...
The images below show the fabrication of the prototype in process. All parts were laser cut. A Test Strip of 20 holes ranging in size by increments of 0.001" diameter was made to find the best diameter for a good joint fit. This determined a dimension for the holes that prevented excessive play, but allowed the joint to move without binding due to friction. The joints in this prototype are 1/4-20 bolts, with washers between each moving part, and a locknut. The locknut was not tightened to the point of clamping the parts together in order to maintain their freedom of movement.
In addition to the laser cut links, a few 3D printed spacers were used where necessary to achieve the required clearances between legs. A CAD model of the complete prototype is shown below. This model is exactly to scale with the 8-bar mechanism link lengths analyzed in the "analysis" section. The 3D model provided some insight to design considerations that are unconsidered when using only a 2D model. For example, multiple overlapping links must use low profile joints to prevent interference during the tight clearance portions of the movement. In this prototype, low profile bolts were used, with the heads facing inwards to prevent jams. This 3rd dimensional aspect will be taken into consideration for the final design. Ideally, the joints would all be as close to the plane of the baseplate as possible to prevent an excessively cantilevered joint from causing undue deflections in the mechanism.
The video below demonstrates the final prototype, with the addition of a marker holder and trace paper following a path produced by the lower portion of the y axis adjustment. This path is near the shortest possible stride length with this configuration.
| View file | ||||
|---|---|---|---|---|
|
In conclusion, the prototype proved the viability of the 8-bar mechanism for a walking robot with a dynamically adjustable gait. Moving the adjustable joint up and down in the vertical slot has a very noticeable impact on the stroke length when observing the traced lines left by the foot on the paper. This prototype also demonstrated the potential of a simple joint design using only a bolt in a precisely dimensioned hole. This approach is much simpler and cheaper than using bearings, but is quite sensitive to small changes in the hole diameter. Optimized joint design, minimized cantilever length, electronics, and the stroke length adjustment mechanism will be the main areas to focus on for the final design.