A biphasic model of soft tissue has been combined with a simple confined compression test to measure the hydraulic conductivity of soft tissues. The biphasic model treated the tissue as a mixture of an incompressible fluid and porous elastic solid. The equations were linearized to the form of a diffusion equation with constant coefficients. The equation was solved using forward finite differences in time and central finite differences in space as a function of the material parameters. The resulting stress at the boundary was then predicted. The rms error between the measured stress at the boundary and the model predicted stress was minimized by varying the material parameters using a Levenberg-Marquardt algorithm. Ten right circular cylinders, about 12.7 mm in diameter and 5 mm in height, were cut from fresh bovine heart muscle and tested within 24 hours of sacrifice. Samples were placed in a 12.50 mm polycarbonate cylinder with a close fitting piston (-0.01 mm). A 40 um sintered filter allowed fluid to escape from the bottom of the cylinder. The piston was placed in a bath of buffered saline maintained at 36 degrees C that was mounted in a commercial tensile machine. A 100 N load cell (0.1 % absolute error) measured load while a LVDT measured displacement (0.1% absolute error). The sample was ramped in compression using displacement control at 0.003 mm/s until a peak force of 3 N was reached. The peak displacement was then held for 900 s. The sample was then unloaded and allowed to recover for a several minutes before loading again. The applied displacement and the resulting force data were collected. The linearized model could not capture the relatively fast changes in the stress as it reached the peak load, but was able to predict the stress relaxation portion of the test. The hydraulic conductivity and the effective elastic constant were significantly different from cycle one to cycle two using a paired two-tailed t-test. As the fluid flows out of one end of the tissue, the fluid channels at this end collapse, resulting in an effective reduction in the local permeability. A higher load is then needed to push additional fluid out. This behavior is not modeled by the linearized equation and is responsible for the mismatch in the load peaks. Since the samples were not allowed to relax completely back to the zero state between cycles, the effect carries over into the second cycle, resulting in new effective values for the material constants.
Proceedings Title: Proceedings of the ASME Summer Bioengineering Conference
Conference Dates: June 20-24, 2007
Conference Location: Keystone, CO
Conference Title: ASME 2007 Summer Bioengineering Conference
Pub Type: Conferences
soft tissue permeability, hydraulic conduction, biphasic model, compression test