 shows the average change (A CD) in current density measured during
sliding wear as a function o f p H for all materials tested. The plotted values were
218
calculated by averaging the current densities found over the potential range o f - 1 0 0 m V to
+100mV as listed in Table 4.
450
e
400
INk
350
"'7.
<.
250
<
~.~
200
L)
<~
150
9 316L
- -~ - Co28Cr6Mo
0N iN
300
~3
Ti-6AI7Nb
100
50
i
pH
Figure 4 - Delta current density (A CD) as a function of p H for all materials tested
Potentiostatic testing under sliding wear demonstrated that, as in the
potentiodynamic testing, 316L stainless steel was the most sensitive material to changes
in pH. The most significant difference between non-wear (potentiodynamic) and sliding
wear testing (potentiostatic) lies in the corrosion behaviour of Ti6AI7Nb. In the absence
of wear, Ti6A17Nb showed excellent corrosion behaviour, as quantified by a high
breakdown potential and a passive current density on the same order as those of other
materials tested. However, Ti6A17Nb showed extremely high values of current density
when placed under wear, indicating that the material requires more energy to repassivate
and will corrode faster if subjected to wear. Although, in the absence of wear, Ti6A17Nb
showed slightly higher relative passive current densities (25% more than Fe22Crl 0NiN
and Co28Cr6Mo, see Table 1), the current density recorded for Ti6AI7Nb during wear
was 200% to 300% (see Table 4) larger than those found for the other materials tested.
From Figure 4, it is evident that the corrosion resistant behaviour of 316L stainless
steel is highly dependent on pH. At pH 6, the material exhibits excellent corrosion
resistance under wear; the average ACD measured at pH 6 for 316L stainless steel (30
IaA/T.S.A.) was well below those measured for the other materials. However, as pH is
decreased to 1, the average current density gradually increases and eventually surpasses
those of other materials, with an increase of more than 12 times in ACD. This indicates
that, at low pH, 316L stainless steel performs very poorly against wear-induced corrosion.
Contrarily, the cobalt and titanium alloys showed very little dependence on pH.
Although there is some fluctuation in the results, no definite trend is seen to suggest that
219
p H has a large effect on the potentiostatic behaviour o f these materials under sliding wear
conditions.
Table 4 - Delta current density (A CD) as a function of p H for all tested materials
[mV(scE~l A C D @ p H 6
-I00
174.6
139.4 41.0
0
100
133.9 29.2
Materials
Ti6Ai7Nb
Co28Cr6Mo
Fe22Cr10NiN
316L
ACD@pH4
153.1
205.6 65.3
185.5
A C D @ pH2.5
157.2
177.6 42.2
151.7
ACD@pH1
152.5
168.6 49.3
232.1
Mean
139.0 45.2
181.4 53.9
161.8 60.0
184.2 94.7
-100
0
100
-100
0
100
78.0 l 1.2
102.0 14.8
112.1 8.7
97.9 18.3
46.8 5.7
44.3 13.7
55.4
92.6 13.5
95.1 ~9.4
96.7& 13.0
94.9 12.1
65.1 11.2
67.6 12.5
69.1 13.5
137.4 15.5
112.3 1 7 . 4
i28.0 12.5
125.7 18.2
101.3 6.2
141.7 9.6
121.6 1 0 . 3
39.2 7.3
66.0
119.6 8.9
76.2 36.5
168.2 13.3
147.6 9.1
135.3
Mean
49.0
67.3
121.5
-100
0
100
27.6
33.1
31.3 6.2
98.8
100.6 5.5
120.7 16.4
190.1
177.8 15.1
n.m.
378.25
391.9 17.8
n.m.
Mean
30.6 6.6
103.9 + 14.5
183.7 13.3
384.8 38.3
Mean
150.0
N O T E : n.m. = not possible to be measured because pitting corrosion stated

As the pH is lowered from 6 to 1, Fe22Cr10NiN shows a slight, but steady,
increase in the values found for the average ACD (300%); still, the increases are much
smaller than those found for 316L stainless steel. More important, however, is the fact
that Fe22Crl 0NiN shows the best overall resistance to wear-induced corrosion over the
pH range o f 2.5 to 6. Therefore, even though it is more sensitive to changes in pH than
the cobalt or titanium alloys, Fe22Crl 0NiN produces the lowest values o f current density
over the pH range o f 2.5 to 6 and is surpassed only by the cobalt alloy at low pH.
Thus, as seen from Figure 4, the titanium alloy generally showed the highest
change and Fe22Crl 0NiN generally showed the smallest change in current density during
wear o f all materials tested at any pH. The two exceptions to this statement occured at
low pH where (1) the ACD o f 316L stainless steel surpassed that o f Ti6A17Nb and (2) the
cobalt alloy behaved slightly better than Fe22Crl 0NiN.
Discussion
and Conclusion
The primary aim o f this study was to damage mechanically the passive oxide layer
o f four materials by sliding wear and to measure the repassivation current transients
220
generated during the redevelopment of the passive film. A secondary goal was to
investigate the effects of scratch time, scratch speed, and pH on the current density
generated during sliding wear.
Initially, potentiodynamic testing and material characterization were performed on
all implant metals in order to obtain a reference set of corrosion data, which could be
compared with data collected during sliding wear. From the results outlined in this report,
one of the most important results to note is that the titanium alloy (Ti6A17Nb), on
average, corrodes at a significantly higher rate during wear than all other materials tested
over the potential range o f - 1 0 0 m V to +100 mV, despite its excellent corrosion resistance
in the absence of wear. Secondly, pH does not greatly affect the wear-induced corrosion
behaviour of Co28Cr6Mo or Ti6A17Nb, but does significantly affect that of 316L
stainless steel and, to a lesser extent, that of Fe22Crl 0NiN.
The current generated during repassivation of any metal is due to at least three
factors: the disruption and rebuilding of the double layer, the passive oxide film
regeneration, and the ionic dissolution of soluble ionic species [ 12]. However, Gilbert et
al. proposed that approximately 80% of the total current generated is due solely to the
oxide film regeneration [12]. Therefore, it can be assumed that the current densities
measured in this study are largely associated with the regeneration of the oxide film.
Gilbert et al. [12] reported values of Ipeakof 13 to 27 A / c m 2 for titanium using a similar
testing method to the one employed in our series or testing, except with high speed
scratching (90~tm scratches in 0.5 to 1.0 ms). It should be noted that, in our study, the
entire scratched surface (7.5mm x 1.5mm) was used to calculate the current densities; this
is an incorrect assumption, for the actual activated area at any given time is considerably
smaller. This explains the discrepancy between the two sets of results; Gilbert et al. used
an extremely small area to calculate the current densities, which more closely
approximate the actual activated surface area at any given time.
It is proposed that the poor performance of Ti6AI7Nb during wear is due to two
factors: (1) The freshly formed titanium oxide layer is relatively softer than those of the
other materials tested and is more easily destroyed by mechanical wear; and (2) higher
current densities are generated during wear because of the four valence electrons needed
to reform the titanium oxide passive layer (mostly made up of TiO2). Titanium is known
for its high corrosion resistance, due to relatively fast regeneration of the passive layer
and its subsequent thickening over time. However, if the passive oxide is fractured or
abraded at a frequency that is greater than its ability to repassivate fully, severe corrosion
of titanium is a definite possibility. Thus, the results presented in this paper indicate that
Ti6A17Nb has the potential to corrode severely if placed under conditions of continual
wear, particularly in the case where oxygen transport is inhibited.
One final point to note is that, at any given current density, the rate of titanium
material loss is less than that of other materials studied, due to its low density and four
valence electrons count. Therefore, although the current densities recorded for titanium
during wear are significantly higher than those found for other materials, the impact on
the corresponding wear rate is not as great as the measured current densities would
suggest. However, the difference in current densities measured is large enough that the
material loss of Ti6A17Nb is still greater than both Fe22Crl 0NiN and Co28Cr6Mo at any
pH.
221
References
[1]
Willert, H.-G., Brob/ick, L.-G., Buchhorn, G.H., Jensen, P.H., K6ster, G., Lang, I.,
Ochsner, P., and Schenk, R., "Crevice Corrosion of Cemented Titanium Alloy
Stems in Total Hip Replacements," Clinical Orthopaedics and Related Research,
No. 333, 1996, pp. 51-75.
[2]
Gilbert, J.L., Buckley, C.A., and Jacobs J.J., "In Vivo Corrosion of Modular Hip
Prosthesis Components in Mixed and Similar Metal Combinations. The Effect of
Crevice, Stress, Motion, and Alloy Coupling," J. Biomedical Materials Research,
Vol. 27, 1993, pp.1533-1544.
[3]
Brown, S.A., Flemming, C.A.C., Kawalec, J.S., Placko, H.E., Vassaux, C.,
Merritt, K., Payer, J.H., and Kraay, M.J., "Fretting Corrosion Accelerates Crevice
Corrosion of Modular Hip Tapers," J. Applied Biomaterials, Vol. 6, 1995, pp. 1926.
[4]
McKellop, H.A., Sarmiento, A., Brien, W., Park, S.H., "Interface Corrosion of a
Modular Head Total Hip Prosthesis," J. Arthroplasty, Vol. 7, 1992, pp. 291-294.
[5]
Verdonschot, N., and Huiskes, R., "Mechanical Effects of Stem Cement Interface
Characteristics in Total Hip Replacement," Clinical Orthopaedics and Related
Research, No. 329, 1996, pp. 326-336.
[6]
Wimmer, M.A., Bluhm, A., Kunze, J., Heckel, L., Morlock, M.M., and Schneider,
E., "Fretting Wear of Titanium and Cobalt-Chromium Alloys against BoneCement," Proceedings of 44thAnnual Meeting of the Orthopaedic Research
Society, New Orleans, March 1998, p. 353.
[7]
Rabbe, L.M., Rieu, J., Lopez, A., and Combrade, P., "Fretting Deterioration of
Orthopaedic Implant Materials: Search for solutions," Clinical Materials, 15,
1994, pp. 221-226.
[8]
Hirakawa, K., Bauer, T.W., Stulberg, B.N., Wilde, A.H., and Secic, M,
"Characterisation and Comparison of Wear Debris from Failed Total Hip Implants
of Different Types," J. Bone andJointSurgery, Vol. 78A, 1996, pp. 1235-1243.
[9]
Rieker, C., Shen, M., and KSttig, P., "In-Vivo Tribological Performance of 177
Metal-on-Metal Hip Articulations," Worm Tribology Forum in Arthroplasty, C.
Rieker, S. Oberholzer and U. Wyss, Eds., Huber Verlag, Bern, 2001, pp. 137-142.
[10] Brown, S.A., and Merritt, K., "Fretting Corrosion of Plates and Screws: An In
Vitro Test Method," Corrosion and Degradation of Implant Materials, ASTM
STP 859, A.C. Fraker and C.D. Griffin, Eds., ASTM International, West
Conshohoeken, PA, 1985, pp. 117-135.
[11] Steinemann, S.G., "Corrosion of Implant Alloys," TechnicalPrinciples, Design
andSafety ofJointlmplants, G.H. Buchhorn and H.G. Willert, Eds., Hogrefe &
Huber Publishers, 1994, pp. 168-179.
[ 12] Gilbert, J.L, Buckley, C.A., and Lautenschlager, E.P., "Titanium Oxide Film
Fracture and Repassivation: The Effect of Potential, pH and Aeration," Medical
Applications of Titanium and Its Alloys: The Material and Biological Issues,
ASTM STP 1272, S.A. Brown and J.E. Lemons, Eds., Philadelphia, 1996, pp.
199-215.
Dieter Wirz 1, Bruno Zurfluh 2, Beat G6pfert3, Feng Li4, Willi Fricks and
Erwin W. Morscher6
Results o f in Vitro Studies about the M e c h a n i s m of W e a r in the Stem-Cement Interface of T H R
Reference: Wirz, D., Zurfluh, B., G6pfert, B., Li, F., Frick, W. and Morscher, E. W.,
" R e s u l t s of in Vitro Studies about the M e c h a n i s m of W e a r in the Stem-Cement Interface of THR," Stainless Steels for Medical and Surgical Applications, ASTM STP
1438, G. L. Winters and M. J. Nutt, Eds., ASTM International, West Conshohocken, PA,
2003.
Abstract: Experiments were performed in a weal- machine especially constructed to
simulate the natural in vivo process of wear in a de-bonded metal-cement interface. Four
different bone cements were fretted against S-30 stainless steel with mat surface finish.
The validation of the metal surfaces of the in vitro experiments revealed excellent correspondence to examined retrieved hip stems. No scratches or fretting orientation was
noted on the polished surfaces. ZrO2 particles did not play a decisive role in the wear
mechanism. PMMA as a softer material than S-30 steel did not mechanically abrade the
metallic surface. Further no steel particles could be found within the wear debris. Metal is
removed by dissolution in a process of fretting and crevice corrosion. Fretting corrosion
causes the formation of debris in the form of a mixture of particles of metallic oxides (but
not metallic particles!), polymer and serum.
K e y w o r d s : Arthroplasty; Biomechanics; Bone Resorption; Corrosion, Crevice; Corrosion, Fretting; Foreign-Body Reaction; Hip Prosthesis; Macrophages; Metals; Microscopy, Electron; Middle Age; Particle Size; Polymethylmethacrylate; Steel, Stainless
I MD, Laboratory of Orthopaedic Biomechanics, University of Basel, Burgfelderstr. 101,

