Professional Documents
Culture Documents
2021 Compar son of ISO 14243-1 to ASTM F3141 n terms of wear ng of knee prostheses - Sc enceD rect
Download
Clinical Biomechanics
Volume 63, March 2019, Pages 34-40
Show more
Highlights
• This study compared differences between ASTM F3141 and ISO 14243-1.
• ASTM F3141 produced knee kinematics that were more similar to human gait.
• Simulations according to ASTM F3141 and ISO 14243-1 produce similar volumetric
wear rates.
• The wear region from the ASTM F3141 was smaller but deeper than ISO 14243-1.
Abstract
Background
The wear properties of knee implants need to be thoroughly evaluated prior to clinical use to
Download
ensure implant longevity. ISO 14243-1:2009 and ASTM F3141-17 are the two standards typically
used for evaluating wear, with the ISO standard being more common; ASTM F3141-17 was first
released in 2015. The aim of this study is to compare differences between these two standards in
terms of wearing on a knee prosthesis.
Methods
Using finite element analysis based on Archard's law, this study evaluated anterior-posterior and
internal-external motion, contact area, contact force, contact stress, volumetric wear rate, wear
depth, and wear distribution on the knee prosthesis.
Findings
The results show that simulations performed according to ASTM F3141 produced knee
kinematics that were more similar to human gait. The maximum wear depth occurred on the
medial side of the tibia. However, the region of peak contact stress did not always correspond
with the region of the maximum wear depth, indicating that considering the contact stress alone
is not sufficient for evaluating wear as the sliding distance also plays an important role. The
resulting wear region from the ASTM F3141 simulation was smaller but deeper than the wear
region from the simulation per ISO 14243-1. However, the volumetric wear rates were very
similar, with 13.48–55.26 mm3/million for ASTM F3141 and 13.64–54.9 mm3/million for ISO
14243-1.
Interpretation
The resulting rate of wear is almost identical between ISO 14243-1 and ASTM F3141. However,
there are differences in wear contours and wear depth.
Previous Next
Keywords
ISO 14243-1; ASTM F3141; Tibial wear; Finite element analysis; Knee kinematics
1. Introduction
Total knee arthroplasty (TKA) is a common surgical treatment for degenerated knees to alleviate
patient pain and increase knee function. Polyethylene wear and wear-related complications, such
as implant loosening, are still a leading cause of revision (Harman et al., 2010; Liang et al., 2018;
https://www.sc enced rect.com/sc ence/art cle/p /S0268003318307605?casa_token=AgKa1DuEu wAAAAA:nEw4lCC8le4T_hcn46c7_9FjcdxzxMrzSg… 2/19
12.04.2021 Compar son of ISO 14243-1 to ASTM F3141 n terms of wear ng of knee prostheses - Sc enceD rect
Pinol et al., 2014; Sharkey et al., 2014). Implant retrieval studies are the most effective way of
Download
directly analyzing wear on the knee joint, but such studies require samples from suitable patients
with a detailed assessment of individual physiological and pathological conditions. In addition, it
is often time consuming and costly (Gebert de Uhlenbrock et al., 2012; Huang et al., 1999; Huang
et al., 2002; Lu et al., 2010).
Wear simulators are often used to evaluate the wear properties of knee implants prior to clinical
use, with ISO 14243-1:2009 being the most commonly used standard governing the test methods
(Accutek, 2018; Endolab, 2018; FDA, 2003; ISO, 2009). ISO 14243-1 is also recommended by FDA
guidance documents for evaluating knee prostheses, with the requirement that if the thickness of
the tibial bearing (polyethelene) is below 6 mm, a wear test should be performed to demonstrate
that the tibial bearing can survive 10 million cycles under normal physiological loading (walking)
(FDA, 2003; ISO, 2009). Another option is to comply with ASTM F3141-17 which has similar input
rules for knee wear simulators under walking conditions.
