A Biphasic Finite Element Model of In Vitro Plowing Tests of the Temporomandibular Joint Disc

Disorders of the temporomandibular joint (TMJ) afflict 3-29% of people aged 19-40 years. Degenerative joint disease (DJD) of the TMJ generally occurs 15 years earlier than in other human joints and 1.5-2 times more often in women than men. The TMJ disc is the primary stress distribution mechanism within the joint. Mechanical failure of the TMJ disc precedes clinical signs of DJD. Unlike postcranial synovial joints, biomimetic replacements of the disc have not been successful, probably due to the paucity of knowledge about TMJ biomechanics. Translation of stress-fields mediolaterally across the TMJ disc may lead to fatigue failure because of the effect of traction forces on the tissue surface and because the disc is relatively weak in this aspect. Traction forces are composed of friction forces, which are known to be low in the TMJ, and plowing forces which are relatively much higher and result from movement and pressurization of fluids within the tissues due to translating surface loads. In the in vitro plowing experiment, a rigid curve-ended indenter is lowered into a TMJ disc that has been mounted on a stage with pressure gauges, and the indenter is then translated in a prescribed mediolateral motion that is intended to simulate the motion of the mandibular condyle on the TMJ disc in vivo. As a first step, these plowing experiments have quantified the variables thought to be important in tissue failure. A next step is to define the full role of these variables in the pathomechanics of TMJ disc tissue through a validated model. Therefore, the aim of this study was to develop and test a finite element model of the plowing experiments based on an orthotropic biphasic description of the soft tissue behavior of the TMJ disc. For this plowing model, the arbitrary Lagrange Eulerian method was used to approximate the moving load problem, where in vitro the indenter slid along the tissue's superior surface. Approximate validation of the plowing model was based on comparisons of model-predicted temporal and spatial distribution of indenter displacement and total normal stresses (+/- 15%) and laboratory measurements during one complete cycle of plowing motion. Other useful predictions from the plowing model include spatial and temporal distributions of biomechanical variables of interest that cannot be measured experimentally, such as total stress, pressure, strain, and the relative significance of the orthotropic solid phase properties.