CH-4012 Basel, Switzerland.
2 PhD, Laboratory of Orthopaedic Biomechanics, University of Basel, Burgfelderstr. 101,
3 g s Eng., Laboratory of Orthopaedic Biomechanics, University of Basel, Burgfelderstr.
101, CH-4012 Basel, Switzerland.
4 MS Eng., Laboratory of Orthopaedic Biomechanics, University of Basel, Burgfelderstr.
101, CH-4012 Basel, Switzerland.
5 PhD, Laboratory of Orthopaedic Biomechanics, University of Basel, Burgfelderstr. 101,
6 Professor era., MD, Laboratory of Orthopaedic Biomechanics, University of Basel,
Burgfelderstr. 101, CH-4012 Basel, Switzerland.
222
www.astm.org
WlRZ ET AL. ON STEM-CEMENT INTERFACE OF THR
223
Introduction
The main problem of total hip replacement (THR) remains aseptic implant
loosening due to osteolysis. The main mechanism of osteolysis appears to be enzymatic
induced resorption of wear particles by macrophages. Wear particles are generated first in
the articulation itself, especially through wear of polyethylene. However, wear particles
can be generated at any interface o f a THR, such as the implant-bone interface in non-cemented endoprostheses or the stem-cement interface in cemented ones. Wherever particles are generated they may be distributed in the "effective joint space" [ 1], thus leading
to osteolysis.
Various clinical observations showed the importance of the stem-cement interface
as a place where osteolysis has its source.
The tapered Exeter stem, an Fe-alloy (FeCrNiMnMo) device was introduced into
clinical practice in 1970 with a smooth surface. Between 1976 and 1980 this stem was
provided with a mat surface. As a consequence of the change of the surface an increased
rate of aseptic loosening of this stem was observed. Since 1986, the Exeter has a smooth
polished surface [2,3] .The superiority of the polished version of the Exeter stem over the
mat one has been confirmed by Howie et al.[4] and the Swedish Implant Register [5].
The Mfiller Straight Stem, introduced in clinical practice in 1977, originally was
made of Protasul 10 (CoCr alloy) and stainless steel (AISI 316-L). From 1985 until 1994,
it was manufactured of a titanium alloys including Ti6A14V and Ti6AI7Nb. In the early
nineties, several Swiss orthopedic institutions reported a high rate of early development
of osteolysis with the cemented Mfiller Straight Stem manufactured from titanium alloy
with mat surface, which led to early revisions [6]. The early development of osteolysis
was unexpected, since several cemented titanium femoral stems were - and still are - on
the market with excellent mid-term results: C6raver [7], Perfecta [8], Bicontact [9] etc.
The poor results of the cemented Mfiller Titanium Straight Stem, however, could not be
confirmed at our institution [10]. There are currently contrasting clinical results with the
same type of cemented stem. While there was a high rate of aseptic loosening with
revision rates of 12 % and 8 % in clinics A and B, the clinics C (our own institution) and
D had revision rates of only 1.6 % and 1.5 %, respectively. These different results were
achieved with exactly the same endoprosthesis and after about the same time interval of
6 - 8 years! In light of these results, it is obvious that when a significantly different
number of osteolyses and revisions is observed in different institutions using the same
endoprosthesis - identical in design, surface, material, - some other variables must be
considered as likely causes for the differences in osteolysis and revision rates. These variables, i.e., the reason for the above-mentioned, different results, may stem from a difference in bone cement, cementing technique, operative technique and/or surgeon. It is to be
noted that all four clinics used modern cementing techniques, but clinics A and B, the
ones with the poor results, used a different cement than clinics C and D. The logical
conclusion, therefore, would be that bone cement does not equal bone cement! This
statement was also convincingly made by the Swedish and the Norwegian Implant Registers [5,11-13]. From the Norwegian Implant Register of the year 2000, it became obvious
that the bone cements Palacos and Simplex are superior to Sulfix 6 or Boneloc. The latter
two have been withdrawn from the market in the meantime. The risk ratio with 95 %
confidence interval are shown in the Swedish National Hip Register of the year 2000 [5]
224
for all revisions, in all diagnoses, and aseptic loosening in osteoarthritis. The risk ratio,
Sulfix 6 being the nominator (equals one) is 0.49 for Palacos- Gentamicin, 0.51 for
Palacos, 0.60 for Simplex and 0.73 for CMW. An equally good outcome for the Palacos
cement is given by the Norwegian Implant Register [ 11-13]. This register also indicated
that bone cement has a greater influence on the outcome of a THR than the prosthesis
itsel~
From these clinical observations, we conclude that a verdict on a total hip replacement component should never be made in an uni-dimensional way, in other words,
not without assessment of the whole system. In any case, the design, the surface characteristics, the material, the quality, and the operative the cementing technique, must be
considered.
The significant difference in clinical outcomes of THRs with different bone cements was also unexpected, because the results of the preclinical test methods of bone
cements did not reveal such important differences [14]. The question, therefore, arises:
Are current testing methods adequate to assess bone cements? Mechanical testing of different brands of bone cement in compression, tension or bending alone did not reveal any
significant differences. There is no evidence that bone cement with higher compressive
and tensile strengths improves the clinical longevity of the cement mantle. In contrast to
static test methods, dynamic fatigue testing has proven to be a more relevant test method
in the laboratory. Significant differences with regard to the quality of bone cements in
terms of survival rates, in accordance with the results of the Swedish and the Norwegian
Implant Register, were found by Harper and Bonfield [ 15] with a dynamic fatigue testing
method. On the other hand, Hopf and Fritsch/Germany [ 16,17] demonstrated
convincingly that the application of the W6hler method [ 17] for testing the fatigue
strength of PMMA is inadequate and misleading as polymers - such as PMMA - behave
differently than metals. With the W6hler method to measure fatigue strength of metals, a
sample has to undergo a certain number of cycles at a defined bending moment without
breaking. According to W6hter, 3 - 5 million cycles are appropriate for the testing of
metals. Hopf and Fritsch performed their fatigue tests up to a minimum of 20 million
loading cycles if breakage of the polymer specimens did not occur [ 17]. They found a
correspondence between the maximum number of loading cycles and the clinical results
of bone cements. These tests convincingly show that failures of the cement mantle, i.e.,
breakage, may even occur a long time after implantation in contrary to failures of the
(metallic) hip stem.
Wear particles - as mentioned earlier - are thought to be the cause of osteolysis
and subsequent aseptic loosening of implants [ 18]. However, very little is known about
the interaction of different bone cements with different prosthetic surface finishes when
fretted, simulating a debonded stem/cement interface. Nothing is known about possible
differences of bone cements under fretting. Different combinations of bone cement types
and surfaces might hypothetically lead to different quantities of wear particles (bone cement and metal), thus explaining the variant clinical outcome with one type of prosthesis.
In order to learn more about the basic mechanisms of the generation of wear debris originating from the cement-stem interface, it was decided to perform wear studies in
collaboration with the Research Department of Sulzer Orthopedics Ltd.
It was hypothesized that the amount of wear particles, their geometry and size after debonding in the cement-stem interface depends mainly on three factors:
WIRZ ET AL. ON STEM-CEMENT INTERFACE OF THR
1.
ZrO2 particles as x-ray additives in PMMA break off and then act as third
body particles under wear conditions
2.
Bone cements with proven good clinical outcomes (for example Palacos)
"generate" less wear debris(particles)
3.
The rougher the surface of the stem, the more cement is abraded and the
more particles are generated.
The wear studies were focused first - as mentioned above - on the basic mechanism of the development of wear particles at the metal/cement interface. Secondly, it was
attempted to examine why different cements behave differently under wear conditions.
Thirdly, the role of the radio-opaque additives, i.e. ZrO2 was assessed, and, finally, it was
aimed at getting more insight into the impact of the surface finish of the implants. A
basic hypothesis was that, the more wear debris are generated, i.e. abraded - both, from
the metallic the cement surface - the more particles are produced.
In order to elucidate experimentally the wear mechanism at the metal/cement interface, the laboratory set-up has to be able to replicate the in vivo mechanisms of wear
generation. The respective surfaces and the particles generated must then be studied and
compared to the findings from retrieved loosened stems and bone cements fragments.
Material and Methods
In Vitro Simulation of the Wear Mechanism

For this purpose, a wear machine had to be designed and constructed. Simulating
the natural fretting mechanisms at the stem/cement interface must take into account the
materials fretting against each other, the conditions of the fretting surfaces, the pressure
between the two surfaces facing each other, the amplitude of the two surfaces' movement
and, last but not least, the direction of the (respective) movements. Movement can only
occur when the two surfaces are physically separated from each other, i,e., when they are
debonded. Once debonded, the direction of the wear motion depends on the direction of
the forces and the geometry of the implants surface, i.e., the stress distribution from the
stem to the cement mantle and vice versa. These forces (which are eventually also
responsible for the loosening of the endoprosthesis stem) become effective in three
dimensions during gait: in the horizontal plane as rotational forces, "moving" the stem
into retroversion (1) and in the axial direction, as demonstrated by subsidence of the
femoral stem (2). On the other hand, axial loading increases the stability of collarless
tapered stems in the cement mantle. Finally, in the frontal plane, the forces are
transmitted to the calcar and the stem is "moved" into varus (3). The stress distribution in
a total hip replacement, therefore, is a complex combination o f forces acting in all three
dimensions. From the burnishing pattern on retrieved hip stems it is known that the
interface micromotion is not uni-dimensional but rather consists o f translational and
rotational components. A potential migration and movement of the metal/cement surfaces
against each other occurs in two dimensions, the rotation of the stem within the cement
mantle being the most important one [ 19-2 l ]. Simulating the "natural" loosening process
in vitro, movements of the two surfaces fretting against each other must be multi-
225
226
dimensional, respecting the alternating direction of forces between the prosthesis stem
and the cement/bone complex.
The fretting machine, therefore, has to fulfill the following requirements: 1. twodimensional movements, 2. simulation of subsidence by the superposition of a slow
linear component to the cyclic loading, 3. variable amplitudes between 20 and 200 ~tm, 4.
normal pressures up to 4 MPa, 5, fretting frequencies of about 1 Hz (the cyclical
frequency is comparable to the average walking frequency of the human body), 6. a
physiological environment using lubrication with calf serum, 7. body temperature of
37~
The fretting samples have a contact area of 10 x 10 mm. Frequency, temperature,
amplitudes and pressure have to be measured and acquired by a computer over all the
experiment.
Tests Performed
1.
Evaluation of the in vivo process of wear in the cement-stem interface: Surfaces
of retrieved MS-30 stems with mat surface were compared with the surfaces of test specimens after fretting, both manufactured both manufactured of S-30 steel, a high nitrogen
stainless steel (FeCrNiMnMo, ASTM F1586, ISO 5832-9). In addition, one sample was
fretted with 1 Hz, over 1 million cycles in one direction (uni-dimensional). This test was
performed in order to show the differences in polishing by uni- and multi-directional.
2.
The impact of Zr02 particles on the fretting mechanism: PMMA samples either
with or without ZrO2 were fretted against S-30 stainless steel samples with a mat surface
(Table 1).
Table l - Fretting tests with PMMA alone (no Zr02) and Duracem 3 with Zr02 and
Gentamicin (currently not on the market)against S-30 steel
number of
total
Tests
number
Components
Parameters
(2 samples
of Sameach)
ples
3.
4.
Pure PMMA without

ZrO2 / S-30 mat
100 ~m, 1Hz, 1Million

cycl., bi-dimensional
Duracem 3 with ZrO2

and Gentamicin / S-30
mat
100 ~m, 1Hz, 1Million

cycl., bi-dimensional
Differences of the outcome of wear experiments with different bone cements

against S-30 mat surface: Three different bone cements were fretted against a mat
S-30 surface (Table 2).
The impact of longer fretting: 1 sample of S-30 with mat surface was fretted
against Duracem 3 with Gentamicin and ZrO2 for 5 million cycles.
227
Table 2 - Fretting tests." 3 different bone cements against S-30 stainless steel
number of
total numTests
Components
Parameters of movement (2 samples
ber of
each)
Samples
Duracem 3 with
Gentamicin and ZrOz /
S-30 mat
Palacos R with
Gentamicin and ZrOz /
S-30 mat
CMW 2000 and BaSO4 /
S-30 mat
100 ~tm, 1Hz, 1Million

cycl.,
bi-dimensional
100 gm, 1Hz, 1Million
cycl.,
bi-dimensional
100 gm, 1Hz, 1Million
cycl.,
bi-dimensional
5.
The impact of high speed fretting at 5 Hz, over 0.5 million cycles, in order to account for the impact of heat to wear.
All tests where performed with sinusoid movements with a path of 0. l ram, either uni-directional or bi-directional with slightly different frequency, all mat specimen have a Ra
value of 1.4 ~tm.
Evaluation of the Specimens

1.
Photo documentation of the surfaces before and after fretting.
2.
Scanning Electron Microscopy (Philips XL30 ESEM and Philips XL20, Philips
Inc, Eindhoven NL) of the surfaces before and after fretting.
3.
Laserprofilometry with UBM Laser Surface Profilometer (UBM Corporation,
Sunnyvale CA). The shape, the roughness and the reflecting power were examined before
and after fretting. Mean, standard-deviation and graphs of the stainless steel sample's
surface were performed.
4.
Measurement of weight loss because of fretting. The amount of debris worn off as
determined by measuring the loss of weight of the samples using a Mettler AT 261 Delta
Range scale (Mettler-Toledo Inc., Greifensee, Switzerland) with a precision of+ 0.01mg.
Results:
The photographs and the SEM pictures of fretted, mat S-30 stainless steel surfaces
correspond well with the appearance of in vivo retrieved stems with the same original
surface. They both revealed no distinct direction of scratches, indicating that the
polishing of a mat metal surface is caused rather by multi- than by uni-directional
movements (Figure 1).
228
F i g u r e 1 - Comparison of the surface of a mat S-30 sample after 5 million wear

cycles in laboratory simulation (a) and a femoral MS-30 stem retrieved for aseptic loosening 50 months postop. (b). During wear against bone cement, the S-30
surface lost its roughness resulting in a surface with polished areas and isolated
'valleys '.
Surface of PMMA-bone cement samples before (left) and afterfretting

(right). a: Zr02 particles totally embedded in PMMA, b: in Back Scattering
Electron technique (BSE) shining through, c: Surface o f PMMA bone cement
sample after fretting o f I million cycles against sand blasted surface of S-30 steel
in SEM technique and d: in BSE technique. 1: Clusters of ZrO2 particles, in (a, b)
wholly embedded in PMMA visible on the surface and partially abraded in (c, d).
2: Debris consists in a mixture of PMMA debris, 3: ZrO2 particles and bovine
serum building a sludge covering the surface partially.
Figure 2 -
229
At the end oft_he fretting tests, very small particles (< 0.5 gin) which could be
identified as Zr-oxide were detected within the sludge on the abraided cement surface
(Figure 2).
ZrO2 in the pre-polymerized stage is found in clusters of 5 to 30 gin. Apparently,
the larger Zr-oxide conglomerates are broken up into very small particles.
Laser profilometry showed non significant differences only in the polishing effect
of bone cement with or without the X-ray additive ZrO2. The Ra-values of the S-30 samples fretted against pure PMMA without ZrO2 are reduced by 7.7 12%, the ones fretted
against Duracem 3 with Gentamicin (with ZrO2) by 15 21% compared to the original
surface before fretting.
Figure 3 shows the number of measured points with a reflection of more than 60%
on the surface of the matte S-30 steel samples. Surprisingly - and in Contrast to our
hypothesis that a "good cement" reveals less wear - Palacos R (clinically proven as bone
cement with the best perforrnance[5,11-13]) produced even more reflection and, therefore, revealed a higher polishing effect on the mat stem surface than other bone cements
as determined by laser profilometry. This finding is significant (Figure 3). The percentage
of light reflection can be taken as a measure of polish. Ra measurements revealed the
highest roughness reduction by Palacos too.
Figure 3 - Number of measured Points with a reflection of more than 60% on the
surface of the matte S-30 steel samples after fretting against four different bone
cements (I million cycles', 4 MPa pressure, two-dimensional fretting I OOlam).No
significant difference can be found for Duracern 3 (with Gentarnicin and Zr02)
and pure PMMA (without Zr02). Zr02 has no special polishing effect to a S-30
steel surface.
Note the significantly higher polishing of S-30 steel afier fretting against Palacos
R (with Gentamicin and Zr02)!
The weight loss of bone cement when fretting CMW 2000 (-19.78 -4- 12.59 rag)
against mat S-30 (-0.40 0.24 rag) is about 50 times larger than the metal loss. In the
case of pure PMMA (-5.02 2.41 mg), Duracem 3 with ZrO2 and Gentamicin (-0.56
2.02 rag) and Palacos R with Gentamicin (-4.87 4.15 rag), this ratio is much smaller.
230
STAINLESS STEELS FOR MEDICAL AND SURGICALAPPLICATIONS
Palacos apparently removed less S-30-material from the mat S-30 surface than
pure PMMA or Duracem 3, and the same amount as CMW 2000. The difference,
however, is not significant. (Figure 4)
Furthermore, the only fretting particles detected in the sludge were identified as
ZrO2 particles. No steel particles from the S-30 alloy could be found! The ZrO2 particles,
however, were smaller than 0.2 gm, and, therefore much smaller than the original particles seen in the PMMA before fretting. In other words Zr-Oxide particles were mechanically broken up by the fretting process.
Figure 4 - Weight loss of the mat S-30 steel samples after fretting against four
different bone cements (1 million cycles, 4 MPa pressure, two-dimensional