A reliable standard for wear performance is crucial for the development of knee prostheses as it
predicts the lifetime of the tibial bearing. There are two control modes for wear testing during
simulated gait: displacement control according to ISO 14243-3, and load control according to ISO
14243-1 and ASTM F3141 (ASTM, 2017; ISO, 2002; ISO, 2009; ISO, 2014). The inputs for
displacement control are anterior-posterior (AP) displacement and internal-external (IE) angle,
and the inputs for load control are AP load and tibial rotation (TR) torque. This study focuses on
load control according to ISO 14243-1 and ASTM F3141, as there is only a load control mode
defined in the ASTM standard. Given its longer history, ISO 14243-1 is more common and
established in industry than ASTM F3141. ASTM F3141 was first released in 2015 and then
revised in 2017. ISO 14243-1 is still recommended for use by FDA guidance documents as well as
a number of multinational research organizations (Accutek, 2018; Endolab, 2018; FDA, 2003).
ASTM F3141 is based on studies by Bergmann et al. (Bergmann et al., 2014; Bergmann and
Graichen, 2014), while ISO 14243-1 is based on decades of data (Bergmann et al., 2014). However,
the differences between ISO 14243-1 and ASTM F3141 are still unclear in terms of wearing of
knee prostheses. Comparing the current revision of ISO 14243-1:2009 with the previous 2002
revision, the definition of soft tissue constraints has changed, but the definition of input curves
remains the same (ISO, 2002; ISO, 2009). The definition of input curves in ASTM F3141 has not
changed from its initial publication through to its revision in 2017 (ASTM, 2017).
There are four inputs to the knee simulator required by ISO 14243-1: 2009 and ASTM F3141-17;
flexion, axial load, internal-external (IE) torque and anterior-posterior (AP) load. The flexi
flexi flexion, axial
load and IE torque are similar between the two standards, but the definition of AP load is
different. For illustration, variations in these four inputs against gait cycles are plotted in Fig. 1
for both standards.
Download
Fig. 1. Four inputs from ISO 14243-1: 2009 and ASTM F3141:2017. (A) Flexi
Flexion angle; (B) axial load;
(C) internal-external (IE) torque; (D) anterior-posterior (AP) load.
In a healthy knee joint, when the knee flexes between 5° and 60°, the tibia moves anteriorly. This
is due to the fact that knee flexion
flexi and femoral rollback occur simultaneously (Bergmann et al.,
2014; Dimitriou et al., 2016; Iwaki et al., 2000; Reinschmidt et al., 1997). As seen in Fig. 1A, knee
flexion noticeably increases in the range of 40% to 70% of the gait cycle. ASTM F3141 correctly
flexi
gives the AP load for the tibia in the anterior direction while ISO 14243-1 states a posterior
direction.
To the best of our knowledge, no studies have been conducted so far to evaluate differences
between ISO 14243-1 and ASTM F3141 in terms of wearing of knee prostheses. The aim of this
study is to quantitatively compare differences between ISO 14243-1 and ASTM F3141 in terms of
wearing of knee prostheses. The hyposthesis is that there are differences in wear rates because of
different input conditions, in particular the AP load as shown in Fig. 1.
2. Methods
Download
Finite element analysis (FEA) models of the femoral component and tibial insert were developed
in Abaqus® 2017 (Abaqus, SIMULIA, East Providence, USA) using 3D scans of a PFC implant (Size
two, PFC, Depuy Synthes, Salt Lake, Utah, USA) (Fig. 2, Fig. 3). Geometrical differences in the
articular surface between the model and physical implant were controlled to be <0.1 mm. The
tibial insert was modeled as a deformable polyethylene (Gur 1050) material. The contact region of
the insert was meshed using hexahedron 8-noded elements C3D8R, and tetrahedron 10-noded
elements C3D10M were used for all other regions (Fig. 3). Based on convergence testing (<2%
difference) of maximum contact stress, an element size of around 1 mm was chosen for the
contact region and 1.6 mm was chosen for all other regions. The femoral component was
modeled as a rigid body to save computational time. The coefficient of friction between the tibia
and femur was 0.04 (Grupp et al., 2009).
Fig. 2. Geometrical differences on the articular surface between the model and physical implant
were controlled to be <0.1 mm.
Download
Fig. 3. FEA model; (A) AP and IE constraints, (B) wear region of tibial insert.
