Journal Information
Vol. 8. Issue 5.
Pages 4642-4650 (September - October 2019)
Download PDF
More article options
Vol. 8. Issue 5.
Pages 4642-4650 (September - October 2019)
Original Article
DOI: 10.1016/j.jmrt.2019.08.008
Open Access
Investigation of wear of ultra high molecular weight polyethylene in a soft tissue behaviour knee joint prosthesis wear test simulator
Erkan Bahçe
Corresponding author

Corresponding author.
, Ender Emir
Inonu University, Department of Mechanical Engineering, Malatya 44280, Turkey
This item has received

Under a Creative Commons license
Article information
Full Text
Download PDF
Figures (14)
Show moreShow less
Tables (2)
Table 1. Literature review of UHMWPE wear study.
Table 2. Wear teat parameters.
Show moreShow less

Knee joint prostheses are commonly employed for therapy in over-use disorder, and as a result of traffic accident and sports injuries. However, errors that can occur due to their use need to be defined. In order to determine the types and rates of wear that might occur in prosthetic components, wear tests were performed in knee joint simulators that mimiced knee joint motion. In this study, the wear was specifically examined in ultra high molecular weight polyethylene (UHMWPE) insert material in the four-axis knee joint prosthesis wear test simulator, and the causes for it were examined. Wear tests were performed for 3 different cycles, 1 × 106, 2 × 106 and 3 × 106. Following wear tests, microscopic images, and mass loss and surface roughness measurements were taken from the UHMWPE insert condyle surface. With an increase in the number of cycles, pitting wear, scratches, agglomerated particles and delamination were seen more clearly on the UHMWPE material surface, due to repetitive forces on different axes. In addition, surface roughness measurements taken from worn surfaces increased the with the number of cycles, and flexion/extension (F/E) motion range increased with surface roughness. Wear rates in the medial region of the UHMWPE insert were higher than in the lateral region.

Knee joint simulator
Wear resistance
Surface quality
Full Text

Total knee prostheses are made of metal and polyethylene materials, creating an artificial joint designed for painless joint movement. With newer implant designs and improved surgical techniques, total knee replacements can be expected to function well for at least 20 to 25 years [1,2]. However, due to repeated loads conditions of use and stresses on different axes, certain wear mechanisms impinge on the prosthetic components. When previous studies are examined, it is seen that the most wear occurred in the UHMWPE insert component [1]. These wear mechanisms on the UHMWPE insert component are pitting wear, delamination, scratches, agglomerated particles and mass losses [1,2]. The most important reason for wear is repetitive loads and the resulting stresses [1,2].

Dynamic in vitro tests are performed outside the body environment in order to determine wear mechanisms before surgery [1], attracting the attention of the orthopedics community. Wear test simulators have been used to detect errors in many knee joint prostheses designed to date and also determine the approximate life of the prosthesis. Previous studies have been carried out to detect wear on the UHMWPE insert material using different types of knee joint simulator (Table 1) [3–10]. The most important of the studies are those with simulators that reflect the effects of soft tissue structure involved in joint movement. However, in these tests movement at the knee joint is only applied in certain axes [9,10].

Table 1.

Literature review of UHMWPE wear study.

Author  Simulator motion  Cycle (million)  Soft tissue behaviour mechanism  Wear type  Wear rate (mg) 
Lancin et al. (2007)  F/E, A/P, I/E, A.F.  2.5  A/P Motion  Burnishing  32 
Flannarry et al. (2008)  A/P, F/E, A.F.  –  Scratches  5.4 
Schwenke et al. (2009)  F/E, I/E, A/P, M/L, V/V  I/E Motion  Strations  70 
Willing et al. (2009)  A/P, F/E, M/L, I/E, A/A  3.5  –  Mass loss  50 
Affato et al. (2011)  F/E, A/P, I/E, A.F.  –  Burnishing  4.5 
Jaber et al. (2015)  A/P, F/E, I/E, A.F.  –  Scratches  20.2 

F/E: Flexion/Extension, A/P: Anterior/Posterior, M/L: Medial/Lateral, I/E: Internal/External, V/V: Varus/Valgus, A/A: Adduction/Abduction, AF: Axial Force.

The similarity of movement of the device for measuring wear to that of the joint is important for the accuracy of the results. Therefore, the ability to observe the effects of muscles, soft tissues and ligaments surrounding the joint are important parameters in obtaining high accuracy data [11]. Previous studies have used springs for studying soft tissue behaviour [3,4]. However, the effects of soft tissue and tendons during movement have generally been ignored in simulators used in wear tests. As a result, reliable wear results have not been obtained.