fretting 100l~m).
Discussion
The presented experiments are the first published data about the mechanism of
wear in the metal-cement interface of cemented total hip replacements based on in vitro
with multi-directional, i.e. relevant movements. The experiments were performed in a
wear machine designed to replicate the natural in vivo process of wear at a debonded
metal-cement interface. The validation of the metal surfaces of the in vitro experiments
revealed excellent correspondence with examined retrieved hip stems. No scratches or
fretting orientation on the polished surfaces does occur, in contrast to uni-directional
wear experiments with larger fretting paths done by others [22-26].
Fretting a S-30 mat surface against bone cement results: 1. In a loss of weight and
decrease of roughness through "removal" of the tops of the asperities. 2. In a more polished and more reflecting surface. This increase of reflection is not due to "valleys" filled
with PMMA debris but due to an increase amount of polished surface by the fretting
process and polishing the "peaks" of the rough surface. 3. In a progressive filling of the
"valleys" between the asperities with PMMA debris as the wear process is going on. The
abraded metal cannot be identified as particles within the wear debris, apparently it is
dissolved.
The respective measurement of the Ra-value of the S-30 stainless steel surface
with Laserprofilometry revealed only a slightly higher reduction of the roughness after
fretting against PMMA with ZrO2 than against PMMA without ZrOz. Thus it must be
concluded that the X-ray additive ZrO2 does not play a significant role in the process of
231
wear generation and for the development of osteolyses. This is in contrast to the tests
perfomaed by Cooper et al. showing an increased abrasion due to ZrO2 [27]. In pin on
disk fretting experiments, neither fretting- nor crevice- corrosion takes place. Further,
ZrOz conglomerates of a size of 5 to 30 ~tm are pulverized into very small particles with a
size of less than 0.2~tm and particles of this size are supposed to be too small to induce
enzymatic bone resorption [28].
Palacos - a clinically proven, good bone cement - has a significantlyhigher polishing effect. A more extensively polished surface is supposed to generate less wear
debris by abrasion. On the other hand, Palacos produced the smallest loss of metal weight
with fretting. This, however, was not significant.
By wearing S-30 stainless steel against pure PMMA the rough surface of the
S-30 stainless steel is polished by removing material though pure PMMA contains
no hard materials like ZrO2 that might abrade and polish the surface mechanically. Further, no S-30 particles could be found neither within the wear debris nor in the bovine
serum. A different process than mechanical abraison of metal particles alone thus must be
brought into consideration. This might be described as chemical dissolution of oxide
layers in the process of fretting corrosion, or, under very specific conditions, even crevice
corrosion in a strongly acidic, oxygen depleted medium [29].
Fretting corrosion causes the formation of debris in the form of a mixture of polymer, serum and metallic oxides (but not metallic particles!). Serum is rich in chlorides,
promoting the process of corrosion. [29]. The metal dissolves, forms complexes with
peptides or proteins or forms insoluble salt deposits as is the case on titanium surfaces
[30]. Furthermore, oxygen depletion in confined pockets of serum enhances crevice
corrosion.
According to these observations, we conclude that the differences in the quahty of
bone cements consist first of all in their capability to induce a process of fretting corrosion, which also greatly depends on the surface characteristics of the metal (roughness,
nature of the oxide surface, material etc) [2].
CoCr alloys show a lower de-passivation pH-value, and thus show less corrosion
in crevice conditions than alloys of austenitic stainless steel as e.g. S-30. It, therefore
must be assumed that endoprosthetic femoral stems manufactured from CoCr-alloys
reveal different fretting behaviour in the cement-stem interface than steel prostheses. On
the other hand, clinical experience with polished steel prostheses has shown excellent
long term results with extremely low incidence of osteolyses.
A complete cement mantle covering the tip of the stem, modem cementing technique used together with a clinically proven good bone cement avoid debonding of hip
prosthesis stem with a mat surface. Such implants obtain good results, at least in the midterm [3 t ]. Nevertheless, for tapered collarless femoral stems, the use of polished surfaces
is recommended and the use of a mat surface should definitively be abandoned as subsidence within the cement mantle is no more considered with aseptic loosening of a prosthesis. If the stem is implanted in combination with an "open" centralizer, i.e. a
centralizer designed in order to allow subsidence, the use of a mat surface has to be
avoided.
Further wear studies fretting polished steel surfaces and CoCr alloys will have to
show whether there is a potential for further improvement of results of cemented THR by
improving implant surfaces and improving bone cements. For a specific endoprosthesis
232
system the design, the implants surface, the materials and the operative, i.e., cementing
technique must be considered as an entity. Most failures in endoprosthetics are the result
of"improvements" based on conclusions from a mono-causal approach!
Progress in THR will continue, though we find ourselves already in the upper area
of the asymptotic curve of the "margin revenue". Despite that, we can get even better results of THRs with better implant designs, better surfaces, better materials, including better bone cements, better operative, i.e., cementing techniques, and corrosion resistant
metals. However, we must not forget that the surgeon still is the greatest variable.
Industry should keep in mind that better clinical results at lower costs are not
achieved with ever new implant designs, but with continuous improvements of existing
systems and, even more importantly, with measures to improve the operating surgeon's
knowledge and skills.
Acknowledgments
The authors would like to thank Dieter Stoll and Scott Wilson (Sulzer INNOTEC,
Winterthur, Switzerland) for performing the Laserprofilometry, Ralf Klabunde (Sulzer
Orthopedics, Winterthur, Switzerland) for supplying the samples and the scientific support, Richard Guggenheim, Marcel D/Jggelin (REM-Lab., University of Basel) and Udo
M. Spornitz (Institute of Anatomy, University of Basel) for their support in electron microscopy. This work was supported by the Hardy and Otto Frey-Ziind Foundation, Basel,
Switzerland and Sulzer Orthopedics, Winterthur, Switzerland.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
Schmalzried, T.P., Jasty, M. and Harris, W.H. "Periprosthetic Bone Loss in Total
Hip Arthroplasty. Polyethylene Wear Debris and the Concept of the Effective
Joint Space," Journal of Bone Joint Surgery Am. Vol. 74, No. 6, 1992, pp. 849863.
Ling, R.S. "Observations on the Fixation of Implants to the Bony Skeleton,"
Clinical Orthopedics and Related Research Vol. 210, 1986, pp. 80-96.
Fowler, J.L., Gie, G.A., Lee, A.J. and Ling, R.S. "Experience with the Exeter Total Hip Replacement since 1970," The Orthopedic clinics of North America. Vol.
19, No. 3, 1988, pp. 477-489.
Howie, D.W., Middleton, R.G. and Costi, K. "Loosening of Matt and Polished
Cemented Femoral Stems," Journal of Bone Joint Surgery Br. Vol. 80, No. 4,
1998, pp. 573-576.
Malchau, H., Sodermann, P. and Herberts, P. "Swedish Hip Registry: Results with
20-Year Follow Up with Validation Clinically and Radiographically," Orlando,
67th Annual Meeting of the American Academy of Orthopaedic Surgeons, 2000.
Maurer, T.B. and Ochsner, P. "Results and Guidelines after 78 Revisions of Cemented Titanium Stems," Beaune, European Hip Society meeting, 1998.
LeMouel, S., Allain, J. and Goutallier, D. "Analyse actuarielle ~ 10 ans d'une cohorte de 156 proth6ses totales de hanche ciment6es ~ couple de frottement alu-
[8]
[9]
[ 10]
[ 11]
[12]
[13]
[14]
[15]
[ 16]
[17]
[ 18]
[19]
[20]
[21 ]
233
mine/poly-ethylbne," Revue de chirurgie orthoprdique et reparatrice de rappareil

moteur. Vol. 8, 1998, p. 345.
Van der Straeten, C., Goossens, M. and Gruen, T. "Cemented and Non-Cemented
Hip Arthroplasty with a Proximal Plasma Sprayed Titanium Stem," Journal of
Bone Joint Surgery Br. Vol. 81, Supplementum II, 1999, p. 173
Eingartner, C., Volkmann, R., Kummel, K. and Weller, S. "Niedrige Lockerungsrate einer zementierten Titan-Geradschaftprothese im langerfristigen Verlauf,"
Swiss Surgery Vol. 3, No. 2, 1997, pp. 49-54.
Acklin, Y.P., Berli, B.J., Frick, W., Elke, R. and Morscher, E.W. "Nine-Year Results of Muller Cemented Titanium Straight Stems in Total Hip Replacement,"
Archives of Orthopaedic and Trauma Surgery. Vol. 121, No. 7, 2001, pp. 391398.
Havelin, L.I., Espehaug, B., Vollset, S.E. and Engesaeter, L.B. "Early Aseptic
Loosening of Uncemented Femoral Components in Primary Total Hip Replacement. A Review Based on the Norwegian Arthroplasty Register," Journal of Bone
Joint Surgery Br. Vol. 77, No. 1, 1995, pp. 11-17.
Havelin, L.I., Espehaug, B., Vollset, S.E. and Engesaeter, L.B. "The Effect of the
Type of Cement on Early Revision of Charnley Total Hip Prostheses. A Review
of Eight Thousand Five Hundred and Seventy-Nine Primary Arthroplasties from
the Norwegian Arthroplasty Register," Journal of Bone Joint Surgery Am. Vol.
77, No. 10, 1995, pp. 1543-1550.
Havelin, L.I., Engesaeter, L.B., Espehaug, B., Fumes, O., Lie, S.A. and Vollset,
S.E. "The Norwegian Arthroplasty Register: 11 years and 73,000 Arthroplasties,"
Acta orthopaedica Scandinavica. Vol. 71, No. 4, 2000, pp. 337-353.
Kfihn, K.D. "Up-to-Date Comparisons of Physical and Chemical Properties of
Commercial Materials," Berlin, Heidelberg, New York: Springer, 2000.
Harper, E.J. and Bonfield, W. "Tensile Characteristics o f t e n Commercial Acrylic
Bone Cements," Journal of biomedical materials research. Vol. 53, No. 5, 2000,
pp. 605-616.
Hopf, T., Zell, J. and Hanser, U. "Methods of Fatigue Testing of PMMA Bone
Cement," Medizinisch Orthop~idische Technik. Vol. 105, 1985, pp. 20-25.
Hopf, T. and Fritsch, E.W. "Fatigue Strength of Vacuum-Mixed Polymethyl
Methacrylate Bone Cement," Orthopedics International Vol. 5, 1997, pp. 98-104.
Wilson-MacDonald, J., Morscher, E. and Masar, Z. "Cementless Uncoated
Polyethylene Acetabular Components in Total Hip Replacement. Review of Fiveto 10-Year Results," Journal of Bone Joint Surgery Am. Vol. 72, No. 3, 1990, pp.
423-430.
Alfaro-Adrian, J., Gill, H.S. and Murray, D.W. "Cement Migration after THR. A
Comparison of Charnley Elite and Exeter Femoral Stems using RSA," Journal of
Bone Joint Surgery Am. Vol. 81, No. 1, 1999, pp. 130-134.
Alfaro-Adrian, J., Gill, H.S. and Murray, D.W. "Should Total Hip Arthroplasty
Femoral Components be Designed to Subside?: A Radiostereometric Analysis
Study of the Charnley Elite and Exeter stems," Journal of Arthroplasty. Vol. 16,
No. 5, 2001, pp. 598-606.
Bergmann, G. "In vivo Messung der Belastung von Hfiftimplantaten," Berlin,
Verlag Dr. K6ster, 1997.
234
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31 ]
Stoll, D. "Knochenzementversuche, Interner Bericht Sulzer Irmotec," 1998.

Charri&e, E. "Projet de diplrme en ing6nierie biomrdicale, EPFL Lausanne,"
1998.
Wimmer MA, Nassutt R, Lersner A, Kunze J, Schneider E. "Schwingverschleiss
von Titan gegen Knochen verursacht Abriebpartikel kleiner als Zellgrrsse", in:
Zuin Gohr, Reibung und Verschleiss, DGM Verlag, 1996.
Wimmer MA, Schneider E, "Fretting Wear of Titanium and Cobalt-Chromium
Alloys Against Bone Cement", Annual Meeting of the Orthopedic Research Society. 353, New Orleans, 16.-22.3.1998.
Wimmer MA, Bluhm A, Hansen I, Heckel L, Schneider E, "Fretting Wear of Titanium against Bone-Cement depends on the Surface Finish of Titanium", Annual
Meeting of the European Orthopedic Research Society, Amsterdam (NL), 1998.
Cooper, J. R, Dowson, D. Fisher, J. and Jobbins, B. "Ceramic Bearing Surfaces in
Total Artificial Joints: Resistance to Third Body Wear Damage from Bone
Cement Particles," Journal of Medical Engineering and Technology., Vol. 15, No.
2, 1991, pp. 63-67.
Ingham, E. and Fisher, J. "Biological Reactions to Wear Debris in Total Joint
Replacement," Proceedings of the Institution of Mechanical Engineering, Vol.
214,No. 1, 2000, pp. 21-37.
Tritschler, B., Forest, B. and Rieu, J. "Fretting corrosion of materials for
orthopaedic implants: a study of a metal/polymer contact in an artificial physiological medium," Tribology International, Vol. 32, 1999, pp. 587-596.
Willert, H.G., Broback, L.G., Buchhorn, G.H., Jensen, P.H., Koster, G., Lang, I.,
Ochsner, P. and Schenk, R. "Crevice corrosion of cemented titanium alloy stems
in total hip replacements," Clinical Orthopedics and Related Research. Vol. 333,
1996, pp. 51-75.
Berli, B., Elke, R. and Morscher, E.W. "MS-30 stems - matte versus polished surfaces; clinical and radiological results ofa 5 year follow-up," Annual Meeting of
the European Federation of Orthopaedic Research and Traumatology, Rhodos,
Greece 6.4. 2001.
Mengke Zhu I and Markus Windler 2
Fretting and Anodic Current of the Taper Interface of Stainless Steel Hip
Stems/Cobalt-Chromium Femoral Heads
Reference: Zhu, M. and Windier M., "Fretting and Anodie Current of the Taper
Interface of Stainless Steel Hip Stems/Cobalt-Chromium Femoral Heads," Stainless
Steels for Medical and Surgical Applications, ASTM STP 1438, G. L. Winters and M. J.
Nutt, Eds., ASTM International, West Conshohocken, PA, 2003.
Abstract: Three types of hip stems, MS-30 T M high-nitrogen stainless steel, Natural
HipT M cobalt-chromium, and APR | II porous titanium were interfaced with cobaltchromium femoral heads. They were tested according to ASTM F 1875-98. In the initial
1 Hz cyclic loading up to 480 cycles, the average fretting current generated from the taper
interface was 0.4, 0.6, and 1 pA, respectively for the stainless steel, cobalt-chromium,
and titanium stems. In the following 2 Hz cyclic loading up to 3600 cycles, the average
fretting current was 1.3, 0.6, and 2.25 pA, accordingly. In the potentiodynamic
polarization, the titanium stems showed relatively high anodic current in the range of
-100 to 500 mV. The cobalt-chromium stems had the highest anodic current in the range
of 500 to 1200 mV. The stainless steel stems exhibited the best corrosion resistance in
the whole potential range of polarization.
Keywords: stainless steel, cobalt-chromium, titanium, fretting corrosion, orthopedic hip
stems, femoral head, taper interface
Introduction
Modular orthopedic hip prostheses are widely used in total hip arthroplasty due to
their clinical flexibility. However, clinically retrieved modular hip prostheses have
manifested corrosion related problems particularly at the taper interfaces [ 1, 2]. The
taper interfaces of hip stems and femoral heads are subjected to micro-motion that might
lead to fretting corrosion. The modular connection of the hip stem and head of different
materials is susceptible to galvanic corrosion. Moreover, stainless steel hip implants have
~Senior Research Engineer, Research and Analysis Department, Sulzer Orthopedics Inc.,
9900 Spectrum Dr., Austin, TX 78717, USA.
2Manager of Materials, Research and Analysis Department, Sulzer Orthopedics Ltd.,
P. O. Box 65, CH-8404, Winterthur, Switzerland.
235
www.astm.org
236
been also used clinically, which requires special attention to their corrosion behavior as a
less frequently used type of material. In vivo corrosion of Charnley and Mtiller 316L
stainless steel hip implants was reported occurring around the one piece design stems [3].
However, little is reported on the corrosion behavior of modular stainless steel hip
implants. The purpose of this study was to evaluate the in vitro fretting corrosion
behavior of a high-nitrogen stainless steel stern/cobalt-chromium (COCO head taper
interface and compare it to that of CoCr stem/CoCr head and titanium stem/CoCr head
taper interfaces.
Materials
The following number of hip stems and femoral heads (Figure 1) were used in this
study.
6
MS-30 TM high-nitrogen stainless steel hip stems, size 12
(Protasul S-30, ASTM F 1586,
stainless steel stems)
5
Natural-Hip TM CoCr stems, size 4 (ASTM F 1537, CoCr stems)
5
APR | II Porous Ti-6A1-4V hip stems, size 15 mm
(ASTM F 136, titanium stems)
16
Sulzer CoCr femoral heads, size 28 mm, neutral (ASTM F 1537,
CoCr heads)
Figure 1- Parts tested