The boundary conditions were set according to the ISO and ASTM standards (ISO, 2009; ASTM,
2017). The axial loading axis was offset to the medial side by a distance 0.07 times the width of
tibia. Three of the loading curves (axial load, AP load and IE torque, Fig. 1B, C & D) were applied
to the tibia and the flexi
flexion angle curve (Fig. 1A) was applied to the femur. The tibia insert was
constrained in flexi
flexion but free to move in all other directions; medial-lateral, AP, superior-
inferior, valgus varus and rotation. The femoral component was only permitted to move in
flexion during the gait simulation. The flexi
flexi flexion axis was defined as a line connecting the lateral
femoral center and medial femoral center (ISO, 2009).
Nonlinear constraints were applied to simulate the integrative effect of all ligaments, tendons and
muscles according to ISO 14243-1 in the AP and IE directions (ISO, 2009). In the AP direction, the
soft tissue restraint stiffness was set as 9.3 N/mm if displacement exceeded 2.5 mm, but was zero if
<2.5 mm. In IE rotation, the restraint stiffness was set as 130 Nmm/° if angular rotation exceeded
6° (including internal and external rotation), but was zero if within 6° (including internal and
external rotation).
The restraints identified above were validated. As shown in Fig. 4, the two curves in each plot
reached their peak values simultaneously at approximately 58% of the gait cycle; refer to Table 1
for peak values recorded. The stiffness could then be calculated from these peak values:
stiffness = force / (displacement - 2.5 mm), or stiffness = torque / (rotation angle - 6°). The results
corresponded with the theoretical values (9.3 N/mm or 130 Nmm/°).
Download
AP constraint IE constraint
Max AP displacement (mm) Max AP load (N) Max rotation (°) Max torque (Nmm)
The tibial insert was evaluated in terms of wear depth, volumetric wear rate and volumetric wear.
The wear depth was estimated by Archard's wear law (Archard, 1953):
(1)
where H is wear depth (mm), K is wear coefficient (mm3/Nm), P is the contact pressure, and S is
the sliding distance of the tibia relative to the femur (mm). Values for K from literature vary from
2.64 ∗ 10−7 mm3/Nm to 10.656 ∗ 10−7 mm3/Nm (Barnett et al., 2002; Ding et al., 2018; Knight et al.,
2007; Wang et al., 2017; Willing and Kim, 2009). This study presents all wear results based on both
2.64 ∗ 10−7 mm3/Nm and 10.656 ∗ 10−7 mm3/Nm.
Using the user-defined subroutine UMESHMOTION in Abaqus 2017, the wear depth at each
node on the articular surface could be calculated using Archard's law. According to the calculated
wear depth, each node on the surface was then moved in the direction normal to the articular
https://www.sc enced rect.com/sc ence/art cle/p /S0268003318307605?casa_token=AgKa1DuEu wAAAAA:nEw4lCC8le4T_hcn46c7_9FjcdxzxMrzSg… 7/19
12.04.2021 Compar son of ISO 14243-1 to ASTM F3141 n terms of wear ng of knee prostheses - Sc enceD rect
surface. An adaptive remeshing procedure was employed to simulate the progression of surface
Download
wear and was calculated for 5 million cycles according to ISO 14243-1. The mesh on the surface of
the tibial insert was updated every 500,000 cycles (1 step), which has been shown to produce
results with a difference of only 2.75% to 4.8% with a step size of 125,000 cycles (Knight et al.,
2007; Wang et al., 2017).
The wear rate was calculated as the average wear volume per million cycles (mm3/million cycles).
This approach was developed and validated in previous studies on TKA in our laboratory (Ding et
al., 2018; Wang et al., 2017).
All finite element models and analyses were setup and performed in accordance with the journal
requirements for finite element studies (Viceconti et al., 2005).
3. Results
Results are presented for AP translation, IE rotation, contact area, contact force, contact stress,
volumetric wear rate, wear depth and wear distribution.
Download
Fig. 6. Contact areas and contact forces from ASTM and ISO standards. (A) Contact areas; (B)
contact forces.
Download
Fig. 8. The respective locations of the peaks of maximum contact stress: A, B, C, and D.
Download
Table 2. Predicted wear rate, volumetric wear and wear depth for ASTM and ISO standards.
