Contact stress aberrations following imprecise reduction of simple tibial plateau fractures.

Despite the well-recognized association between poorly reduced intraarticular fractures and late degenerative changes, current guidelines regarding the reduction precision necessary to avoid excessive cartilage pressures are based largely on anecdotal clinical observations. To gain a quantitative appreciation of the relation between local pressure elevations and fracture reduction imprecision, a simplified laboratory cadaver model of minimally displaced tibial plateau fractures was developed. Cartilage contact stress distributions were measured as a function of depressed fragment malreduction in seven knees, using high-resolution (100 pixels/mm2) digital image scans of Fuji-film stain patterns. The contact stress data showed a general trend of increases of peak local pressure with increasing fracture site incongruity, and in a few isolated instances the effect was very pronounced. Across the whole series, however, statistically significant departures from anatomic pressure levels did not occur until the fragment stepoff was greater than 1.5 mm. Even at the 3-mm stepoff level, for which the depressed fragment usually no longer made contact with the femoral condyle, the peak local pressure values on the intact side of the fracture line averaged only approximately 75% greater than those prevailing anatomically. Given the successful clinical outcomes normally achieved for conservatively managed simple tibial plateau fractures having stepoff magnitudes (5-10 mm) clearly sufficient to insure fragment articular noncontact, the present laboratory results suggest that nominally factor-of-two peak local pressure elevations, provided that they occur over only small portions of the cartilage surface, are probably within the long-term overall tolerance range of an articular joint.

Experimental determination of the linear biphasic constitutive coefficients of human fetal proximal femoral chondroepiphysis.

The mechanical properties of the cartilaginous regions of the proximal femoral epiphysis are an important factor in load transmission through the hip joint of young children. Cylindrical test specimens excised from the chondroepiphysis of human stillborn femoral heads were subjected to uniaxial loading in peripherally-unconfined compression, using a ramp/plateau input strain history. The corresponding load vs time curves were analyzed in terms of a recent analytical solution for a linear biphasic material (the well-known KLM model), allowing calculation of that model's three fundamental constitutive coefficients (permeability, equilibrium modulus and solid-phase Poisson ratio) for this material. The numerical algorithm developed to evaluate the biphasic solution yielded very precise replication of previously published KLM parametric plots. When fitted to experimental load histories, however, the model provided only a rather loose approximation of specimen behavior, due apparently to a substantial underestimation of the transient response component associated with interstitial fluid transport. Averaged over the series, the best-fit values for permeability (2.51 X 10(-15) m4 Ns-1) and equilibrium modulus (0.699 MPa) were in the range of values accepted for human adult articular cartilage. A consequence of the coarseness of the analytical curve fits was that a solid-phase Poisson ratio of 0.0 was inferred for all specimens. The permeability vs equilibrium modulus exhibited a nearly linear (r = 0.74) inverse relationship similar to that reported for adult articular cartilage.

Dynamic performance characteristics of the liquid metal strain gage.

Performance characteristics of the liquid metal strain gage (LMSG) were evaluated by both static and dynamic bench testing. Statically, the devices were found to have outputs closely proportional to engineering strains, up to strain levels of 40%. While individual gage factors varied appreciably (up to 50%), each of the gages studied showed excellent reproducibility of behavior. Dynamically, the response to sinusoidal strain inputs was frequency-independent up to 50 Hz, and there was no detectable phase shift. Similarly, the LMSG response to constant-speed displacement inputs was velocity-independent over the range of nominal strain rates from 20 s-1 to 0.02 s-1. The devices proved capable of maintaining stable outputs when held stretched to fixed lengths, even if such tests were performed immediately following stepwise displacement inputs. Thermal artifacts were found to be modest (0.185% apparent strain per degree C), and there was no appreciable sensitivity to non-axial strains. When mounted on an in vitro ligament preparation, the LMSG measured apparent ligament strain similar to that detected by a video dimension analyzer. A protocol by which an implanted LMSG could be used to infer in vivo muscle forces was demonstrated, based on recordings of tendo-Achilles strains developed by a rabbit during slow hopping.