Dynamic Friction Analysis of Articular Cartilage
Daniela Warnecke1, Maxi Klengel1, Luisa de Roy1, Anita Ignatius1, Lutz Dürselen1
Insitute of Orthopedic Research and Biomechanics, Trauma Research Center, Ulm University – Medical Center, Germany
1
Disclosures: Daniela Warnecke (N), Maxi Meßemer (N), Luisa de Roy (N), Anita Ignatius (N), Lutz Dürselen (N)
INTRODUCTION: Articular cartilage exhibits a remarkably low friction coefficient. Furthermore, cartilage friction is complex as it depends on a variety of
parameters like time, lubricant, sliding velocity, applied normal load and opposing surface [1, 2, 3]. Most previous studies investigated cartilage friction
under constant normal load and sliding velocity. There is only one single study assessing the friction under sinusoidal loads at constant sliding speed [4]. In
the knee joint, however, the velocity of femoral and tibial surfaces relative to each other and the load are subjected to vary considerably during gait. The
question arises how friction in the knee joint is affected by these dynamic loading and motion conditions. Therefore, the aim of this study was to establish an
experimental setup to investigate the frictional behavior of cartilage against cartilage under simulated physiological conditions in the knee joint during
walking. As a first parameter study, dynamic friction was assessed for two different simulated walking speeds.
METHODS: A dynamic pin-on-plate test setup was developed. Here, a dynamic materials testing machine (ElectroForce 5500, TA Instruments, New
Castle, USA) applied the dynamic normal forces FN acting in the knee joint during walking to the pin, while the plate slid cyclically against it with varying
velocities derived from stance and swing phase for walking speeds of 1 km/h and 5 km/h. Flat cartilage samples (n= 4) harvested from intact bovine femur
condyles (plate, length approx. 50 mm) were tested against cylindrical samples (n= 4) retrieved from the tibial plateaus (pin, Ø= 6 mm). A double peak load
regimen typically occurring during gait was applied representing the load during stance phase (p max≅ 0.7 MPa) followed by a low load plateau (p≅ 0.2 MPa)
simulating the swing phase. During stance phase, the stroke length was 6 mm and 25 mm during swing phase (Fig. 1, A-1, red line). The total duration of
each test was 125 minutes. The applied normal forces FN and the resultant friction force FR were continuously recorded to calculate the friction coefficient µ
(µ= FR/FN) at the onset (µ0) and at the end of the experiment (µend) during stance phase and also during swing phase. The starting walking speed was
randomly chosen and after a recovery time of >12h, the tests were repeated with each friction pairing at the corresponding alternative walking speed. Bovine
synovial fluid was used as a lubricant. Due to the pilot character of this study, the results were only evaluated descriptively.
RESULTS: In all testing configurations, it could be observed that the friction coefficient during swing phase was higher compared to stance phase. Further,
this increase was more marked at 1 km/h walking speed than at 5 km/h. However, no variation over time were found not only for the friction coefficients of
the stance phase but also for the swing phase (Fig 1, B).
DISCUSSION: A test setup for the assessment of joint friction under simulated physiological loading and motion conditions was introduced for the first
time. It allows to monitor the friction coefficient between two cartilage surfaces under varying loads and sliding velocities. The first measurement revealed a
higher friction coefficient during swing phase under low normal force and at high velocity compared to stance phase with a high normal forces and lower
velocity at both the onset of the experiment and after 2 hours of testing. This contradicts the results of e.g. Gleghorn et al., who found decreasing friction at
equilibrium with higher speed and lower strain [2]. However, they tested against glass and under constant strain and speed conditions. This discrepancy
could indicate that friction is even more multifactorial than thought by now. Krishnan et al., who reported on friction under sinusoidal loads found
dramatically increasing friction coefficients during the unloading phase at low sliding speed against glass [4]. This effect could not be observed in our study,
which is most likely attributable to the missing pressurization of the interstitial fluid of the sliding partner glass. Another interesting finding is that walking
speed seems to affect the difference between friction coefficients in stance and swing phase of gait. To understand these effects, further measurements are
necessary.
SIGNIFICANCE: The new testing system allows new insights in joint friction mechanics. More tests e.g. on degenerated cartilage could possibly help to
better understand the development of tissue degeneration in human joints.
REFERENCES: [1] Forster H & Fisher J (1996) Proc Inst Mech Eng H 210: 109-119; [2] Gleghorn JP & Bonassar LJ (2008) J Biomech 41: 1910-1918; [3]
Neu CP et al. (2008) Tissue Eng Part B Rev 14: 235-247; [4] Krishnan R et al. (2005) J Biomech 38:1665-1673.
ACKNOWLEDGEMENTS: This work was supported by the German Research Foundation (DFG/DU254/8-1).
IMAGES:
ORS 2017 Annual Meeting Poster No.1440