Fig. 9 shows the wear depth on the tibial insert for both standards. The wear region is smaller but
deeper for the ASTM standard, while the region is more dispersed for the ISO model. The
maximum depth is present on the medial side of the insert for both models.
Download
Fig. 9. Wear contours.
4. Discussion
The most important finding of this study is that whether ASTM F3141 or ISO 14243-1 is used for
simulating wear on the tibial insert, the resulting volumetric rate of wear is almost identical.
However, there are differences in wear contours and wear depth between the two standards.
The wear region was smaller but deeper for the ASTM model, but the two standards had an
almost identical volumetric wear rate. This can be attributed to several causes. Firstly, for the ISO
14243-1 model, the AP load acted on both the posterior and anterior region of the tibia (Fig. 1D),
which resulted in not only femoral rollback but also anterior femoral translation, which is
opposite to normal knee motion. This caused a more dispersed wear region for the ISO model
(Fig. 5). Secondly, the IE torque and mean axial load from the ASTM standard are greater (Fig. 1B
and C), which resulted in more IE rotation and greater contact stresses (Fig. 5, Fig. 7). Therefore,
the wear depth was deeper in the ASTM model. Thirdly, the wear rate for a volume is calculated
by integrating the wear area and wear depth per million cycles. Therefore, almost equivalent
volumetric wear rates were calculated for both standards.
The accuracy of the FEA models in this study was confirmed by the following aspects: (i) The wear
region from the ISO 14243-1 model was similar to that reported by Morra and Postak (1999), in
which shallow wearing was found in the anterior region but deeper wearing in the posterior
region. (ii) The calculated volumetric wear rates in this study (13.64 mm3/million to
54.9 mm3/million from ISO 14243-1 with different K values) are comparable to in vitro
experimental studies by Barnett et al. (18.1 mm3/million on average with a standard deviation of
3 mm3/million, and 41 mm3/million on average with a standard deviation of 14 mm3/million)
(Barnett et al., 2001; Barnett et al., 2002) and McEwen et al. (22.12 mm3/million on average with a
standard deviation of 6.02 mm3/million) (McEwen et al., 2001). (iii) The wear algorithm and
calculations were confirmed by experimental results published in previous research from our
laboratory (Ding et al., 2018; Wang et al., 2017). (iv) The FEA method and model was validated
according to the experimental results (Supporting file).
From the results presented in Fig. 5, ASTM F3141 can be considered to more closely represent
normal gait. During knee flexion,
flexi the tibia rotates internally and the femur rolls back (Bergmann
et al., 2014; Dimitriou et al., 2016; Iwaki et al., 2000; Reinschmidt et al., 1997).
Variations in the contact area, total contact forces, and contact stress (Fig. 6, Fig. 7) show a similar
pattern to the axial load inputs from the ASTM and ISO standards in Fig. 1B, and the locations of
maximum depth can be seen on the medial side of the tibial insert. These results could be
anticipated as the axial load dominates over other loading forces, such AP or ML loads. As a
Download
consequence, the total contact force appears to be primarily influenced by the axial load. The
loading points on the tibia and femur were offset by a distance 0.07 times the width of tibia to the
medial side according to ISO 14243-1, which resulted in the medial side bearing a greater load
(ISO, 2009).
The locations of peak contact stresses (Fig. 8A, B, C, D) did not necessary correspond with the
location of maximum wear depth (Fig. 9). This indicates that considering contact stress alone is
not sufficient for evaluating wear as the sliding distance also plays an important role for
estimating the wear depth and volume (Liau et al., 2002; Simpson et al., 2008).
There are some limitations to this study that should be noted. Firstly, Archard's law does not
account for pitting, delamination and third body wear modes and has a limited ability for
predicting abrasive/adhesive wear (Pal et al., 2008). Secondly, neither the effect of creep nor cross-
shear was considered in this study. Thirdly, the geometrical models and material properties were
not sourced directly from the manufacturer, but the accuracy of the model in this study was
confirmed to be within 0.1 mm of the physical implant. Despite these limitations, the current
study provides insights into the differences between ISO 14243-1 and ASTM F3141 and the
variability in wear calculations.
5. Conclusion
The resulting wear region is smaller but deeper when simulated according to ASTM F3141.
