The purpose of this study was to quantify human rnetatarsophalangeal joint stiflkess and compare it to an existing artificial foot which is a candidate for the foot in a shoe testing robot. The study was conducted by examining the foretbot flexion produced by external moments about a Z axis, defined by the horizontal line segment between the I& and 5& metatarsal heads, a vertical Y axis, and an X axis normal to each.

It was found that for the 6 human subjects tested, the largest stiflhess coefficient was about the Z axis with a mean stifiess of 1.1 Nm/deg and a standard deviation of 0.1. An MTS machine was used to measure shoe stiflhess, which was then compared with the stiffiess of the human foot measured previously using a force plate and motion control cameras. It was found that the stiffuess of the shoe about the Z axis was relatively small in comparison to the human foot stifiess. These results showed that shoe stifiess does not dominate that of the foot about the Z axis and that foot stiflhess must be incorporated into the artificial foot.

The artificial foot chosen was the Sede Foot by Seattle Limb Systems Inc. The stifhess of the Seattle Foot was similarly determined by the MTS machine and was compared with the stiffness of the human MP joint The stShess of the prosthetic about the Z axis was found to be 37% of the average human stiffhess. It was also found that while a human joint rotates about a point, this particular prosthetic foot bends as a beam throughout the forefoot Thus, a modification to the foot was proposed in which stiff plates would prevent bending dong a selected portion of the forefoot, thus increasing the net stifhess of the forefoot. The required dimensions of the plates were calculated based on a theoretical model, and upon testing the st3hess of the modified foot experimentally, it was found that the stifE~ess had increased to 94% of the desired -as, which was within hatf a standard deviation,