From left to right: MS-30 stainless steel, Natural-Hip CoCr, APR II Porous titanium
All parts had a 12/14 mm taper with threads on the stem cone.
Lactated Ringer's Injection USP, Baxter Healthcare Corp., Deerfield, IL 60015
ZHU AND WINDLER ON HIP STEMS/FEMORAL HEADS
237
Methods
This study was conducted per the ASTM Standard Practice for Fretting Corrosion
Testing o f Modular Implant Interfaces: Hip Femoral Head-Bore and Cone Taper
Interface (F1875-98). Fretting corrosion tests were conducted in two configurations
corresponding to Procedure B and Procedure A o f Test Method II of the standard. A total
o f five tests described in Test Method II were conducted; two tests were performed in
Configuration I and three tests were in Configuration II. Each o f the 16 stem and head
assemblies was set up and went through the five tests in the sequence described as
follows, with the most destructive test executed at the end.
Setup of the Hip Stem and the Femoral Head
Each set o f a hip stem and a femoral head was assembled on an Instron material test
machine in a dry condition. The stem was fixed with the neck aligned with the load axis.
The femoral head was pushed onto the stem cone by applying a static load to 2000 N.
The fretting corrosion tests were conducted on a servo-hydraulic MTS material test
machine. The stem and head assembly was set up in an inverted position as per ASTM
Cyclic Fatigue Testing o f Metallic Stemmed Hip Arthroplasty Femoral Components
without Torsion (F1440). The femoral head articulated against an acetabular insert that
was fixed in a fluid chamber mounted on the lower platen o f the MTS test machine. The
distal portion o f the stem was fixed on the upper platen o f the machine (Figures 2 and 3).
The stem and head were either unloaded or cyclically loaded according to the
requirements o f individual fretting corrosion tests. If required, cyclic loads of 2000 N in
a sinusoidal waveform, with a minimum o f 40 N and a maximum o f 2040 N, were
applied to the stem and head assembly at either 1 or 2 Hz.
Fretting Corrosion Testing
Configuration I Fretting Current at Open Circuit Potential - The stem and head
assembly was the working electrode (A in Figure 2), while an identical hip stem as the
test stem (B in Figure 2) was employed as the counter electrode. Both working and
counter electrodes were immersed in 500 mL lactated Ringer's solution in the fluid
chamber. The current generated at the taper interface was measured using a galvanic
corrosion program o f an EG&G 273A Potentiostat/Galvanostat, which served as a zero
resistance ammeter.
238
Figure 2 - Setup in Configuration I

The current was measured while the stem was cyclically loaded in the following two
loading periods.
Test 1 Cyclic loading at 1 Hz Current was measured for a total of 600 seconds (s),
including 120 s while the stem was not loaded and 480 s while it was cyclically loaded
(current increase due to fretting was recorded in this period).
Test 2 Cyclic loading at 2 Hz Current was measured for 120 s at 360, 1000, 1800,
and 3600 cycles periodically.
Configuration II Corrosion potential (Rest Potential) and Anodic Current at Applied
Potentials - In this test configuration, the working electrode was the stem and head
assembly (A in Figure 3). A platinum wire counter electrode (B in Figure 3) was placed
close to the CoCr head and the lugging probe of a saturated calomel reference electrode
(SCE, C in Figure 3) was positioned near the taper interface. To reduce the effect of size
and geometry differences between the three types of stems, the stem and head assembly
was immersed in 160 mL Ringer's solution such that only the CoCr head and the stem
neck portion was exposed to the solution. The following three tests were conducted
using the EG&G 273A Potentiostat/Galvanostat.
Test 3 Corrosion potential measurement Corrosion potential of the working electrode
was measured for one hour while the stem was cyclically loaded at 1 Hz. To display the
effect of fretting, measurement was first performed without applying the loads for 5
minutes.
Test 4 Potentiostatic polarization Potentiostatic polarization was conducted at 50
mV/5 minutes for five steps from -150 mV through 50 mV, The current was measured
for five minutes at each applied potential. Cyclic loads at 1 Hz were applied in this test.
Test 5 Potentiodvnamic polarization Potentiodynamic polarization was conducted at
0.667 mV/s from the then open circuit potential to 1200 mV. No cyclic loads were
applied in this test.
ZHU AND WINDLER ON HIP STEMS/FEMORALHEADS
239
Figure 3 - Setup in Configuration II

The taper interfaces of the stems were examined using an SEM before and after the
tests. The stem and head couples were disassembled by using a head extractor.
Results
Fretting Current at Open Circuit Potential (Configuration I)

Test 1 - Figure 4 shows the initial response of current to fretting of the modular
interface. Current jumped in response to the commencement of the cyclic loads at 120 s.
In the initial 8 minutes of cyclic loading at 1 Hz, several stems showed unstable current
variation, and the titanium stems scattered most in the current output. When the curves
of the stems in each type were averaged, the fretting current was approximately 0.4 ~tA,
0.6 ~tA, and 1.3 ~tA, respectively for the stainless steel, CoCr, and titanium stems
(Figure 5).
Test 2 - In the following periodical current measurements with cyclic loading at 2
Hz, no consistent variation was observed between the current of each stem and the cycle
number. In view o f that, all measurements (started at 360, 1000, 1800, and 3600 cycles)
of the stems in each group were averaged (Figure 6). The averaged fretting current
values corresponding to the three types of stems were approximately 1.3 ~tA, 0.6 ~tA, and
2.25 ~tA. In the cyclic loading at 2 Hz, the current of the individual stems appeared
relatively stable. In comparison to the values in the initial 1 Hz cyclic loading, the
averaged fretting current of the stainless steel and titanium stems rose by approximately
225% and 73%, respectively. In contrast, no apparent change was observed for the CoCr
stems.
240
S Steel Stems/CoCr Heads (n=3)

5
Titanium Stems/CoCr Heads (n=5)
~-
CoCr Stems/CoCr Heads {n=4)
?+"
2~
9
~ ~
~.."
~:-
1O0
Cto
...+
5~" - ,, - u, . " '. c c~_ , -
'
--
,", . ~ , - _. .,. -. ., ~ L
200
'
300
TimR/.ql
'
"
"2 ~
400
~++m++++
..... ~
'
500
600
F i g u r e 4 - Current in 1 Hz cyclic loading
Cyclic loads started at 120 s
F i g u r e 5 - A v e r a g e current in 1 Hz cyclic
loading Cyclic loads started at 120 s
ZHU AND WINDLER ON HIP STEMS/FEMORALHEADS
241
Figure 6 - Current in 2 Hz cyclic loading

Average of results measured at 360, 1000, 1800, and 3600 cycles
and the stems from the same group
Corrosion Ppotential (Rest Potential) and Anodic Current at Applied Potentials

(Configuration 11)
Test 3- Corrosion potential dropped when the modular interface was subjected to the
cyclic loads at 1 Hz at 300 s (Figure 7). The titanium stems showed the highest (noblest)
corrosion potential, while the stainless steel and CoCr stems appeared to be comparable.
Test 4 - In the potentiostatic measurement with the 1 Hz cyclic loading, current
increased with increased applied potential for all stems, as represented by an stainless
steel stem (Figure 8). At an applied potential o f - 100 mV, the three types of stems
showed similar levels of anodic current (Figure 9). When the applied potential reached
50 mV, titanium stems showed the highest anodic current, while the stainless steel stems
the lowest (Figure 10).
242
STAINLESS
STEELS FOR MEDICAL AND SURGICAL APPLICATIONS
S Steel Sterns/CoO" Heads (n=3)
0.(30
. .
- - 9 Titanium Sterns/CoO" Heads (rl=5)

" ,, .
-0.10
t'.~t::"
.
-
9 CoCr Sterns/CoCr Heads (n=4)
2,
. ' - -
" * - ' :.: -. . . .
. ..' _ -. : =.= . . .. . . .
. . . .
LU
(.9
09
-0.20
".
....
" ~ - ' , , . . . l ~
-0.30
.......
-0.40 " ~
-0.50
...~=
'
500
'
1000
'
1500
'
2000
"nine (s)
'
2500
'
'
3000
F i g u r e 7 - Corrosion potential measured in 1 Hz cyclic loading

Cyclic loads started at 300 s
3.60
3.40
3.20
3.00
@50mv
2.80
@Omv
2.60
@ -50 mv
2.40
"E
4-1 O0mv
@-150 mv
2.20
2.00
1.80
1.60
~ - ~
1,40
1.20
1.00
0.80
0.60
'
I
400
Time (S)
I
200
4OO0
I
300
F i g u r e 8 - Potentiostatic scan @50 m V/5 min, stainless steel stem
ZHU AND W I N D L E R ON HIP STEMS/FEMORAL HEADS
8.00 -1
',
7.00
~
~,
6.00
S Steel Stems/CoCr
Heads (n=4)
" " " Titanium Stems/CoCr

|
--
9 CoCr Stems/CoCr
Heads (n=4)
Heads (n=4)
|
5.00
--
4.00 - -
3.00
ZOO
-.--< : _ , . . . . , -
1.00
....
....':'_T'_
0,00
L
50
100
150
200
Time (S)
250
300
Potentiostatic measurement
@ - 100 mVvs. SCE
cyclic loads @ 1 Hz
Figure 9 -
1100
10,00 - -
- I
S Steel Stems/CoCr Heads (n=4)
" " " Titanium Sterns/CoCr Heads (n=5)

~
CoCr Sterns/CoCr Heads (n=5)
- -
9,00
8.00
=k
7.00
14
6.00
5.00
9
- . , . o
4.00
3.00
~
" " " "~
zoo
. . . . . . .
1.03
'
50
'
100
2.2. . . . . . . . . . .
"- Z Z ~
- - r
'
150
I
200
. g,..
'
ram, , , ~
'
250
T i m e (S)
Potentiostatic measurement
@ 50 mVvs. SCE
cyclic loads @ 1 Hz
Figure 10 -
t
30o
243
244
Test 5 - In the potentiodynamic measurement without cyclic loading (Figure 11),

starting at 550 mV, the CoCr stems showed much higher anodic current than the other
two types o f stems. The severity of corrosion o f the CoCr stems was indicated
1.2
9 9
1.0
....
".1t ,I
S Steel StemslCoCr Heads (n=5)

Titanium Stems/CoCr Heads (n=5)
CoOt Stems/CoCr Heads (n=4)
,7/" /i
~/"
//
/ j - - / / I ./
/i.//
0.8
0.6
0 ~ / : -~ "
~W
I.-
..
0.4
:...
0.2
0.0
,,,
-7
........
-6
........
-5
........
-4
..................
-3
. ....... ,
-2
........
-1
CURRENT(Log A)
Figure 11 - Potentiodynamic scan @ O.67 m V/S
by the size o f the corroded area and its extent o f discoloring on the CoCr head and neck
(Figure 12), and by the thick and dark electrolyte left after the test. The titanium stems
showed relatively higher anodic current in the potential range up to 500 mV. The
stainless steel stems exhibited the least anodic current in the whole polarization range.
Figure 12 - Corrosion damage after
the potentiodynamic polarization

From left to right: Natural Hip CoCr, MS-30, APR 1I Titanium
ZHU AND WINDLER ON HIP STEMS/FEMORAL HEADS
245
SEM Examination of the Stem Taper Sections

The threaded taper appeared similar on the three types of stems. Typical for all
stems, no detectable corrosion damage was found in the threaded taper sections covered
by the femoral heads. Corrosion damage due to the potentiodynamic polarization was
primarily present on the area close to the taper and neck junction (Figure 13) and the stem
neck section below the taper as shown in Figure 12.
(a) before corrosion test

(b) afier potentiodynamic polarization
Figure 13 SEM pictures of the threaded taper section (MS-30)
Discussion
Corrosion at the femoral stem/head taper interface may result from fretting due to
micro motion, crevice at the taper mismatch, and galvanic coupling of dissimilar
materials or a combination of these three. This study investigated the fretting current due
to cyclic loading and galvanic corrosion behavior of the tree types of modular interfaces.
Since all parts had the same taper design, a large variation in the crevice of the tapers can
be dismissed.
The effect of cyclic loading was demonstrated in both the initial fretting current and
the corrosion potential measurements. Fretting at the taper interface creates fresh surface
areas that contribute to an increased current. The variation of the fretting current may
indicate how stem cones of different materials fit in the CoCr head bore. The fretting
current varied greatly in the initial cyclic loading due probably to the irregular micro
motion at the taper interface in the "sit-in" phase. The titanium stems appeared to have
most initial micro motion, while the stainless steel stems seemed to have the best initial
match with the CoCr head. As the cyclic load frequency increased to 2 Hz, both the
stainless steel and titanium stems revealed to be sensitive to the increased frequency. The
CoCr stems appeared to have the best fit with the CoCr head by showing the very low
fretting current and insensitivity to the frequency change. Since fretting current results
from changes of the surface state, its magnitude should mainly depend upon the materials
of the components and has little to do with the galvanic effect of mixed materials. The
24 6
generally higher fretting current of the titanium group appears to be characteristic of

titanium that reacts and repassivates sensitively to surface changes.
The CoCr head was a universal component of the modular taper interface. However,
the degree of corrosion of the modular portion as a whole was greatly affected by the
stem cones and necks of different materials. The coupling of the titanium stems had the
noblest corrosion potential, resulting probably from the sintering process of the stem. As
demonstrated by both potentiostatic and potentiodynamic polarization measurements,
titanium stems showed the highest anodic current in the lower range of applied potential
(- 50 mV to 500 mV). Although made of the same material, the coupling of the CoCr
stems and heads exhibited the highest anodic current or corrosion in the upper scan range
of 550 mV to 1200 mV. Noticeably, the coupling of the stainless steel stems had the best
corrosion resistance in the whole range of polarization. These results demonstrate that
the anodic current of a hip stem/head assembly is less dependent on the galvanic coupling
of dissimilar materials than on the polarization behavior of a particular material. In fact,
the polarization behavior of the three groups shown in Figure 11 reflected that of the
CoCr head, which dominated the polarization process because of its higher anodic current
than components of the other two materials [4].
Although only short-term testing was performed, this in-vitro study covered several
conditions that a hip stem may encounter in-vivo, including cyclic loading and a wide
range of applied potential. A limitation of this study was that the proximal bodies of the
stems were partially immersed in the electrolyte in the fretting current measurement,
which had different surface areas and shapes. Whether the additional surface area would
affect the fretting current was not examined. Under such a circumstance of testing and
with the said design features, the stainless steel stems behaved well, if not superior to the
titanium and the CoCr stems.
References
[1] Collier, J. P., Surprenant V. A., Jensen, R., Mayor, M. B., and Surprenant, H.,
"Corrosion Between the Components of Modular Femoral Hip Prostheses," Journal
of Bone and Joint Surgery [Br], Vol. 74-B, No. 4, 1992, pp. 511-517.
[21 Gilbert, J. L., Buckley, C. A., and Jacobs, J. J., "in vivo Corrosion of Modular Hip
Prosthesis Components in Mixed and Similar Metal Combinations. The Effect of
Crevice, Stress, Motion, and Alloy Coupling," Journal of Biomedical Materials
Research, Vol. 27, 1993, pp. 1533-1544.
[3] Walczak, J., Shahgaldi, F. and Heatley, F., "In vivo Corrosion of 316L Stainless
Steel Hip Implants: Morphology and Elemental compositions of Corrosion
Products," Biomaterials, Vol. 19, 1998, pp. 229-237.
[4] Zhu, M., "Fretting Corrosion Testing of the Taper Interface of Four SOAG Hip
Femoral Stem/Head Combinations," Technical Report 00026G, Department of
Research and Analysis, Sulzer Orthopedics Inc., 2000.
Clinical Experience
Bernhard Berli l, Reinhard Elke2, Erwin W. Morscher 3

The Cemented MS-30 Stem in Total Hip Replacement, Matte versus Polished
Surface: Minimum of Five Years of Clinical and Radiographic Results of a
Prospective Study
Reference: Berli, B., Elke, R., and Morscher, E. W., "The Cemented MS-30 Stem in
Total Hip Replacement, Matte versus Polished Surface: Minimum of Five Years of
Clinical and Radiographic Results of a Prospective Study," Stainless Steels for
Medical and Surgical Applications, ASTM STP 1438, G. L. Winters and M. J. Nutt,
Eds., ASTM International, West Conshohocken, PA, 2003.
Abstract: The MS-30 femoral stem is a three-dimensionally tapered, collarless implant
for cement fixation in THR, manufactured from an FeCrNiMnMoNb-alloy with either a
matte or polished surface. One hundred twenty-seven MS-30 stems with matte surface
and 128 stems with polished surface, respectively, were implanted in an alternated
manner.
The mean observation period was 5.6 years. No patients were lost to the followup. In both groups 88 patients with 108 stems were included in the clinical and
radiographic follow-up examinations. Only one stem with a matte surface was revised
for aseptic loosening. There was no death related to surgery and no infection.
Conclusion: Both, the clinical and radiographic evaluation (radiolucencies,
osteolyses and subsidence) did not show any statistically significant difference between
the two groups.
The main reason for the equally excellent performance of both the matte and the
polished version of the MS-30 stem may be the design of the centralizer which resists
both debonding of the stem, and subsequent subsidence as a precondition for fretting in
the metal/cement interface.
Clinical relevance: Since, on the one hand, debonding of the stem/cement
interface cannot entirely be excluded, we recommend the discontinuation of tapered
stems with a matte surface - especially if they do not have a centralizer which resists the
stem to subside.
Keywords: Total Hip Replacement, Cement Fixation, Tapered Stem, Implant Surface
MD, Orthopaedic Department, University of Basel, Burgfelderstr. 101, CH-4012

Basel, Switzerland
2 Private Dozent, MD, Orthopaedic Department, University of Basel, Burgfelderstr.
101, CH-4012 Basel, Switzerland
3 Professor em., MD, Laboratory of Orthopaedic Biomechanics, University of Basel,
Burgfelderstr. 101, CH-4012 Basel, Switzerland
249
www.astm.org
250
Introduction:
The MS-30 femoral stem (SULZER Orthopedics, Baar/Switzerland, Figure 1),
manufactured in stainless steel (FeCrNiMnMo ISO 5832-9) and designed for cemented
fixation in Total Hip Replacement, was developed by the senior author (EWM) and L.
Spotorno/Italy. The primary goal for the development of this stem was to generate an
optimal cement mantle, which is known to be the most important factor for the
longevity of a cemented femoral stem.
Figure 1 - MS-3O femoral stem with centralizer and modular head. Left: Matte surface,
metallic head, Right: Polished surface, ceramic head.