However, the calculated volumetric wear rates were very similar for both standards. The results
show that knee kinematics simulated according to ASTM F3141 are also more similar to human
gait. For both ISO 14243-1 and ASTM F3141, the trends of contact area, total contact forces, and
contact stress curves were similar to the axial load inputs. The location of maximum depth
occurred on the medial side of the tibia for both standards, and the region of peak contact stress
did not always correspond with the region of the maximum wear depth.
Declarations of interest
None.
Acknowledgement
This study was supported by a Grant from the Chinese National Science and Technology
Program (National Key Research and Development Plan) (Grant number: 210YBXM2016110002),
and a grant from National Natural Science Foundation of China (Grant number: 81572180). The
authors would like to thank Mr. Colin McClean for his assistance with proofreading this
manuscript, and Prof. Chuan Li at the National Yang Ming University for his support and advice
through numerous communications.
https://www.sc enced rect.com/sc ence/art cle/p /S0268003318307605?casa_token=AgKa1DuEu wAAAAA:nEw4lCC8le4T_hcn46c7_9FjcdxzxMrzS… 13/19
12.04.2021 Compar son of ISO 14243-1 to ASTM F3141 n terms of wear ng of knee prostheses - Sc enceD rect
References
Accutek testing laboratory, 2018 Accutek testing laboratory
https://www.element.com/more-sectors/medical-device/wear-testing-of-medical-devices
(2018)
Barnett et al., 2001 P.I. Barnett, J. Fisher, D.D. Auger, M.H. Stone, E. Ingham
Comparison of wear in a total knee replacement under different kinematic conditions
J. Mater. Sci. Mater. Med., 12 (10−12) (2001 Oct–Dec), pp. 1039-1042
View Record in Scopus Google Scholar
Barnett et al., 2002 P.I. Barnett, H.M. McEwen, D.D. Auger, M.H. Stone, E. Ingham, J. Fisher
Investigation of wear of knee prostheses in a new displacement/force-controlled simulator
Proc. Inst. Mech. Eng. H, 216 (1) (2002), pp. 51-61
View Record in Scopus Google Scholar
Dimitriou et al., 2016 D. Dimitriou, T.Y. Tsai, K.K. Park, A. Hosseini, Y.M. Kwon, H.E. Rubash, G.
Li
Weight-bearing condyle motion of the knee before and after cruciate-retaining TKA: in-
vivo surgical transepicondylar axis and geometric center axis analyses
J. Biomech., 49 (9) (2016), pp. 1891-1898
Article Download PDF View Record in Scopus Google Scholar
Ding et al., 2018 W.Y. Ding, S.Q. Ma, X.C. Zhou, Y.J. Yang, Y. Sun, F. Zhao, Y.B. Fan
Wear simulation of tibiofemoral joint surface of total knee prosthesis with high
conformity
J. Med. Biomechanics, 33 (3) (2018), pp. 7-13
CrossRef View Record in Scopus Google Scholar
Grupp et al., 2009 T.M. Grupp, C. Kaddick, J. Schwiesau, A. Maas, S.D. Stulberg
Fixed and mobile bearing total knee arthroplasty—influence on wear generation,
corresponding wear areas, knee kinematics and particle composition
Clin. Biomech., 24 (2) (2009), pp. 210-217
Article Download PDF View Record in Scopus Google Scholar
M.K. Harman, S. Schmitt, S. Rössing, S.A. Banks, H.P. Sharf, M. Viceconti, W.A. Hodge
Download
Polyethylene damage and deformation on fixed-bearing, non-conforming unicondylar
knee replacements corresponding to progressive changes in alignment and fixation
Clin. Biomech., 25 (6) (2010), pp. 570-575
Article Download PDF View Record in Scopus Google Scholar
Huang et al., 2002 C.H. Huang, F.Y. Ho, H.M. Ma, C.T. Yang, J.J. Liau, H.C. Kao, T.H. Young, C.K.
