EWA - Manufacturing process and design iterations

Many of the changes made as the design evolved focused on attempting to achieve smooth motion of the slider along the helically shaped path.  A first decision was whether to have the slider rail be an extrusion from the lower arm plate (Figure 1) or a helically shaped bar attached to the lower arm plate (Figure 2). Using a rod would allow a harder rail for the slider than an extruded rail made from 3d printed ABS (acrylonitrile butadiene styrene) or PLA (polylactic acid), and this would be advantageous.  Specialized a freeform pipe bending CNC machines are capable of complicated pipe bending and could theoretically be used for making the bar, however we did not have one available for our use and wanted to produce our design with the means available to us.  Another possibility discussed was the use of a hand bending tool or roller.  The problem with bending tools is that they can only make 2-dimensional bends of a certain radius.  A roller would allow more flexibility in the radius of the curve made, but again is designed for making a 2-dimensional bend.  The technique attempted was the construction of a custom cylindrical jig sized for the curve being made and hand bending the rod round the jig, as shown in Figure 3.  Using this method we were able to bend a piece of 5/16” steel rod into an approximately helical shape.  This technique was unable to achieve a sufficiently accurate curve, so the decision was made to print the rail.

Figure 1. Extruded rail                   Figure 2. Attached rail                          Figure 3. Rail bending method

                A first iteration of the slider mechanism used a 3d printed slider, as shown in Figure 4 below.  This apparatus, as expected, resulted in a slider that experienced significant friction.  Some sort of linear bearing was needed, but the curvature of the slide rail does not allow the use of a standard linear bushing, such as that shown in Figure 5.  A posting on Grabcad,¹ shown in Figure 6, provided an inspiration for a possible solution.  Splitting the slider into a front and back piece and allowing some flexibility between the pieces might allow the slider to follow the helically shaped rail.  This basic design was attempted in several iterations, shown in figure 7 through figure 9 below.  All prototypes used 5mm x 11mm x 4mm bearings.  The wood and printed PLA parts were held together using nuts and bolts, and the aluminum prototype used 5mm shafts with set screws to hold the pieces together.  The PLA sheets were used connect the front and back portions of the aluminum slider.  None of these sliders allowed sufficiently smooth motion when used in the mechanism.

Figure 4. Printed slider            Figure 5. Standard linear bushing¹     Figure 6. Slider design from Grabcad²

Figure 7. Slider in wood                          Figure 8. Slider in PLA.                 Figure 9. Slider in aluminum.

                Several changes were made to the slide rail to improve smoothness of motion, such as printing in ABS rather than PLA, increasing the infill of the printed part to prevent indentation of the rail, and increasing the resolution of the conversion to SLS file in the printed part to increase smoothness of motion.  These changes all helped, but none of the above designs allowed sufficiently smooth motion when force was applied at an angle not parallel to the direction of the rail, as happens when the slider is used in the brace mechanism. 

                The slider was then redesigned with printed U groove shaped covers attached to the bearings to grasp the rail and better distribute the applied force.  These pieces were fit into a flat plate made of printed plastic and secured with set screw, and then this assembly was attached to the hinge.  This change finally allowed operation of the mechanism. The design of the final slider piece is found in Figure 10, with the final slider shown in Figure 11.

 Figure 10. Final slider design.                                       Figure 11. Final slider.

                The manufacturing of the hinge on the upper arm and the hinge attached to the slider was straightforward with no design changes.  Flanged bearings  (1/4 ID, 1/2 OD) were inserted into either end of a piece of aluminum and a 1/4" steel shaft inserted through holes in the printed PLA pieces.  This was all held together using a retaining clip on the outside of the bearings.  A 3/8" steel shaft between these two hinge pieces was held in place using 1/4" set screws, allowing alteration of the length of the rod and adjustment of the relative angle of the two hinges.  These pieces are shown in Figure 12 below. 

Figure 12. Hinges.

                Parts for the attachment of the mechanism to the upper and lower arm were constructed using 3d printed PLA and ABS.  Foam padding was added to the undersides of the plastic to increase comfort for the wearer.  Velcro straps were threaded through holes in the plastic parts for attaching them to the arm.  During testing it was found that there was too much movement of the lower piece when the mechanism was used.  To correct for this movement and add stability in the mechanism a handle was added to the lower arm piece, shown in Figure 14.  The final product is shown in Figure 15. 

Figure 13. Arm attachment straps, upper arm.              Figure 14. Lower arm piece with handle.      Figure 15. Complete prototype.