The quality of the cement mantle itself depends on the implant design, its
surface conditions, the material characteristics (both, of the implant and the cement)
and, last but not least, on the operative, i.e. the cementing technique. In contrast,
inadequate cementing techniques were the main cause of the high rates of aseptic
loosening and, thus,of the increase in revision rates in the past. Main causes of failure
of cemented femoral stems are an insufficient medial support [1], an insufficient
cement mantle with metal/bone contact [2], an insufficient cement/bone interdigitation
[3] and an inadequate positioning of the stem [4 - 6]. Therefore, a series of various
factors, which are closely related to each other, play a decisive role for the outcome of a
THR.
The design of the implant determines the amount and the direction of the forces
transmitted from the bone to the prosthesis and vice versa. Furthermore, the design and
size of the implant determine the thickness of the cement mantle. Preoperative
planning, therefore, is vital for the success of a THR. According to experimental and
clinical experiences, the optimum thickness of the cement mantle can be defined as
asymmetric and non-uniform. It is thicker in the region where the main forces are
transmitted, i.e. in Green zones 7, 3 and 5. Overall, the thickness of the cement mantle
around the stem should be not less than 2 mm [1, 7].
BERLI ET AL. ON CEMENTED MS-30 STEM
The MS-30 stem is three-dimensionally tapered. Thus, if the cement debonds