In this study, the tibial rotation and axial force movements in the simulator are determined from the tibial component, while the F/E and anterior/posterior (A/P) motions are given from the femur component. Unlike previous studies, a knee joint wear test simulator, which exhibited soft tissue behaviour by using springs in axial force and A/P motions, was used. Wear tests were carried out for 3 different cycles, 1 × 106, 2 × 106 and 3 × 106. In this way, the types and rates of wear occurring in the polyethylene material were determined. In addition, previous studies were discussed and the effects of soft tissue behaviour on prosthetic wear were determined.

2Materials and methods2.1Knee joint biomechanics and wear simulator design

In the ISO 14243-3 test standard, 4 different axes and motion pairs were specified for the knee joint simulator. These were axial force, tibial rotation, anterior/posterior (A/P) motion and flexion/extension (F/E) motion [12]. The prosthetic components and directions of motion are given in Fig. 1. The F/E motion and A/P motion were given from the femur component, the tibial rotation and the axial force movement were given from the tibial component. In another phase of the design, engines were selected for the load and displacement of 4 different motion pairs. Servo motors (1.5 kW and 0.75 kW) were used to perform F/E and tibial rotation motions with high precision. For the axial force and A/P linear motions, linear actuators were used to convert the circular motion from the DC motor to linear motion (500 N and 2500 N).

Fig. 1.

Knee joint wear simulator.

2.2Data input for total knee joint prosthesis wear simulator

The study was set up to enable the motors to implement the required standardized ISO 14243-3 standard motion. Motion profiles determined to complete a cycle within 2 seconds are given in Fig. 2. The tibial rotation was performed at 0°–5.7° (Fig. 2a). The F/E motion value was applied at 0°–58° (Fig. 2b). These 2 motion profiles were obtained according to the displacement values of the knee joint in the gait cycle, based on the ISO 14243-3 standard.

Fig. 2.

Input data profile for; a) tibial rotation, b) F/E motion, c) A/P motion, d) axial force.


Linear actuators used for A/P and axial force motion were operated with position control. In the axial force motion profile, the maximum value was 2400 N (Fig. 2c). Another movement profile, A/P motion, was limited to a maximum of 5.2 mm. Since linear actuators did not work as precisely as servo motors, motion profiles were created taking into account the maximum values given in the ISO 14243-3 standard (Fig. 2c–d).

Control operation was performed for the motors used according to the motion profiles given in Fig. 2. Programmable logic control (PLC) was used for servo motors during the control phase. Limit switches and relays were used for the control of linear actuators. In this way, A/P and axial force motions were run between distances based on ISO 14243-3 test values. Finally, the number of cycles required for the test was determined by the number of steps a healthy person would take daily, and by taking into account when critical wear occurred. Therefore, tests were performed for 1 × 106, 2 × 106 and 3 × 106 cycles. A size 5 ligament protective knee joint prosthesis was used by 3 test specimens (Fig. 3). Tests were performed in an environment without body fluid. The wear test parameters are given in Table 2.

Fig. 3.

Size 5 ligament protect knee joint prosthesis.

Table 2.

Wear teat parameters.

Motion variations  Maximum test value 
Flexion (°)  58 
Anterior/posterior (mm) 
Tibial rotation (°)  5.2 
Axial force (N)  2500 
Cycle (million) 
2.3Measurement of surface defects, wear rate and surface roughness

It was important to determine the errors that might occur in the prostheses used for experimentation. For the knee joint prosthesis, the defects caused by wear were mostly in the UHMWPE component. The most common types of defects previously reported were pitting wear, oxidation, agglomerated particles, delamination and scratches of the UHMWPE material [2]. Therefore, measurements were designed to analyse these components, as described below.

Following testing, defects on the surface were examined both micro and macroscopically. A LEICA microscope and LEO-EVO 40 scanning electron microscope (SEM) were used to obtain images at different magnifications (50×, 100×, 250×, 500× and 1000×) from the medial and lateral regions. The images were taken for each of the tests performed at the 3 different cycles. The measurements were taken from the regions where the wear rates were highest, as described in the literature (Fig. 4).

Fig. 4.

Microscope imaging zone.


