Specimens of bovine, rabbit, and human corneas were systematically tested in uniaxial tension to experimentally determine their effective nonlinear stress-strain relations, and hysteresis. Cyclic tensile tests were performed over the physiologic load range of the cornea, up to a maximum of 10 percent strain beyond slack strain. Dimensional changes to corneal test specimens, due to varying laboratory environmental conditions, were also assessed. The measured stress-strain data was found to closely fit exponential power function relations typical of collagenous tissues when appropriate account was taken of specimen slack strain. These constitutive relations are very similar for rabbit, human and bovine corneas; there was no significant difference between the species after preconditioning by one cycle. The uniaxial stress strain curves for all species behave similarly in that their tangent moduli increase at high loads and decrease at low loads as a function of cycling. In the bovine and rabbit data, there is a general trend towards more elastic behavior from the first to second cycles, but there is little variation in these parameters from the second to third cycles. In comparison, the human data demonstrates relatively little change between cycles. Increases in width of corneal test specimens, up to a maximum of 2 percent were found to occur under 95 percent relative humidity test conditions over 10 minutes elapsed time test periods, while specimens which were exposed to normal laboratory conditions (45 percent RH) were found to shrink in width up to a maximum of 9.5 percent over the same elapsed time period. The thickness of the test specimens were observed to decrease by 3 percent in 95 percent relative humidity and by 12 percent in 45 percent relative humidity over the same elapsed time period.

No abstract

A new knowledge-based approach for the synthesis of mechanisms, referred to as Pattern Matching Synthesis, has been developed based on a combination of committee machine and Hopfield neutral network models of pattern classification and matching applied to coupler curves. Computational tests performed on a dimensionally-parameterized four bar mechanism have yielded 15 distinct coupler curve groups (patterns) from a total of 356 generated coupler curves. This innovative approach represents a first step toward the automation of mapping structure-to-function in mechanism design based on the application of artificial intelligence programing techniques. Demonstrative examples of its application to “real-world” mechanism synthesis problems, including the design and evaluation of a two-stroke pump mechanism and the redesign of a variable-stroke engine mechanism have been included, establishing its viability for creative mechanism synthesis.

No abstract

The use of a perforated, titanium funicular shell to support the proximal femoral cortex in total hip arthroplasty was evaluated with the aid of both analytical and numerical techniques. The principal interactions between the femoral cortex, the metal shell, the implant stem and the acrylic bone cement were modeled using beam on elastic foundations theory and two-dimensional elasticity theory. Subsequent formulation of this model as a nonlinear design optimization problem enabled the determination of the dimensions of the implant and reinforcing shell which minimized an objective function based on a simplified material failure criterion. Two cases were examined, each with two cervico-diaphyseal angles: case A: with a rigid contact between a proximal prosthesis collar and the calcar femorale and case B: no collar contact (a collarless prosthesis or post-operative loosening). Case A achieved an optimal solution at a stem diameter 11-23 percent of the cortex inner diameter, a stem length to diameter ratio of 12-40, shell diameter 22-53 percent and thickness 0.2-7.2 percent of the cortex inner diameter and thickness, respectively. Case B achieved an optimal solution at a stem diameter 67-92 percent of the cortex inner diameter, length to diameter ratio of 4-6, and no shell. In case A the collar support makes the type of internal fixation unimportant, while in the more realistic case B, the shell is not recommended.