from the stem, the system functions according to a press-fit concept. The selected stem
is - compared to the medullary canal prepared by reamers - undersized, in order to
provide the necessary space for the bone cement between implant and bone. This is
already taken into account in the preoperative planning. The edges of the stem are
rounded off, in order to minimize stresses within the cement mantle. The flanges on the
proximo-lateral part of the prosthesis increase its rotational stability. Furthermore,
rotational stability is improved by a high neck resection of 30 ~ to the horizontal plane.
An integral part of the system is the centralizer to position the component
distally, hence avoiding metal/bone contact and - in combination with excessive lateral
reaming of the proximal end of the f e m u r - avoiding malaligmnent [8].
The cemented MS-30 stem was introduced in clinical practice with a matte
surface in 1990. From March 1990 until December 31, 2001, 1,465 stems were inserted
at the Orthopaedic Department of the University of Basel/Switzerland.
Aseptic loosening caused by osteolysis through wear particles [9] is still the
most limiting factor for the longevity of a hip arthroplasty. Wear particles are generated
predominantly in the articulating surface of a THR. However, fretting corrosion at the
cement/implant interface has been identified as an additional cause of osteolysis of
cemented femoral stems [10 - 12]. For both, the generation of wear particles and for
fretting corrosion, the implant surface and the quality of the bone cement play a
decisive role.
Only when the interface between the cement mantle and the endoprosthetic stem
is debonded and relative movements allow friction, both wear particles are produced
and corrosion at the cement/stem interface can take place. As the literature documents
that the tapered collarless Exeter stem yielded better results with a polished rather than
a matte surface [13 - 16], SULZER Orthopedics Ltd, Switzerland, also offered the MS30 stem in a polished version. Our own results with this stem with a matte surface
revealed excellent results with no revision and no osteolysis during a four-year period
from 1990 to 1994.
In order to compare the outcome of the MS-30 stem in relation to the surface
characteristics, we decided to perform a prospective study including a total number of
255 stems: 127 with a matte, 128 with a polished surface. The first MS-30 stem of this
series was inserted on November 15, 1994, the last one on March 20, 1997. A
prospective study with the careful use of criteria for inclusion and exclusion allowed
the investigators to minimize the number of variables [17]. The only variable of the
study was the characteristic of the surface.
The main goals of the study were to elucidate whether there is a significant
difference with regard to radiolucencies, osteolyses and stem subsidence on the X-rays,
and the clinical survival rate between stems with matte and polished surface.
Patients and Methods:
The study included 127 stems with a matte and 128 stems with a polished
surface. All but one of these stems were combined with the Morscher Press-Fit Cup and
a 28 mm head (metal or ceramic). In 21 patients, the MS-30 stem was inserted
251
252
bilaterally during the index period, in 16 cases in one operation, and in 5 cases in two
stages. All the implants were assigned by the head nurse of the operating theater in a
randomized, blinded manner. In cases where a patient was to receive a total hip
replacement in both hips in one stage surgery, one prosthesis had a matte, the other one
a polished surface. There was one exception, where a polished stem was inserted in
both hip joints. The operative procedures were performed by 15 different surgeons of
the Orthopaedic Department of the University of Basel/Switzerland, including the
senior author EWM, with different lengths of training and experience levels.
Forty patients (19 women, 21 men)already had undergone a total hip
replacement on the opposite side prior to the index surgery period, and 17 patients (10
women, 7 men), were treated on the opposite side after the index surgery period. Five
patients had a bilateral total hip replacement done in two stages during the index
surgery period (2 with a matte surface and 1 with a polished surface in both hips: 2
patients received a matte in one and a polished stem in the opposite hip). Sixteen
Patients had a bilateral THR done in one stage (15 with a mat on one and a polished on
the opposite side, 1 patient got a polished stem on both sides). In the first group with a
matte surface, there are 61 men with a mean age at surgery of 65 years, and 66 women
with a mean age of 73 years. In the second group with a polished stem surface, there are
63 men with a mean age of 64 years, and 65 women with a mean age of 70 years. The
right hip was treated in 53 cases with a matte, in 73 hips with a polished stem; 71 stems
with a matte and 58 stems with a polished surface were inserted on the left hip. Sixteen
patients in the first group with a matte surface and 14 patients with a polished MS-30
surface in the second group, i.e., a total number of 30 patients (14 men and 16 women),
died during the observation period. Six patients with 7 hips (2 with a matte, 5 with a
polished surface) were not able or not willing to come to the last, 5-year, follow-up.
They were, however, interviewed by telephone. All had their endoprosthesis stem in
situ. One hundred and eight hips with a MS-30 stem with a matte surface and 108 hips
with a polished surface had a minimum 5-year clinical and radiological check. No
patient was lost to follow-up (Table 1).
Table 1- Patients Data: N = 234 Patients/255 Hips, 127 mat, 128 polished surface
Total Uni-lateral THR
Bi-lateral THR
mat
polishe mat-mat mat-polished polished d
polished
mat polished
47
55
Patients male
111
1
6
2
59
52
Patients female
123
1
11
0
65 Y
66 Y
76 Y
60 Y
52 Y
Age at surgery male
(41-86) (35-88)
(52-72)
(44-60)
Age at surgery
72 Y
70 Y
54 Y
65 Y
(54-89) (40-91)
female
(56-73)
Hips male
120
47
55
2
6
6
4
Hips female
135
59
52
2
11
11
0
Stem revisions
1
1
0
0
0
0
Deceased
30/31
15
14
0
1
0
(until Dec. 2001)
Lost for FU
0
0
0
0
0
0
0
1
0
Not able to come
6/7
1
4
Patients controlled
197
89
89
2
15
2
Hips controlled
2/6
89
89
4
15
15
4
Table 2 - Preoperative Dia~,nosis:
Surface
Total matte
Primary Osteoarthritis
198
105 (82.7%)
Avascular Necrosis
24
9 (7.1%)
Dysplasia
17
7 (5.5%)
Posttraumatic OA
9
4 (3.1%)
Rheumatoid Arthritis
7
2 (1.6%)
Total N:
255
127 (100%)
polished
93 (72.7%)
15 (11.7%)
10 (7.8%)
5 (3.9%)
5 (3.9%)
128 (100%)
Out of the 255 hips, the preoperative diagnosis was osteoarthritis in 198 (77 %),
avascular necrosis in 24 (9.5%), dysplasia in 17 (7%), rheumatoid arthritis in 7 (3 %)
and post-traumatic osteoarthritis in 9 cases (3.5%) (Table 2).
Stem sizes were not different in the two groups (p > 0.05) (Fig. 2).
253
254
Figure 2 - Sizes of the MS-30 Stem
The only criterion for exclusion from the study was the surgeon's decision to use
a non-cemented femoral stem. During the study period, this was the case in 52 hips.
As an acetabular component, the senior author's Press-Fit Cup (SULZER
Orthopaedics, Baar/Switzerland) was used for all patients except in one with posttraumatic osteoarthritis, where a Ganz Reinforcement Ring was inserted. The
articulation chosen was metal on conventional polyethylene in 7 cases of MS-stems
with a matte, and in 4 cases with a polished stem, with ceramic on conventional
polyethylene in 120 cases with a matte, and in 124 cases with a polished surface.
The patients were operated on in supine position and all THRs were inserted by
a lateral approach. The rasp is oversized compared to the stem, in order to provide the
necessary space for the cement. The medullary canal was irrigated and packed with
sponges until immediately before introducing the bone cement. The cementing
technique included the use of a cement gun, plugging of the distal femoral canal and
pressurization. The centralizer designed to position the stem distally is an integral part
of the MS-30 system. It serves to prevent contact between implant and bone and to
avoid malalignment. The centralizer selected is in general 1 size larger than the
respective stem size. The femoral component was fixed with Gentamicin-Palacos
(Kulzer GmbH, Bad Homburg/Germany), according to the manufacturer's instructions.
The mean duration of surgery was 11/2hours (range: 1 - 21A hours).
Because the MS-30 stem with distal centralizer was at first introduced in the
matte version only, and because the controversy about the condition of the surface
arose only after that time, the patients were not informed about the nature of the surface
condition of their cemented stem. On the X-rays, it cannot be distinguished whether the
stem surface is matte or polished. The evaluating authors were informed about the
surface condition only after the clinical and radiological evaluation, i.e., the evaluation
part of the study was performed in a double-blind fashion.
The perioperative regimen was the same for all patients. It included the use of
antibiotics (Mandokef, 2 g preoperatively) as well as prophylaxis against venous
thrombosis (Heparin 5000 U and Sintrom from the 2nd postoperative day). The
postoperative treatment (rehabilitation) was uniform in the whole series. The patients
were mobilized either the evening after the operation or the first postoperative morning.
Partial weight bearing was allowed up to half body weight for 6 weeks. Physiotherapy
was prescribed for about 3 months if there was no specific indication for a prolonged
aftertreatment.
The average observation time, i.e. the time between surgery and the last
radiographic follow-up, was 67.3 (60 - 87) months. The minimum observation time,
therefore, was 5 years.
Clinical assessment occurred according to the IDES forms of the Institute of
Documentation of the M.E.Mfiller Foundation in Berne/Switzerland [18].
The clinical evaluation was done according to the Harris Hip Scores [19].
Radiographic evaluation
For the radiographic assessment, an X-ray of the pelvis including both hip joints
was taken. The radiographic appearance of the initial cementing was graded on the 6week check according to the classification of Barrack et al. [1992]: A: complete filling
of the medullary cavity by cement, a so-called "white out" at the cement/bone interface,
B: slight radiolucency of the cement/bone interface, C: radiolucency involving 50% to
99% of the cement/bone interface or a defective or incomplete cement mantle, D:
radiolucency at the cement/bone interface of 100% in any projection, or a failure to fill
the canal with cement to such an extent that the tip of the stem was not covered [20].
Fixation of the acetabular socket (Press-Fit Cup) was also assessed with regard
to osteolyses and radiolucencies in zones I to III of DeLee & Charnley [21].
The radiograph of the latest follow-up was examined for radiolucent lines at the
cement/bone interface which had not been present on the 6-week radiograph. The
cortical bone was checked for the development of osteolyses, the stem for varus/valgus
position, and the cement mantle for cement fractures. Subsidence of the stem within the
cement mantle was measured as the difference of the distance between the upper
circumference of the prosthesis shoulder and the sclerotic line above the shoulder [22,
23].
In the radiological evaluation, osteolysis was defined as a newly developed,
cystic lesion with endosteal scalloping and/or migration, which had not been noticed on
the immediate postoperative radiograph [24]. The sites of radiolucent lines (< 1 mm, 12 mm, >2mm) at the cement/bone interface) and osteolyses were recorded as being
present in one or more of the 7 Gruen zones.
Ectopic ossification was assessed using the method of Brooker et a1.[25].
The data were analyzed and the descriptive statistics (means and standard
deviations) were calculated with the Excel 95 (Microsoft Corp., Redmont, CA)
program. Survivorship was corrected according to Kaplan-Meyer [26].
Results:
Clinical Results
The mean preoperative Harris Hip Score improved from 67 points (range 44 84) to a postoperative score of 93 points (67 - 100). There was no significant difference
between the hips with a matte or polished femoral stem with regard to the overall score
255
256
(p > 0.05), the subscore for pain (p > 0.05) or the subscore for function (p > 0.05) The
overall results were excellent in the group with stems with a matte surface in 79
patients (73%), and in 87 patients (81%) with a polished surface. Good results were
observed in 21patients (19%) and 11 patients (10%), moderate results in 5 patients
(5%) and 7 patients (6%), poor results in 3 patients (3%) with a matte surface, 3
patients (3%) with a polished surface.
None of the 21 patients with bilateral hip replacement, one with a matte MS-30,
the other one with a polished surface, indicated a preference for either hip.
Complications
A preoperative fracture pf the great trochanter occurred in 6 cases, five of them
were stabilized by cerclage wires. In two additional cases a crack of the proximal femur
occurred that had to be eerclaged as well.
There was no death related to surgery and no infection has been observed.
Revisions
One revision had to be done for aseptic loosening of a stem with a matte
surface 4 years after surgery. Two hips (acetabula) had to be revised for recurrent
dislocation. The overall revision rate is, thus, three of 255 (1.2%), and one of 255 for
aseptic loosening (0.4%).
Radiological Results
37 MS-30 stems out of the 216 stems (17%, 27 of the stems with matte and 12
stems with polished surface) showed a varus position between 3 and 6-0(av. 4.2~
Femoral cementing was classified on the postoperative radiographs in the 108
hips with a matte MS-30 stem as grade A in 72, grade B in 34 and grade C in 2 cases.
There was no grade D. In the cohort of stems with a polished surface, the respective
numbers are 78 grade A, 30 grade B and none with grade C and D.
Radiolucencies - all below 2 m m - in 1 to 3 zones were observed in 16 cases of
the group with matte surface and in 15 cases in the group of the polished stems.
The number, the extension and distribution of osteolyses according to the seven
Gruen zones is shown in Table 3.
Subsidence of less than 2 mm was found in 3 cases of the group with a matte
surface and in 2 cases in the group with a polished one, subsidence of 2 - 5 mm in 6
stems with a matte and 7 stems with a polished surface (Table 4).
Table 3 - Osteolyses (accordin8 to the Gruen zones)
Gruen zones Matte surface Polished surface
1
1
0
2
3
4
5
0
0
0
0
6
7
0
1
0
0
0
2
J1 2
1
2
Table 4 - Subsidence
Surface condition Matte surface
Polished surface
< 2mm
3
2
2 - 5 mm
6
7
>5mm
1
0
Subsidence was related significantly to the quality of the cement mantle: out of
150 stems with Barrack A cementing 15 (10%) subsided between I and 3 mrn (av.
1.9ram), out of 64 stems with B cementing 10 (16%) subsided between I and 8 mm
(av. 2.9 mm). The stem that subsided 8 mm has a matte surface and shows severe
osteolysis in the proximal part of the femur. The patient is asymptomatic up to now.
Nevertheless, revision surgery will be necessary in the not too distant future. On 12
radiographs no plug could be detected. Despite this, cementing was qualified A
according to Barrack in 7, as B in 5 cases. Two stems only subsided within the cement
mantle, one 2, the other one 4 mm. In both cases cementing had to be qualified as B.
Two cement fractures in zones 3/5 were observed.
Two periacetabular ossifications (PAO) III according to Brooker were identified
in each group, but no PAO IV were seen.
Acetabular cup
There was no osteolysis and no revision for aseptic loosening, however, two
revisions for dislocation for the Morscher Press-Fit Cup.
Discussion:
In contrast to the experience with the Exeter stem reported by Ling [13], Howie
[14] and the Swedish Implant Register [16], no statistically significant difference in the
outcome between the MS-30 stem with a matte or a polished surface was found in this
prospective study over a 5-year observation period. There is, up to now, no difference
between the MS-30 stem with a matte or a polished surface, not in relation to
survivorship, subsidence or the incidence of osteolyses. However, there seems to be a
tendency for a higher percentage of excellent clinical scores in the polished group vs. a
higher percentage of good clinical scores in the matte group. Furthermore the only
revision for aseptic loosening had to be performed with a stem with a matte surface and
the only THR with severe osteolysis of the femur has a stem with matte surface as well.
A follow-up of 60 to 87 months, with an average time of 5.6 years, is, of course,
too short to give a final verdict, at least concerning survivorship. There is still the
theoretical possibility of a "type-II error", in other words, the probability of concluding
that no difference between groups exists, when, in fact, there is a difference [27]. TypeII errors are to be expected especially in studies with a small sample size. In the present
study, however, the sample size was 127 stems with a matte, and 128 stems with a
polished surface. It was the aim of this study to compare the rate of early loosening of
the MS-30 stem with respect to the surface conditions. Our findings should be
considered sound, as it is known that debonding with subsidence of a stem within the
cement mantle starts within the first 2 years after surgery. Furthermore, it is known that
257
25 8
migration is more pronounced during the first year, then becomes slower and more
continuous [28 - 32].
The main reason for the difference of the rate of osteolyses may be explained
by the difference in the design of the centralizer of the Exeter and the MS-30 stem.
Whereas the centralizer of the Exeter stem is directly put over the stem's tip and has a
hollow space underneath to allow for the stem to subside, the centralizer of the MS-30
stem is made from PMMA and therefore bonds strongly with the surrounding bone
cement and is as well fixed with a metallic pin into a hole at the tip of the stem. This
provides resistance against debonding and subsidence.
The rationale to allow a tapered stem to subside (migrate) is the recognition of
the fact that subsidence within the cement mantle has not only no correlation with
regard to pain [13, 38], but allows the stem to restabilize as is also shown with the
tapered M/iller Straight Stem [37, 38].
According to the clinical experience with the Exeter stem [13, 14, 33, 34] it
must be assumed that once debonded a MS-30 stem with a matte surface will produce
more panicles and/or corrosion compared to a polished surface. Despite the almost
equally good results observed to date with the matte version, we, therefore, recommend
the exclusive use of the MS-30 stems with polished surface. Furthermore, it has been
shown that subsidence of a tapered polished stem is not harmful, but quite in contrast,
is advantageous [13, 33].
On the other hand, according to studies of stem surface roughness and creepqnduced subsidence of Norman et a1.[35], it appears that stem subsidence is not important
for the maintenance of a "taper-lock"; and creep-induced subsidence does not result in
an increase of normal stress patterns at the stem/cement interface as previously
proposed [36].
When designing the MS-30 stem, the main goal was to improve the quality of
the cement mantle without abandoning the proven concept of a tapered design ( pressfit principle), realized, i.e., in the M/iller Straight Stem or the Exeter Stem. Thus, there
is the possibility of a "second line of defense" in case of debonding of the metal/cement
interface with migration (subsidence) [37].One has to bear in mind that, with a straight
stem, the risk of a thin cement mantle in zones 8 and 9 increases [38]. In order to avoid
this risk of metal/bone contact, Wroblewski et al. [6] and Breusch et a1.[39] consider a
low-neck osteotomy and aggressive removal of the posterior calcar femorale a
necessity. However, with a low resection of the femoral neck, rotational stability is
seriously compromised [40, 41]. Since loosening of a femoral stem rotates the stem into
retroversion and a high neck resection secures the prosthesis stem effectively against
migration into retroversion, we prefer the "high", i.e., 30 ~ instead of 45 ~ resection, as it
is part of the "fixation philosophy" of the MS-30 stem.
In conclusion, no significant difference in the outcome of MS-30 stems with a
matte or a polished surface could be found during a 5-year observation period. Neither
did the incidence of radiolucencies, osteolyses or subsidence reveal a significant difference between the two surfaces. The main reason for the almost equally excellent performance of both the matte and polished version of the MS-30 in contrast to the matte
Exeter stem may be the design of the centralizer of the MS-30 stem, which resists both,
a debonding of the stem from the cement, and subsequent subsidence as a precondition
for fretting between the two surfaces at the metal/cement interface. Another reason for
the fact that no significant difference was found to date between the two implant
surfaces, may be differences in the bone cement used in the different studies. However,
this is a matter of ongoing research at the ,,Laboratory of Orthopaedic Biomechanic,, of
the University of Basle/Switzerland. Furthermore, since it could be shown in our own
experiments that the main mechanism for inducing osteolysis is fretting corrosion at the
interface of bone cement (PMMA) and a matte stainless steel surface [12], only tapered
stems with a polished surface must be used in combination with a hollow centralizer.
References:
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
Ebramzadeh, E., Sarmiento, A., McKeUop, H.A., Llinas, A., Gogan, W. "The
cement mantle in total hip arthroplasty" J Bone Joint Surg Am 76, 1994, pp.7787.
Draenert, K. and Draenert, Y. "Die Adaptation des Knochens an die
Deformation durch Implantate - Strain-Adaptive Bone Remodelling". Art &
Science Miinchen, 1992
Miller, J. and Johnson, A. "Advances in cementing techniques in total hip
arthroplasty". In: Stilwell, W.T. (ed.) The art of total hip arthroplasty, 1987, pp.
277-292, Grunde & Stratton.
Markolf, K.L. and Amstutz, H.C. "A comparative experimental study of stresses
in femoral total hip replacement components: the effects of prosthesis
orientation and acrylic fixation". J Biomech. 9, 1976, pp. 73-79.
Gruen, T.A., McNeice, G.M. and Amstutz, H.C. "Modes of failure of cemented
stem-type femoral components. A radiographic analysis of loosening". Clin
Orthop. 141, 1979, pp.17-27.
Wroblewski, B.M., Siney, P.D., Fleming, P.A. and Bobak, P. "The calcar
femorale in cemented stem fixation in total hip arthroplasty". J Bone Joint Surg
Br 82, 1979, pp. 842-845.
Estok, D.M., Orr, T.E. and Harris, W.H. "Factors affecting cement strains near
the tip of a cemented femoral component". J Arthroplasty. 12, 1997, pp. 40-48
Morscher, E.W., Spotorno, L., Mumenthaler, A. and Frick, W. "The Cemented
MS-30 Stem". In: Morscher, E.W.(Ed.) Endoprosthetics, 1995, pp.211-219,
Springer Berlin, New York.
Willert, H.G. and Semlitsch, M. "Reactions of articular capsule to wear
products of artificial joint prostheses". J Biomed Mater Res. 11, 1977, pp. 157164.
Morscher, E.W., Wirz, D. "Current state of cement fixation in THR". Acta
Orthop Belg. 2002 (in press).
Tritschler, B., Forest, B. and Rieu, J. "Fretting corrosion of materials for
orthopaedic implants: a study of a metal/polymer contact in an artificial
physiological medium. Tribol Int. 32, 1999, pp. 587-596.
Wirz, D., Zurfluh, B., Goepfert, B., Li, F., Frick, W., Klabunde, R., Stoll, D.
and Morscher, E.W. "Results of in vitro studies about the mechanism of wear in
the stem-cement interface of THR, ASTM, Symposium on Steels for Medical
and Surgical Applications, Pittsburgh PA,USA, May 6-7, 2002.
259
260
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
Ling,R.S. "The use of a collar and precoating in cemented femoral stems is unnecessary and detrimental". Clin Orthop. 285, 1992, pp. 73-83.
Howie, D.W., Middleton, R.G. and Costi, K. "Loosening of matte and polished
cemented femoral stems". J Bone Joint Surg. Br 80, 1998, pp. 573-576.
Rockborn, P. and Olsson, S.S. "Loosening and bone resorption in Exeter hip
arthroplasties. Review at a minimum of five years". J Bone Joint Surg. Br 75,
1993, pp. 865-868.
Malchau, H. and Herberts, P. "Prognosis of total hip replacement. Revision and
rerevision rate in THR: A revision risk study of 148, 359 primary operations".
Scientific exhibition, 65th Annual Meeting AAOS, February 19-23, 1998, New
Orleans/USA
Clark, C.R. "The prospective, randomized, double-blind clinical trial in
orthopaedic surgery". J Bone Joint Surg. Am 79, 1997, pp. 1119-1120.
Johnston, R.C., Fitzgerald, R.H., Harris, W.H., Poss, R., Miiller, M.E. and
Sledge, C.B. "Clinical and radiographic evaluation of total hip replacement". J
Bone Joint Surg. Am 72, 1990, pp. 161-168.
Harris, W.H. "Traumatic arthritis of the hip after dislocation and acetabular
fracture: treatment by mold arthroplasty. J Bone Joint Surg. Am 51, 1969, pp.
737-755.
Barrack, R.L., Mulroy, R.D. Jr and Harris, W.H. "Improved cementing
techniques and femoral component loosening in young patients with hip
arthroplasty. A 12-year radiographic review". J Bone Joint Surg. Br 74, 1992,
pp. 385-389.
DeLee, J.G. and Charnley, J. "Radiological demarcation of cemented sockets in
hip replacement". Clin Orthop. 121, 1976, pp. 20-33.
Kelly, A.J., Lee, M.B., Wong, N.S., Smith, E.J. and Learmonth, I.D. "Poor
reproducibility in radiographic grading of femoral cementing technique in total
hip arthroplasty". J Arthroplasty. 11, 1996, pp. 525-528.
Acklin, Y.P., Bedi, B.J., Frick, W., Elke, R. and Morscher, E.W. "Nine-year
results of Miiller cemented titanium straight stems in total hip replacement".
Arch Orthop Trauma Surg. 121, 2001, pp. 391-398.
Joshi, R.P., Eftekhar, N.S., McMahon, D.J. and Nercessian, O.A. "Osteolysis
after Chamley primary low-friction arthroplasty". J Bone Joint Surg. Br 80,
1998, pp. :585-590.
Brooker, A.F., Bowerman, J.W., Robinson, R.A. and Riley, L.H. Jr "Ectopic
ossification following total hip replacement. Incidence and a method of
classification". J Bone Joint Surg. Am 55, 1973, pp. 1629-1632.
Kaplan, E.L. and Meier, P. "Nonparametric estimation from incomplete
observations". J Am Star Assoc. 53, 1958, pp. 457-481.
Lochner, H.V., Bhandari. M. and Tometta III, P. "Type-II error rates (Beta
Errors) of randomized trials in orthopaedic trauma". J Bone Joint Surg. Am 83,
2001, pp. 1650-1655.
Alfaro-Adri~in, J., Gill, H.S., Murray, D.W. "Should total hip arthroplasty
femoral components be designed to subside? A radiostereometric analysis study
of the Charnley elite and Exeter stems". J Arthroplasty. 16, 2001, pp. 598-606.
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
Harris, W.H., McCarthy, J.C. and O'Neill, D.A. "Femoral component loosening
using contemporary techniques of femoral cement fixation". J Bone Joint Surg.
Am 68, 1986, pp. 1064-1066.
Kiss, J., Murray, D.W., Turner-Smith, A.R., Bithell, J. and Bullstrode, C.J.
"Migration of cemented femoral components after THR: Roentgen
stereophotogrammetry analysis". J Bone Joint Surg. Br 78, 1996, pp. 796-801.
K~irrholm,J. and Snorrason, F. "Subsidence, tip and hump micromovement of
noncoated ribbed femoral prostheses". Clin Orthop. 287, 1993, pp. 50-60.
Soballe, K., Toksvig-Larson, S. and Gelinek, J. "Migration of hydroxyapatite
coated femoral prostheses: a roentgen stereophotogrammetric study". J Bone
Joint Surg. Br 75, 1993, pp. 681-687.
Fowler, J.L., Gie, G.A., Lee, A.J.C. and Ling, R.S.M. "Experience with the
Exeter total hip replacement since 1970". Orthop Clin North Am. 19, 1988, pp.
477-489.
Alfaro-Adri~n, J., Gill, H.S. and Murray, D.W. "Cement migration after THR.
A comparison of Charnley elite and Exeter femoral stems using RSA". J Bone
Joint Surg. Br 81, 1999, pp. 130-134.
Normann, T.L., Thyagarajan, G., Saligrama, V.C., Gruen, T.A. and Blaha, J.D.
"Stem surface roughness alters creep induced subsidence and "taper-lock" in a
cemented femoral hip prosthesis". J Biomech. 34, 2001, pp. 1325-1333.
Lee, A.J.C., Perkins, R.D., Ling, R.S.M. "Time dependent properties of
polymethylmethacrylate bone cement". In: John, O. (Ed) In implant bone
interface. Springer New York, 1990, pp. 85-90.
Wilson-MacDonald, J. and Morscher, E. "Comparison between straight- and
curved-stem Mtiller femoral prostheses; 5- to 10-year results of 545 total hip
replacements". Arch Orthop Trauma Surg. 109, 1989, pp. 14-20.
R~iber,D.A., Czaja, S, and Morscher, E.W. "Fifteen-year results of the Miiller
CoCrNiMo straight stem". Arch Orthop Trauma Surg. 121, 2001, pp. 38-42.
Breusch, S.J., Lukoschek, N., Kreutzer, J., Brocai, D. and Gruen, T.A.
"Dependency of cement mantle thickness on femoral stem design and
centralizer". J Arthroplasty. 16, 2001, pp. 648-657.
Freeman, M.A.R. "Why resect the neck?". J Bone Joint Surg. Br 68, 1986, pp.
346-349.
Nunn, D., Freeman, M.A.R., Tanner, K.E. and Bonfield, W. "Torsinal stability
of the femoral component of hip arthroplasty. Response to an anteriorly applied
load". J Bone Joint Surg. Br 71, 1989, pp. 452-455.
261
Robert M. Urban, 1Joshua J. Jacobs, 1 Jeremy L. Gilbert,2 Anastasia K. Skipor, 1 Nadim J.