The amount of wear occurring in the UHMWPE component was measured according to the gravimetric method [13]. The mass losses of UHMWPE material of the 3 knee joint prostheses used during the test were measured in milligrams on precision scales. Surface roughness of the UHMWPE surface was measured with a MITUTOYO SJ-210 surface roughness tester, operated at a probe feed rate of 0.5 mm/s and cut off at 0.8 mm. The medial (M) and lateral (L) regions of the UHMWPE component were measured for surface roughness. In addition, due to the greater freedom and femoral roll-back of motion in the medial region than in the lateral region, 4 different flexion angles were measured (Fig. 5). In order to assess the reproducibility of the surface roughness values, 3 measurements were taken for both methods.

Fig. 5.

Surface roughness measurement degrees.

3Result and discussion

As different cycle numbers were found to have significantly different effects on surface defects and surface roughness of the UHMWPE component, these aspects were discussed separately, below.

3.1Surface defects and wear rate

As a result of abrasion tests performed for different cycle numbers, pitting wear, scratches and agglomerated particles were formed in UHMWPE insert condyles. While tests were performed with 3 different prostheses, microscopy images were taken from a single prosthesis. In addition, measurements of wear were taken from the same regions of the insert.

Pitting wear on the surface of the insert was distinctly increased by the number of cycles (Fig. 6). The formation of pitting wear was evaluated at 3 stages. First level pitting wear appeared as traces under the surface, as a result of 1 × 106 cycles (Fig. 6a). With an increase in the number of cycles, the wear marksbecame more distinct and second level pitting was found (Fig. 6b). At this stage, an increase in the stresses under the surface of the contact areas of the UHMWPE insert and the femur component were evident. In Fig. 6c, the wear marks showed growth due to the mechanical properties of the UHMWPE material, and the particles begin to break after a determined stress value. This produced the third level level of the wear, due to stresses in different axes on the surface or under the surface, resulting in continuously growing cracks [2,14,15].

Fig. 6.

Microscope images taken from UHMWPE material as a result of different cycle numbers a) 1 × 106 cycle, b) 2 × 106 cycle, c) 3 × 106 cycle.


SEM images showed that the same directional scratches were formed with A/P motion (Fig. 7a). Particles separated from polyethylene and metal materials as well as from bone cement, which were loosened by repeated load, resulted in scratch formation. Particulates formed distinct scratches due to both pressure and friction on the surface, with the component force generated by a combination of A/P and F/E motion between the UHMWPE and the femur components [16,17]. In addition, scratches were wider on the medial region compared to the lateral region. On average those in the lateral region measured 9.1 µm and those in the medial region 12.2 µm (Fig. 7b). This difference was due to both the load imbalance caused by the body mechanical axis and the greater mobility of the medial region compared to the lateral region during joint motion. In addition, significant scratches were observed in the femoral component condyles corresponding to the UHMWPE insert component condyles (Fig. 8). This indicated that the scratches in the insert conduits were caused by movement between the femur and the insert component. The other consequence of scratch formation was delamination. This was due to tangential growth of cracks on the UHMWPE insert condyle surfaces (Fig. 9).

Fig. 7.

UHMWPE SEM image; a) scratching, b) scratching width.

Fig. 8.

Femoral component SEM image scratching.

Fig. 9.

Delamination of UHMWPE insert codyle surface.


In the microscope images, significant wear marks were observed in the anterior areas of both condyle surfaces of the UHMWPE insert (Fig. 10). In cases where the F/E motion angle increased due to knee joint kinematics, the tibia joint moved in the anterior direction due to femoral roll-back motion. With this movement, the load increased in the anterior region due to the effect of the axial force. Due to the mechanical properties of the UHMWPE material as a result of repeated loads, deformations occurred on the condyle surfaces above a certain setback value. This resulted in wear marks in the anterior region of the UHMWPE insert component. Previous studies have also indicated that extensive wear occurred in the anterior areas of the lateral and medial condyles of the UHMWPE insert material [18,19].

Fig. 10.

UHMWPE material anterior region wear marks; a) medial, b) lateral.


Large amounts of agglomerated particles formed on the condyle surfaces of the UHMWPE material in the A/P motion direction (Fig. 11). This structure was the result of 2 or more particles interacting with each other [20]. The main reason for this structure, formed at the particle-sub-layer interface by a continuous increase in deformability due to use, could be explained by the aggregation of polymer matrices. This resulted in a non-uniform formation. Similar results have previously been observed in studies on UHMWPE wear [20–22].

Fig. 11.

Agglomerated particle on UHMWPE condyle surfaces.


