The primary goal of the Jansen mechanism is to replicate the motion of walking, specifically in regards to the path of the foot joint. The foot joint (i.e. joint HI) should follow something resembling the path of a human foot. This path can be verified by comparing a simulated footpath from MATLAB to an experimental path.
The simulated path is generated from the same MATLAB script used in the previous section. It is available for download in the appendix. The plot below shows the X,Y potions of the foot joint HI for all values of input angle.
This path is to be compared to the path below which is the path of the foot joint HI on one of the prototype legs. Joint HI was made capable of drawing by inserting a piece of mechanical pencil lead through the holes in the links along with the toothpick. The mechanism drew this path in a full rotation of link M.
The most obvious difference between these figures is that the first is one loop while the second has a figure-8 shape. This is due to error in the physical prototype. The small loop on the left side of the figure, specifically, is error. This error is likely due to links being made close to, but not exactly, the correct length. When in a position that has the foot on the left portion of the figure-8, the foot joint moves much faster than in the rest of the cycle. This means that the error only exists for a very small portion of input angles, meaning that the foot stays on the desired path for most of each cycle, then flies into error for a few degrees of each cycle.
Disregarding that error in the prototype, the paths are extremely similar. Both include a sharp edge on the bottom right which is the point that the foot rises up to begin a stride. The foot moves upwards in a concave-up shape then descends down to the ground. Once the foot moves from striding to propulsion, it moves back and downwards since it is propelling the body up and forwards. The main differences between these two shapes is that the peak is slightly to the left in the simulated trajectory, and the propulsion trajectory is mostly concave-down in the simulated trajectory while its mostly concave-up in the experimental trajectory. However, these two differences are marginal in regards to the goal of this project. The differences can likely be attributed to the same errors that caused the extra loop in the experimental drawing.
In all, the Jansen mechanism was a success. Aside from the figure-8 error and the stickiness in practice, the mechanism was able to produce a foot trajectory that was verifiable with simulations in MATLAB. Both of these foot trajectories resemble a natural human foot trajectory.