Hallab, 1 Katalin Mikecz, ] Tibor T. Giant, 1 J. Lawrence Marsh, 3 and Jorge O. Galante 1
Corrosion Products Generated from Mechanically Assisted Crevice Corrosion of

Stainless Steel Orthopaedic Implants
Reference: Urban, R. M., Jacobs J. J., Gilbert, J. L., Skipor, A. K., Hallab, N. J., Mikecz,
K., Giant, T. T., Marsh, J. L., and Galante, J. O., "Corrosion Products Generated from
Mechanically Assisted Crevice Corrosion of Stainless Steel Orthopaedic Implants,"
Stainless Steels for Medical and Surgical Applications, ASTM STP 1438, G. L. Winters
and M. J. Nutt, Eds., ASTM International, West Conshohocken, PA, 2003.
Abstract: Accelerated corrosion of metallic implants in vivo can generate both soluble
and insoluble products that can be detected locally and systemically. Retrieved stainless
steel implants for trauma fixation or spinal instrumentation demonstrate iron and
chromium-containing solid products of corrosion deposited around corroded modular
junctions and as phagocytosable particles in the adjacent tissues. In some cases, the
resulting adverse local tissue reaction has been associated with pain, inflammation and
osteolysis, requiring removal of the implant. In vitro cell and organ culture studies
confirm that corrosion products such as particles of chromium phosphate can elicit
proinflammatory cytokine secretion from macrophages and promote macrophagemediated bone resorption. Systemically, soluble corrosion products of chromium can be
detected in the serum of selected patients with accelerated corrosion of chromiumcontaining implants. Metal-protein binding studies indicate that the high molecular
weight serum proteins including immunoglobulins have the highest affinity for
chromium. These findings stress the importance of the design of modular junctions to
minimize corrosion of stainless steels used in orthopaedic appliances.
Keywords: Stainless steel, corrosion, crevice, fretting, modularity, chromium, particulate
debris, metal ions, osteolysis, cytokines, serum proteins, systemic dissemination
ZDirector, Implant Pathology Laboratory, Professor and Director, Section of Biomaterials
Research, Director, Metal Ion Laboratory, Assistant Professor and Director,
Biomaterials Testing Laboratory, Associate Professor, Professor and Director,
Section of Biochemistry and Molecular Biology, and Professor, respectively,
Department of Orthopedic Surgery, Rush-Presbyterian-St. Luke's Medical Center,
1653 W. Congress Parkway, Chicago, IL 60612.
2professor and Chairman, Department of Bioengineering and Neuroscience, Syracuse
University, Syracuse, NY 13244.
3professor, Department of Orthopaedic Surgery, the University of Iowa College of
Medicine, Iowa City, IA 52242.
262
Copyright*2003by ASTMInternational
www.astm.org
URBAN ET AL. ON MECHANICALLY ASSISTED CREVICE CORROSION
263
Introduction
Type AISI 316L (ASTM F138) stainless steel continues to be a dominant alloy for
fabrication of internal fixation devices where biocompatibility, high strength and ductility
are required. This paper summarizes the clinical, histopathological, micro-analytical and
in vitro findings from previously reported and ongoing studies in our laboratories as they
relate to the nature and significance of corrosion products generated by a variety of
contemporary spinal instrumentation mad trattma fixation devices made from stainless
steel.
Corrosion of Stainless Steel Implants
Corrosion of single-part internal fixation devices is rare [ 1]. Retrieved stainless steel
hardware generally demonstrates a bright, "as-implanted" finish indicative of the low rate
of uniform dissolution of this alloy in its passivated state. The multi-part or modular
design of many internal fixation devices consistently introduces the potential for crevice
corrosion, to which all stainless steels are prone [2]. Screw and plate interfaces in fracture
fixation devices and hook, screw and rod systems for spinal instrumentation provide
environments for fretting and crevice corrosion, which can challenge even the most
corrosion-resistant alloys. The elements of these interface corrosion processes have been
well described and are the subject of current review articles [3,4].
As a result of the susceptibility of multi-component stainless steel implants to
interface corrosion, the incidence of observed localized corrosion in retrieved implants is
high. A study of 250 multi-part stainless steel fracture fixation devices removed in the
late 1970s and early 1980s showed some degree of crevice corrosion in 89% of plates and
88% of screws [5].
In a previously unreported study conducted in our laboratory, 24 consecutively
retrieved, contemporary, stainless steel spinal instrumentation devices (Table 1) removed
after a mean of 68 months (range 4 to 277 months) were examined using a stereo light
microscope (8 to 75 X) for evidence of interface corrosion. In selected cases, scanning
electron microscopy with energy-dispersive X-ray analysis (model JSM-5900LV, JEOL)
was used to determine the elemental composition of adherent corrosion products adjacent
to the modular junctions. Ninety-two percent of the spinal devices demonstrated localized
corrosion at the junctions between rods and connectors. Iron and chromium-containing
degradation products were identified surrounding the sites of corrosion (Figure 1).
Despite the high incidence of interface corrosion with stainless steel implants, most
often, the reaction is focal and usually not severe enough to require implant removal [1,
4, 6]. "Metal intolerance" manifested as late operative site pain is one reason for removal
of stainless steel devices. In practice, however, the relationship among corrosion, local
inflammation and pain is often difficult to demonstrate. Infection [7], mechanical
irritation, or rarely metal allergy [8] plays a possible role in some cases. In plate and
screw devices for fracture fixation, accelerated corrosion occurs at the very small area of
screw/plate contact, limiting the amount of degradation products and the magnitude of
the response in the surrounding tissues. Hence, the importance of corrosion as a clinical
phenomenon is uncertain in this circumstance, However, oxide abrasion can induce large
potential (electric field) transients that may play a role in pain generation [9].
Accelerated corrosion of larger modular junctions of different geometry provides
26 4
Table 1 - Retrieved Stainless Steel Spinal Implants

Implant
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Instrumentation
System 1
Harrington
Rogozinski
ISOLA
ISOLA
ISOLA
TRSH
Simmons
C-D
C-D
Harfington
TRSH
ISOLA
TRSH
C-D
C-D
Simmons
ISOLA
ISOLA
C-D
ISOLA
ISOLA
ISOLA
Simmons
ISOLA
Interface
Corrosion
(Yes / No)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Duration
(Mos.)
208
103
13
33
17
47
13
38
60
277
41
35
175
103
145
44
33
25
86
4
47
10
58
16
Reason for Removal
Flat back deformity

Transition syndrome
Pain, flat back deformity
Flat back deformity
Transition syndrome
Infection
Fell, numbness
Pseudoarthrosis
Prominent hardware
Flat back deformity
Rod breakage
Flat back deformity
Flat back deformity
Broken rod
Broken hardware
Low back pain
Infection
Pain from iliac screws
Painful hardware
Flat back deformity
Transition syndrome.
Flat back deformity
Pseudoarthrosis.
Extension of fusion
1Harrington (Zimmer, Warsaw, IN); Rogozinski (Surgical Dynamics, Norwalk CT);

ISOLA (DePuy AcroMed, Warsaw, IN); TRSH, Texas Rite Scottish Hospital (Sofamor
Danek, Memphis, TN); C-D, Cotrel-Dubousset (Sofamor Danek, Memphis, TN);
Simmons (Smith & Nephew Richards, Memphis, TN).
increased contact area and does appear to generate a greater volume of degradation
products that can result in local inflammation and pain, necessitating removal of an
implant. A well-studied example of clinically significant local reactions to corrosion and
metallic degradation products was described by Jones et al. [ 10], who reported interface
corrosion of a three-piece modular 316L stainless steel intramedullary nail used to treat
femoral diaphyseal fractures. Osteolysis, periosteal reaction or cortical thickening
localized to one or both of the modular junctions were present in 23 of 27 femora at a
mean of 21 months following fracture fixation (Figure 2). The level of thigh pain was
significantly greater in patients with the modular nails compared to a similar group of
patients whose femoral fractures had been treated with single-piece stainless steel nails.
Fretting corrosion at the modular junctions and a granulomatous reaction to particulate
corrosion products in the adjacent tissues was noted in retrieved specimens.
URBAN ET AL. ON MECHANICALLYASSISTED CREVICE CORROSION
F i g u r e 1 - Spinal instrumentation rod removed after 33 months for infection

demonstrated severe pitting at a rod~connector junction. Adherent corrosion products
(arrows) surrounded the site of crevice corrosion.
Radiograph (Left) shows periosteal reaction (white arrows) and focal

osteolysis (black arrows) at the level of a junction of a modular stainless steel
intramedullary nail A retrieved nail (Righ0 shows adherent corrosion products
surrounding a modular junction.
Figure 2 -
265
26 6
Solid Corrosion Products and Local Tissue Effects
Generalized corrosion currently is not known to be of consequence in the tissues

adjoining stainless steel implants, since in histological sections there is no evidence of a
specific response to the slow release of metallic species that characterizes this process
[4]. However, insoluble metallic degradation products are often found surrounding sites
of accelerated corrosion. These products exist as deposits adherent to the implant and as
particulate debris, which can provoke an inflammatory response in the peri-implant
tissues. The adherent deposits may be recognized as black, dark green and rust-colored
encrustations associated with interface corrosion at screw-plate junctions seen at routine
removal of stainless steel internal fixation devices. Both the adherent deposits and the
particulate debris within the tissue can be identified as metallic degradation products
having similar compositions.
The nature of this particulate debris was first reported by Williams and Meachim [11],
Sevitt [12] and Winter [13], who described two types of corrosion products in the tissues
surrounding stainless steel implants retrieved in the 1970s. The first of these particles was
a chromium compound containing iron and a substantial amount of phosphorus that
appeared in histological sections as yellow or "apple-green" plate-like particles. These
particles were termed, "microplates". The microplates ranged in size from as much as
several hundred micrometers down to sub-micron dimensions. The second type of
particles consisted of abundant, 1-3 micrometer, "yellow-brown", hemosiderin-like
granules identified as oxides and hydrated oxides of iron.
More recent studies indicate that corrosion of contemporary stainless steel devices
generates degradation products similar in morphology and elemental composition to
those described above for implants retrieved in the 1970s. Our laboratory conducted
supplementary analyses on the retrieved implants and peri-implant tissues from the study
of Jones et al. [ 10] concerning corroded modular intramedullary nails. Correlated
histological, energy-dispersive X-ray analysis (JSM-5900LV, JEOL) and selected area
electron diffraction (JEM-4000FX, JEOL) revealed the existence of at least two types of
microplates. No crystallinity was detected in either type of microplate particle. The
microplates that appeared yellow by transmission light microscopy of the histological
sections proved on microanalysis to be an iron phosphate hydroxide compound (Figure
3). The microplates appearing green in the light microscope were confirmed to be a
hydrate-rich, chromium phosphate compound (Figure 4). The latter material has been
described previously in association with corrosion of cobalt-chromium alloy head/neck
junctions in hip replacement prostheses [14]. Nanometer size particles of chromium
phosphate have been identified within macrophages in peri-implant tissue. The valance of
chromium in this material has been determined to be 3+ using comparisons of parallel
electron energy loss spectra [ 15]. Microptate-like particles have been reported in
association with corroded junctions of spinal instrumentation devices as well [ 16, 17].
Energy-dispersive x-ray analysis of the yellow-brown, hemosiderin-like granules
indicated high concentrations of oxygen and iron and traces of chromium (Figure 5).
Histopathological examination of the tissues adjacent to the modular junctions of the
intramedullary nails demonstrated an inflammatory response to the various-size particles
of the corrosion products. The cellular response to the two types of microplates, iron
phosphate hydroxide and chromium phosphate, was similar (Figures 6 and 7).
Multinucleated giant cells and macrophages surrounded large deposits of microplates;
267
Fe
C
o
Ca
T
Sill
0
Cr
, ~ ' , , , ' , I ' , , + ~ , , , i
. . . . . . .
"I
....
, ....
i ,~',,
I ....
keV
' ....
'"'
' ',
F i g u r e 3 - Energy-dispersive X-ray spectrum of a "'yellow microplate "particle identified

as an iron phosphate hydroxide corrosion product.
Ca
Cr
0
Cr
C
0
U
N
T
S
Ca
Fe
keV
F i g u r e 4 - Energy-dispersive X-ray spectrum of a "green microplate"particle identified

as a chromium phosphate hydrate-rich corrosion product.
Ca
Fe
0
Cr
Ca
Fe
keV
Figure 5 - E n e r g y - d i s p e r s i v e X-ray spectrum of a "yellow-brown ", hemosiderin-like

particle identified as an iron oxide corrosion product containing trace amounts of
chromium.
268
STAINLESSSTEELS FOR MEDICAL AND SURGICALAPPLICATION'S
Figure 6 - Muhinueleated giant cells (center) adjacent to an accumulation of "yellow
microplate" particles (arrows), identified as an iron phosphate hydroxide, in the periimplant tissue from a corroded stainless steel modular intramedullary nail (H&E, X78).
Figure 7 - Macrophages surround a collection of "green microplate "particles (arrows),

identified as chromium phosphate, in the tissue adjacent to a corroded modular stainless
steel intramedullary nail removed after 20 months. (H&E, X234).
while microplates of 5 micrometers and less were found within macrophages. Abundant
iron oxide granules were present within macrophages and occasionally fibroblasts
(Figure 8). The macrophages and multinucleated giant cells were accompanied by a mild
infiltrate of lymphocytes and plasma cells. As would be expected, the magnitude of the
cellular response was dependent on the size and concentration of the particles.
In vitro cell and organ culture studies have conftrmed that phagocytosable chromium
269
phosphate particles (1.42 + 0.83 micrometers) can elicit proinflammatory cytokine (ILlet, IL-113 and prostaglandin E2) secretion from macrophages, promote macrophagemediated bone resorption [18], and suppress osteoblast procollagen gene expression
[19].
Figure 8 - Hemosiderin-like "yellow-brown"l to 3 micrometer particles imparted a

granular appearance to numerous macrophages (arrows) in this tissue adjacent to a
corroded modular stainless steel intramedullary nail removed after 20 months. The
particles were identified as iron oxides containing traces of chromium. (H&E, X156).
Soluble Corrosion Products and Systemic Dissemination
Much of the metal released by corrosion of implants in vivo is transported

systemically as minute particles and organometallic complexes via the lymphatic and
blood circulatory systems. This is evident from post-mortem studies of implants and
remote organs and from studies of metal ion concentrations in serum and urine of
patients hosting metal implants. Case [20] studied specimens from 13 cadavers with
implants, including 5 with stainless steel hip screws and 5 with stainless steel hip or
knee replacement prostheses, using inductively coupled plasma-mass spectroscopy.
Increased metal levels thought to be from the implants were found in lymph nodes,
liver, spleen, and bone marrow. Post-mortem retrieval studies from our laboratories
using electron microprobe analysis have shown that systemic distribution of metallic
particles generated by total joint replacements to the liver, spleen and lymph nodes is
common [21 ]. One case also demonstrated dissemination of corrosion products from the
screw/plate interfaces of a stainless steel femoral fixation plate to the para-aortic lymph
270
nodes where abundant particles of chromium phosphate were detected. Both of these
post-mortem studies reported that the highest concentrations of disseminated metal were
associated with mechanical failure of the implant.
Chromium can be released and disseminated systemically from stainless steel or
cobalt-chromium alloys. Increased concentrations of metals in the serum and urine,
particularly chromium, have been associated with interface corrosion of multi-part
stainless steel as well as cobalt-chromium implants. In the study of Jones et al. [10],
patients with corrosion at the modular junctions of a three-piece femoral intramedullary
stainless steel nail had a 20-fold increase in the mean level of serum chromium compared
to a control group o f patients without metallic implants, and a fourfold increase compared
to a group with one-piece stainless steel nails. Another study from our laboratories
suggests that fretting corrosion at the modular head/neck coupling of otherwise wellfunctioning cobalt-chromium hip replacements was responsible for increases in serum
and urine chromium levels of as much as fivefold and eightfold, respectively, over
preoperative levels [22]. Serum protein-metal binding studies of patients with cobaltchromium alloy hip prostheses indicate that chromium is bound to serum proteins within
two specific molecular weight ranges at approximately 70 and 180 kilodaltons [23]. The
higher molecular weight range, which includes the immunoglobulins, had the highest
affinity for chromium.
Summary
Interface corrosion of multi-component stainless steel implants can generate both
soluble and insoluble metallic degradation products that can be detected locally and
systemically. The local tissue response to corrosion of screw/plate junctions is limited
and not of clinical importance in most circumstances. However, accelerated corrosion of
modular junctions with different geometries can generate a greater volume of degradation
products, resulting in local inflammation, pain, and the need for removal of an implant.
Spinal instrumentation systems employ connectors of various geometries. Neither the
local tissue response, the potential effects on the proximate neural elements, nor the
systemic metal levels have been studied comprehensively in patients with these devices.
Systemically, serum chromium levels can be substantially elevated in patients with
accelerated corrosion at modular junctions of stainless steel devices and may be used as a
marker for mechanical dysfunction. The long-term biological significance of these
circulating organometallic complexes is currently unknown. These findings stress the
importance of the design of modular junctions to minimize corrosion of stainless steels
used in orthopaedic appliances.
Acknowledgment
This work was supported by grant AR39310 from the National Institute of Arthritis and
Musculoskeletal and Skin Diseases.
References
[1]
Black, J., Biological Performance of Materials. Fundamentals of

Biocompatibility, Second Edition, Marcel Decker, New York, 1992.
URBAN ET AL. ON MECHANICALLYASSISTED CREVICE CORROSION
271
[2]
Rostoker, W., and Galante, J. O., "Materials for Human Implantation," Journal of
Biomechanical Engineering, Vol. 101, 1979, pp. 2-14.
[3]
Gilbert, J. L., and Jacobs, J. J., "The Mechanical and Electrochemical Processes
Associated with Taper Fretting Crevice Corrosion: A Review," Modularity of
Orthopedic Implants, ASTM 1301, Donald E. Marlowe, Jack E. Parr and Michael
B. Mayor, Eds., American Society for Testing and Materials, West
Conshohoeken, PA, 1997, pp. 45-59.
[4]
Jacobs, J. J., Gilbert J. L., and Urban, R. M., "Corrosion of Metallic Implants,'!
Journal of Bone and Joint Surgery, Vol. 80-A, 1998, pp. 268-282.
[5]
Cook, S. D., Thomas, K.A., Harding, A. F., Collins, C.L., Haddad, R. J., Jr.,
Millicic, M., and Fischer, W. L., "The In Vivo Performance of 250 Internal
Fixation Devices. A Follow-Up Study," Biomaterials, Vol. 8, 1987, pp. 177-184.
[6]
Sutow, E. J., and Pollack, S. R., "The Biocompatibility of Certain Stainless