As a result of the abrasions occurring in the UHMWPE material, particle breaks were observed in the condyle surfaces. Due to the fracture particles, mass loss occurred in the insert component. Minimum mass loss was 1.25 mg and the maximum mass loss was 3.91 mg for 3 different UHMWPE materials as a result of 3 different cycles (Fig. 12). In the first stage of wear, the loss of mass occurred due to the rapid increase in load and the resulting fatigue. In the second stage, the size of the pits in the wear areas and the loss in mass gradually increased. However, the rate of increase in mass loss was slower than the first stage. In the last stage, cracks in both the fatigue wear zones and the newly formed areas grew and mass loss increased. When the loss of mass exceeded a certain value, the knee joint prosthesis was unable to perform its prescribed function. These findings on the wear of UHMWPE material was consistent with previous studies [2,6]. The rate of mass loss was higher in simulators that did not take into account soft tissue behaviour [8].

Fig. 12.

Mass loss of the tested polyethylene under gait cycle activity.

3.2Surface roughness

Roughness measurements taken from the medial and lateral condyle surfaces of the UHMWPE component showed an increase in surface quality with an increase in number of cycles (Fig. 13). Maximum and minimum surface roughness were 3.688 µm and 2.392 µm, respectively. With an increase in the number of cycles, surface quality deterioration occurred as a result of gradually increasing pitting wear, scratches and agglomerated particles on the UHMWPE condyle surfaces.

Fig. 13.

Surface roughness of UHMWPE insert component condyle surface.


In the procedure specified in the ISO14243-3 standard, the insert component should be assembled with the femur approximately 0.07 times the width of the axis as the insert [13]. Due to this, more load impinged on the medial areas [16]; therefore, load had a greater effect on this region during motion. As a result, it was understood that the condyles in the femur component did not evenly distribute the load on the insert.

Another result from the roughness measurements was the increase in surface roughness due to an increase in F/E motion angle (Fig. 14). This was related with femoral roll-back motion as defined previously for knee joints [2,11]. Due to this movement, the frictional force caused by the increase in pressure applied to the insert surface by the femur component increased. As a result, particle breakage was increased due to the effect of contact pressure on the surface [23,24]. Concommitantly, surface quality deteriorated as a result of breakage particles.

Fig. 14.

Surface roughness change according to F/E degree.


In this study, a knee joint prosthesis wear test simulator was used to assess the wear performance of UHMWPE material in knee joint prostheses under periodic loads. In addition, the mechanisms of error in the UHMWPE material subjected to the knee joint simulator were examined. The experimental data obtained were as follows:

  • 1

    With an increase in the number of cycles, loss of mass resulting from wear of the UHMWPE material increased.

  • 2

    With an increase in the number of cycles, pitting wear on the surfaces of the UHMWPE material became more evident. In addition, wear rate in the anterior region was greater than in the posterior region.

  • 3

    Agglomerated particles were formed in the direction of A/P motion.

  • 4

    In the medial and lateral regions of the UHMWPE material, surface roughness increased with an increase in the number of cycles.

  • 5

    An increase in the F/E angle increased wear on the surface.

  • 6

    Surface roughness was greater in the medial region than in the lateral region.

  • 7

    Increased surface roughness was due to an increased F/E angle.

In summary, the results supported literature data. However, more accurate results have been obtained in terms of the ratio of hand and wear zones, wear types, numerical data and human walk approach. In future studies, it will be an important reference for the design of knee joint prosthesis in different geometries by testing the knee joint prostheses in different geometries and the results obtained from the simulator to improve wear zones and sizes as a result of tests.

Conflict of interest

The authors have no conflict of interest to declare.


This research was supported by OTIMED Medical Company and Inonu University BAP (project number: FYL-1194).