Cheng
Particle size and morphology of UHMWPE wear debris in failed total knee arthroplasties
—a comparison between mobile bearing and fixed bearing knees
J. Orthop. Res., 20 (5) (2002), pp. 1038-1041
Article Download PDF View Record in Scopus Google Scholar
Tibiofemoral movement 1: the shapes and relative movements of the femur and tibia in
Download
the unloaded cadaver knee
J. Bone Joint Surg. (Br.), 82 (8) (2000), pp. 1189-1195
View Record in Scopus Google Scholar
Knight et al., 2007 L.A. Knight, S. Pal, J.C. Coleman, F. Bronson, H. Haider, D.L. Levine, M.
Taylor, P.J. Rullkoetter
Comparison of long-term numerical and experimental total knee replacement wear
during simulated gait loading
J. Biomech., 40 (7) (2007), pp. 140-146
Google Scholar
Liang et al., 2018 H. Liang, J.K. Bae, C.H. Park, K.I. Kim, D.K. Bae, S.J. Song
Comparison of mode of failure between primary and revision total knee arthroplasties
Orthop. Traumatol. Surg. Res., 104 (2) (2018), pp. 171-176
Article Download PDF View Record in Scopus Google Scholar
Liau et al., 2002 J.J. Liau, C.K. Cheng, C.H. Huang, W.H. Lo
The effect of malalignment on stresses in polyethylene component of total knee
prostheses—a finite element analysis
Clin. Biomech., 17 (2) (2002), pp. 140-146
Article Download PDF View Record in Scopus Google Scholar
Lu et al., 2010 Y.C. Lu, C.H. Huang, T.K. Chang, F.Y. Ho, C.K. Cheng, C.H. Huang
Wear-pattern analysis in retrieved tibial inserts of mobile-bearing and fixed-bearing total
knee prostheses
J. Bone Joint Surg. (Br.), 92 (4) (2010), pp. 500-507
View Record in Scopus Google Scholar
McEwen et al., 2001 H.M. McEwen, J. Fisher, A.A. Goldsmith, D.D. Auger, C. Hardaker, M.H.
Stone
Wear of fixed bearing and rotating platform mobile bearing knees subjected to high levels
of internal and external tibial rotation
J. Mater. Sci. Mater. Med., 12 (10–12) (2001), pp. 1049-1052
View Record in Scopus Google Scholar
Google Scholar
Download
Pal et al., 2008 S. Pal, H. Haider, P.J. Laz, L.A. Knight, P.J. Rullkoetter
Probabilistic computational modeling of total knee replacement wear
Wear, 264 (2008), pp. 701-707
Article Download PDF View Record in Scopus Google Scholar
Reinschmidt et al., 1997 C. Reinschmidt, A.J.V.D. Bogert, A. Lundberg, B.M. Nigg, N. Murphy
Tibiofemoral and tibiocalcaneal motion during walking: external vs skeletalmarkers
Gait Posture, 6 (2) (1997), pp. 98-109
Article Download PDF View Record in Scopus Google Scholar
Sharkey et al., 2014 P.F. Sharkey, P.M. Lichstein, C. Shen, A.T. Tokarski, J. Parvizi
Why are total knee arthroplasties failing today—has anything changed after 10 years?
J. Arthroplast., 29 (9) (2014), pp. 1774-1778
Article Download PDF View Record in Scopus Google Scholar
Simpson et al., 2008 D.J. Simpson, H. Gray, D. D'Lima, D.W. Murray, H.S. Gill
The effect of bearing congruency, thickness and alignment on the stresses in
unicompartmental knee replacements
Clin. Biomech., 23 (9) (2008), pp. 1148-1157
Article Download PDF View Record in Scopus Google Scholar
Wang et al., 2017 C. Wang, F. Zhao, W.Y. Ding, Y. Sun, Y.B. Fan
Finite element study on total knee prosthesis wear during stair ascent
J. Med. Biomechanics, 32 (2) (2017), pp. 109-114
View Record in Scopus Google Scholar
View Abstract
Download
About ScienceDirect
Remote access
Shopping cart
Advertise
Privacy policy
We use cookies to help provide and enhance our service and tailor content and ads. By continuing you agree to the use of cookies.
Copyright © 2021 Elsevier B.V. or its licensors or contributors. ScienceDirect ® is a registered trademark of Elsevier B.V.
ScienceDirect ® is a registered trademark of Elsevier B.V.