Steels," Biocompatibility of Clinical Implant Materials, Vol. 1, David F.
Williams, Ed., CRC Press, Boca Raton, 1981, pp. 45-98.
[7]
Gaine, W. J., Andrew, S. M., Chadwick, P., Cooke, E., and Williamson, J. B.,
"Late Operative Site Pain With Isola Posterior Instrumentation Requiring Implant
Removal. Infection or Metal Reaction?," Spine, Vol. 26, Number 5,2001, pp.
583-587.
[8]
Hallab, N., Merritt, K., and Jacobs, J. J., "Metal Sensitivity in Patients with
Orthopaedic Implants," Journal of Bone and Joint Surgery, Vol. 83-A, 2001, pp.
428-436.
[9]
Goldberg, S., Gilbert, J.L., "Transient Electric Fields Induced by Mechanically

Assisted Corrosion of Ti-6A1-4V," Journal of Biomedical Materials Research,
Vol. 56, 2001, pp. 184-194.
[10]
Jones, D. M., Marsh, J. L., Nepola, J. V., Jacobs, J. J., Skipor, A. S., Urban, R.
M., Gilbert, J. L., and Buchwalter, J. A., "Focal Osteolysis at the Junctions of a
Modular Stainless-Steel Femoral Intramedullary Nail," Journal of Bone and Joint
Surgery, Vol. 83-A, 2001, pp. 537-548.
[11]
Williams, D. F., and Meachim, G., "A Combined Metallurgical and Histological
Study of Tissue-Prosthesis Interactions in Orthopaedic Patients," Journal of
Biomedical Material Research Symposium 5 (part/), 1974, pp. 1-9.
[12]
Sevitt, S., "Corrosion of Implants and Tissue Metallosis," Bone Repair and
Fracture Healing in Man, Churchill and Livingstone, Edinburgh, 1981, pp. 281295.
[13]
Winter, G.D., "Wear and Corrosion Products in Tissues and the Reactions They
272
Provoke," Biocompatibility of Implant Materials, David F. Williams, Ed., Sector

Publishing, London, 1976, pp. 28-39.
[141
Urban, R. M., Jacobs, J. J., Gilbert, J. L., and Galante, J. O., "Migration of
Corrosion Products from Modular Hip Prostheses," Journal of Bone and Joint
Surgery, Vol. 76-A, 1994, pp. 1345-1359.
[15]
Jacobs, J. J., Urban, R. M., Gilbert, J. L., Skipor, A. S., Black, J., Jasty, M., and
Galante, J. O., "Local and Distant Products of Modularity," Clinical Orthopaedics
and Related Research, Vol. 319, 1995, pp. 94-105.
[161
Mody, D. R., Esses, S. I., and Heggeness, M. H., "A Histologic Study of SoftTissue Reactions to Spinal Implants," Spine, Vol. 19, Number 10, 1994, pp. 11531156.
[17]
Vieweg, U., van Roost, D., Wolf, H. K., Schyma, C.A., and Schramm, J.,
"Corrosion on an Internal Spinal Fixator System," Spine, Vol. 24. Number 10,
1999, pp. 946-951.
[181
Lee, S. H., Brennan, F. R., Jacobs, J. J., Urban, R. M., Ragasa, D. R., and Glant,
T. T., "Human Monocyte/Macrophage Response to Cobalt-Chromium Corrosion
Products and Titanium Particles in Patients with Total Joint Replacements,"
Journal of Orthopaedic Research, Vol. 15, 1997, pp. 40-49.
[19]
Vermes, C., Chandrasekaran, R., Jacobs, J. J., Galante, J. O., Roebuck, K. A., and
Glant, T. T., " The Effects of Particulate Wear Debris, Cytokines, and Growth
Factors on the Functions ofMG-63 Osteoblasts," Journal of Bone Joint Surgery,
83-A, 2001, pp. 201-211.
[20]
Case, C. P., Langkamer, V. G., James, C., Palmer, M. R., Kemp, A. J., Heap, P.
F., and Solomon, L., Widespread Dissemination of Metal Debris from Implants,"
Journal of Bone and Joint Surgery, Vol. 76-B, 1994, pp. 701-712.
[21]
Urban, R. M., Jacobs, J. J., Tomlinson, M. J., Gavrilovic, J., Black, J., and Peoc'h,
M., "Dissemination of Wear Particles to the Liver, Spleen and Abdominal Lymph
Nodes of Patients with Hip or Knee Replacement," Journal of Bone and Joint
Surgery, Vol. 82-A, 2000, pp. 457-477.
[221
Jacobs, J. J., Skipor, A. S., Patterson, L. M., Hallab, N. J., Paprosky, W. G.,
Black, J., and Galante, J. O., "Metal Release in Patients Who Have a Primary
Total Hip Arthroplasty," Journal of Bone and Joint Surgery, Vol. 80-A, 1998, pp.
1447-1458.
[23]
Hallab, N. J., Jacobs, J. J., Skipor, A. S., Black, J., Mikecz, K., and Galante, J. O.,
"Systemic Metal Binding Associated with Total Joint Replacement Arthroplasty,"
Journal of BiomedicalMaterials Research, Vol. 49(3), 2000, pp. 353-361.
STP1438-EB/Jan.2003
Author Index
A
Hallab, Nadim J., 262

Haraldsson, Christina, 3
Hicks, Albert G., 61
Hirsiger, Werner, 93
Agonstinho, Silvia M. L., 168

B
Barbosa, Celso A., 168

Befli, Bernhard, 249
Bigolin, Gianni, 93
Bingmann, Dieter, 119
Bogan, Jay-Anthony, 107, 154, 194
Boucher, Bruno, 13
Brauer, Holger, 119
Bullard, Sophie J., 176
Bumgardner, Joel D., 137
Jablonski, Paul D., 176

Jacobs, Joshua J., 262
Jenusaitis, Matthew, 61
L
Li, Feng, 222
Lin, Hsin-Yi, 137
MacDougall, Julie E., 209
C
Covino, Bernard S., Jr., 176
Cowen, Stephen, 3
Craig, Charles H., 28, 61,176
Cramer, Stephen D., 176
Marsh, J. Lawrence, 262

Mikecz, Katalin, 262
Morlders, Martina, 119
Moraux, Jean-Yves, 13
Morscher, Erwin W., 222, 249
de Andrade, Arnaldo H. P., 168

Dennis, J. Zachary, 61
Dichtel, Jean-Paul, 13
Disegi, John A., 50
Pannek, Edward J., Jr., 61

Parker, Suzanne H., 137
Perot, Nicolas, 13
Edwards, Michael R., 28, 61,176

Elke, Reinhard, 249
Eschbach, Lukas, 93
Radisch, Herbert R., Jr., 28, 61,176

Roach, Michael, 107, 154, 194
Fischer, Alfons, 119

Frick, Willi, 222
Friend, Clifford M., 28, 61, t76
Schenk, Rolf, 211

Skipor, Anastasia K., 262
Sokolowski, Alexandre, 168
Steger, Rainer, 39, 72
Galante, Jorge O., 262

Gasser, Beat, 93
Gilbert, Jeremy L., 262
Giant, Tibor T., 262
Gokcen, Nev A., 28, 61,176
Grpfert, Beat, 222
Govier, R. Dale, 28
Tikhovski, Ilia, 119

Trozera, Thomas A., 28
Turner, Paul C., 28, 61,176
U
Urban, Robert M., 262
273
Copyright 9 2003 by ASTM lntcrnational
www.astm.org
274
V
Vesely, Edward J., Jr., 28
Villamil, Ruth F. V., 168
W
Wiemann, Martin, 119

Williamson, Scott, 107, 154, 194
Windier, Markus, 39, 72, 211,235
Winters, Gary L., 72
Wirz, Dieter, 222

Woods, Terry O., 82
Z
Zardiackas, Lyle D., 50, 107, 137, 154,
194
Zhu, Mengkr 235
Ziomek-Moroz, Margaret, 176
Zurfluh, Bruno, 222
STP1438-EB/Jan.2003
Subject Index
intergranular, 176
macrophage cells and, 137
pitting, 3, 176
properties, 39, 72
resistance, 13, 119, 176
wear-induced, 211
Crevice corrosion, 176, 222, 262
Cutting, 13
Cyclic fatigue, 119
Cytokines, 262
Cytotoxicity, 119
316L stainless steel, 50, 194, 211

macrophage cells and corrosion, 137
A
AISI 630, 13
Alloy, 176
Anodic polarization, 107
Arthroplasty, 222
ASTM A 262E, 176
ASTM F 138, 168
ASTM F 746, 176
ASTM F 1314, 3
ASTM F 1586, 3, 39, 168
ASTM F 1801, 194
ASTM F 1875, 235
ASTM F 2129, 176
ASTM F 2180, 50
ASTM G 5, 107
ASTM G 71,107
Austenitic stainless steel, 119, 168, 194
Dilation-balloon expandable coronary

stents, 176
E
Electrochemical studies, 168

Electron microscopy, 222
Elongation, 39
F
Fatigue, 119, 194

resistance, 93
testing, 154, 194
Femoral head, 235
Femoral stem, 39, 249
Foreign-body reaction, 222
Forged hip stems, 39
Fretting corrosion, 222, 235, 262
Bending properties, 93
Biocompatibility, 119, 137
BioDur 108, 107, 154, 194
BioDur 316LS, 176
Biomechanics, 222
Bone plate, fatigue, 93
Bone resorption, 222
Bone screw, 93
Breakdown potentials, 168, 176
Galvanic corrosion, 107
Cable systems, 50
Cell culture, 137
Cement fixation, 249
Chromium, 262
CoCr alloy, 21 I, 235
Compatibility, MRI, 82
Coronary stent, 28, 61,176
Corrosion, 107, 176
crevice, 176, 222, 262
fatigue, 154, 194
fretting, 222, 235, 262
galvanic, 107
Hardness, 13, 39
High cycle fatigue tests, 93
High-nitrogen stainless steel, 39, 72, 119,
211,235
Hip prosthesis, 222
Hip stems, 39, 235
I
Implant materials, 3, 194, 211

Implant surface, 249
275
276
Inclusions, 72
Intergranular corrosion, 176
Intergranular crack initiation, 93
ISO 5832-9, 39
Radiopacity, 28, 176

Repassivation, 211
Ringers solution, 176
L
Low-nickel stainless steel, 93, 107
notch sensitivity and stress
corrosion, 154
M
Macrophage cells, stainless steel

corrosion and, 137
Macrophages, 222
Magnetic resonance imaging, 82
Martensitic steel, 13
Mechanical properties, 3, 39, 72
Medical instruments, 13
Metal ions, 262
Metals, 222
Microstructure, 3, 39, 50, 72
Modularity, 262
N
New PHACOMP, 28
Nickel-free stainless steel, 119
Nitrogen alloyed steel, 13
Nonmetallic inclusions, 3, 50
Notch sensitivity, 154
O
Safety, MRI, 82
Sandvik Bioline High-N, 3
Sendzimir mill, 61
Serum proteins, 262
Sliding wear, 211
Stem-cement interface, wear, 222
Stress corrosion cracking, 154
Surface analysis, 137
Surgical implants, 50
Systemic dissemination, 262
T
Tapered stem, 249
Taper interface, 235
Tensile strength, 39
Titanium alloy, 211,235
Total hip replacement, 222, 249
U
UNS $31683 alloys, 28, 61,176
V
Vacuum arc remelting, 61
Vacuum induction melting, 61
Vickers hardness, 39
Orthopedic implants, 168

Osteolysis, 262
P
P2000, 119
Particle size, 222
Particulate debris, 222, 262
Pitting corrosion, 3, 176
Platinum, 176
Platinum-enhanced radiopaque stainless
steel, 28, 61, 176
Polymethylmethacrylate, 222
Potentiodynamic polarization, 176
W
Wear, mechanism, 222
Wear-induced corrosion, 211
Wear resistance, 13
Wire, 50
X
XM 16, 13
X-rays, 176
Y
Yield strength, 39

0803134592

Uploaded by

Copyright:

Available Formats

0803134592

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

0803134592

Uploaded by

Copyright:

Available Formats

STP 1438

Stainless Steels for Medical and

Gary L. Winters and Michael J. Nutt, editors

ASTM Stock Number: STP1438

Library of Congress Cataloging-in-Publication Data

Peer Review Policy

Printed in (to come)

Characterization of Sandvik Bioline High-N--A Comparison of Standard Grades

Mechanical and Corrosion Properties of Forged Hip Stems Made of High-Nitrogen

PROCESSING, QUALITY ASSURANCE CONCERNS, AND M R I TESTING

Processing Platinum Enhanced Radiopaque Stainless Steel (PERSS | for Use as

MRI Safety and Compatibility of Implants and Medical Devices---TERRYo. WOODS

Comparison of Anodic Polarization and Galvanic Corrosion of a Low-Nickel Stainless

Fatigue Behavior and In-Vitro Biocompatibility of the Ni-Free Austenitie

Influence of Maerophage Cells on 316L Stainless Steel Corrosion--SUZANNErl. PARKER,

Comparison of Notch Sensitivity and Stress Corrosion Cracking of a Low-Nickel

Comparative Electrochemical Studies of F 1586-95 and F 138-92 Stainless Steels in

Corrosion Behavior of Platinum-Enhanced Radiopaque Stainless Steel (PERSS| for

Comparison of Corrosion Fatigue of BioDur| 108 to 316L S.S.

WEAR AND CORROSION-RELATED WEAR

Results of In-Vitro Studies about the Mechanism of Wear in the Stem-Cement

Corrosion Products Generated from Mechanically Assisted Crevice Corrosion of

STAINLESS STEELS FOR MEDICAL AND SURGICAL APPLICATIONS

Alloy Design, Mechanical Properties

Christina Haraldsson 1 and Stephen Cowen 2

Characterization of Sandvik Bioline High-N - A Comparison of Standard Grades

F1314 and F1586

Medical and SurgicalApplications, ASTMSTP 1438, G. L. Winters and M. J. Nutt, Eds.,

Keywords: Stainless steel, implant materials, nonmetallic inclusions, pitting corrosion

1 Research Engineer, Product Development, Metallurgical Research, AB Sandvik Steel,

Copyright*2003 by ASTM International

STAINLESSSTEELS FOR MEDICAL AND SURGICAL APPLICATIONS

Nitrogen-strengthened austenitic stainless steels conforming to ASTM F1314 and

HARALDSSON AND COWEN ON SANDVIK BIOLINE HIGH-N

Table 1 - Chemical compositions (weight-%)

STAINLESSSTEELS FOR MEDICALAND SURGICAL APPLICATIONS

Microstructural investigations were performed in a Leica DMR light optical

Table 2 - Maximum inclusion ratings according to ASTM E45-A

A comparative measurement has been performed using the Swedish standard

HARALDSSON AND COWEN ON SANDVIK BIOLINE HIGH-N

The pitting corrosion resistance has b ~ tested in accordance with Method A in

At the Royal Institute of Technology an electrochemical study of the resistance to

Table 3 - Comp,.xs'itions o f the PBS and simulated crevice solutions

Simulated crevice solution

STAINLESS STEELS FOR MEDICAL AND SURGICAL APPLICATIONS

Figure 3 - Microstructure of F1314material

Table 4 - Inclusion rating according to ASTM E45 method A, plate IlL

to the investigated F1314-material. The calculated area-% of inclusions is an average of

STAINLESSSTEELS FOR MEDICAL AND SURGICAL APPLICATIONS

Simulated crevice solution

Table 8 - Mechanical requirements according to material ,specifications

HARALDSSON AND COWEN ON SANDVIK BIOLINE HIGH-N

~ " i ' ' m

STAINLESS STEELS FOR MEDICAL AND SURGICAL APPLICATIONS

Nicolas Perot, t Jean-Yves Moraux, 1Jean-Paul Dichtel, 2 and Bruno Boucher 3

Keywords: medical instruments, hardness, corrosion resistance, cutting, wear resistance,

STAINLESSSTEELS FOR MEDICAL AND SURGICAL APPLICATIONS

Some Examples of Existing Grades (Table 1)

Steel Type AIS1420B: 13% chromium and O.3% carbon

Steel Type AIS1431:17% chromium, 2% nickel and O.16% carbon