M. Flannery, T. McGloughlin, E. Jones, S. Birkinshaw.
Analysis of wear and friction of total knee replacements: Part I. Wear assessment on a three station wear simulator.
Wear, 265 (2008), pp. 999-1008
Junfen Shı.
Finite element analysis of total knee replacement considering gait cycle load and malalignment, Ph.D. Thesis.
Wolverhampton University England, (2007),
Antonino Romeo.
Design and development of a knee simulator device, PhD Thesis.
Politecnico Di Torino University Belgium, (2015),
Neil Campbell.
Design of knee simulator for the testing of total knee prosthesis, PhD Thesis.
Cape Town University South Africa, (2008),
T. Calliess, S. Schado, B.I. Richter, C. Becher, M. Ezechieli, S. Ostermeier.
Quadriceps force during knee extension in different replacement scenarios with a modular partial prosthesis.
Clin Biomech, 29 (2014), pp. 218-222
S. Abdel-Jaber, C. Belvedere, A. Leardini, S. Affatato.
Wear simulation of total knee prostheses using load and kinematics waveforms from stair climbing.
J Biomech, 48 (2015), pp. 3830-3836
A. Liu, L.M. Jennings, E. Ingham, J. Fisher.
Tribology studies of the natural knee using an animal model in a new whole joint natural knee simulator.
J Biomech, 48 (2015), pp. 3004-3011
P.S. Walker, M.T. Lowry, G. Yildirim, A. Kumar.
Kinematic differences in posterior stabilized total knees determined by a holistic experimental evaluation method.
J Med Devices, 7 (2013),
T. Schwenke, D. Orozco, E. Schneider, M.A. Wimmer.
Differences in wear between load and displacement control tested total knee replacements.
Wear, 267 (2009), pp. 757-762
P. Lancin, A. Essner, S.S. Yau, A. Wang.
Wear performance of 1900 direct compression molded, 1020 direct compression molded, and 1020 sheet compression molded UHMWPE under knee simulator testing.
Wear, 263 (2007), pp. 1030-1033
Loading and displacement parameters for wear-testing machines with displacemet control and corresponding environmental conditions for test, ISO 14243-3.
International Organization for Standardization, (2014),
Implants for surgery-wear of total knee joint prostheses, ISO 14243-2.
International Organization for Standardization, (2009),
A.V. Lombardi, B.S. Ellison, K.R. Berend.
Polyethylene wear is influenced by manufacturing technique in modular TKA.
Clin Orthopaed Relate Res, 466 (2008), pp. 2798-2805
T.M. Wright, S.B. Goodman.
What design factors influence wear behavior in total knee replacement?.
Implant wear in total joint replacement, pp. 224
J.J. Cherian, B.H. Kapadia, S. Banerjee, J.J. Jauregui, K. Issa, M.A. Mont.
Mechanical, anatomical, and kinematic axis in TKA: concepts and practical applications.
Curr Rev Musculoskeletal Med, 7 (2014), pp. 89-95
M. Flannery, T. McGloughlin, E. Jones, C. Birkinshaw.
Analysis of wear and friction of total knee replacements.
Wear, 265 (2008), pp. 999-1008
J.P. Kretzer, E. Jakubowitz, R. Sonntag, K. Hofmann, C. Heisel, M. Thomsen.
Effect of joint laxity on polyethylene wear in total knee replacement.
J Biomech, 43 (2010), pp. 1092-1096
A. Abdelgaied, F. Liu, C. Brockett, L. Jennings, J. Fisher, Z. Jin.
Computational wear prediction of artificial knee joints based on a new wear law and formulation.
J Biomech, 44 (2011), pp. 1108-1116
H.J. Cheng, S.S. Hsiau.
The study of granular agglomeration mechanism.
Powder Technol, 199 (2010), pp. 272-283
A. Liu, L. Richards, C.L. Bladen, E. Ingham, J. Fisher, J.L. Tipper.
The biological response to nanometre-sized polymer particles.
Acta Biomater, 23 (2015), pp. 38-51
K. Ravi, T. Deplancke, K. Ogawa, J.Y. Cavaille, O. Lame.
Understanding deposition mechanism in cold sprayed ultra high molecular weight polyethylene coatings on metals by isolated particle deposition method.
Additive Manuf, 21 (2018), pp. 191-200
B.R. Rawal, A. Yadav, V. Pare.
3rd International Conferences on Innovations in Automation and Mechatronics Engineering, pp. 60-67
J. Zhang, Z. Chen, L. Wang, D. Li, Z. Jin.
A patient-specific wear prediction framework for an artificial knee joint with coupled musculoskeletal multibody-dynamics and finite element analysis.
Tribol Int, 109 (2017), pp. 382-389
T. Villa, F. Migliavacca, D. Gastaldi, M. Colombo, R. Pietrabissa.
Contact stresses and fatigue life in a knee prosthesis: comparison between in vitro measurements and computational simulations.
J Biomech, 37 (2004), pp. 45-53
Copyright © 2019. The Authors
Journal of Materials Research and Technology

Subscribe to our newsletter

Article options
Cookies policy
To improve our services and products, we use cookies (own or third parties authorized) to show advertising related to client preferences through the analyses of navigation customer behavior. Continuing navigation will be considered as acceptance of this use. You can change the settings or obtain more information by clicking here.