Manufacturing and Assembly:
We utilized several different materials and manufacturing processes in the manufacturing and assembly of our machine. At the bottom of this page, there are several images of our final product along with a video to showcase the motions.
Manufacturing:
Cylinder and Spacer: The cylinder was made of ABS and cut with a band saw. It was attached to the wooden pillar with two 3.5 inch wood screws. The cylinder was also spaced from the wooden pillar via a 3D printed PLA spacer. One of the surfaces on this spacer was flat to sit flush with the wooden pillar and the other surface was curved to fit with the profile of the cylinder.
Base: The base was created using 2x4 inch blocks of wood. This wood was cut using a band saw to size, and counter bores were made with a hand drill press to house large hex bearings. The two pillars were approximately 1 ft in height, while the base was 1.5 ft in length. The pillars were screwed into the base using angle brackets and wood.
Cams and followers: The cams were made from 3D printed PLA and had a design to press fit a 3/8 inch hex nut inside. The cam follower was also made from 3D printed PLA and was affixed to the top of the valve stem.
Cylinder cap: The cylinder cap was made from 3D printed PLA to house the valves and valve springs. The openings for the valves were hexagonal to fit a 3/8 inch size hex nut snugly. This was so that when the valves were actuated they would not move in an undesired direction.
Valves: Our valves were created by using a Dremel cut bolt that could accommodate nuts, springs, and spacing sleeves. We decided on this design because of its adjustability and affordability, it would be easy to cut the bolt down with a Dremel as well as adjust valve position by rotating the nut to a different placement.
Piston: The piston was made from 3D printed PLA, it was sized to fit well within the 3.49 inch inner diameter of the cylinder.
Piston linkages: The linkages were originally made of laser cut wood and rotated on an aluminum shaft cut with a band saw. After some testing, the linkage material was changed to aluminum cut using a band saw. Holes were drilled using a drill press and rotary bearings were press fit into the bores. These linkages were attached to the piston with an aluminum dowel cut using a band saw.
Timing pulleys: This project utilized 2 timing pulleys to translate rotary motion from the input crank to the camshaft. The pulleys were made of 3D printed PLA with the cam pulley having a 2 to 1 ratio to the crankshaft pulley. The camshaft pulley has 4 holes at the outer edges of the radius to accommodate screws to attached to the Lazy Susan rotary bearing. It also has a hexagonal hole in the center to accommodate a locking hex nut. The crank pulley was press fit onto the aluminum rotary crankshaft.
Camshaft: The camshaft was a threaded rod that could translate back and forth within the cam pulley and hexagonal bearing in the wooden support pillar. There are hexagonal nuts within the cam pulley and hex bearing that stay in place, these allow for the camshaft to spin and thread back and forth.
Miscellaneous Parts:
Rotary input handle: The input handle was made from 3D printed PLA.
Lazy Susan: The Lazy Susan bearing was fixed to the wooden support pillar and allowed to rotate with the camshaft pulley. It was affixed to the wooden support using wood screws and to the cam pulley using machine screws.
Shafts: All dowel shafts were 6061 Aluminum and cut with either a band saw. The camshaft was a threaded mild steel rod and cut using a rotary abrasive saw.
Assembly:
The machine was assembled first by affixing the cylinder and spacer to a support pillar without installing the pillar onto the base. The 3D printed cams were screwed onto the camshaft and put into place with locking nuts. Then, the camshaft was installed into the hex bearing on both pillars and fixed into place with locknuts. Within the linkage components, rotary bearings were press-fit into all applicable machined holes. The crankshaft is split into 2 parts, the linkages and piston were installed using the pre-cut dowel pins. Both sides of the crankshaft were then press-fit into the piston linkages while also being press-fit into place in the two vertical support pillars. After this, the cam pulley and Lazy Susan were installed, the Lazy Susan was drilled into the support pillar with the piston cylinder assembly. The cam pulley was affixed to the camshaft and bolted on to the Lazy Susan. The crank pulley was then press fit onto the crankshaft. These pillars were then attached to the base first using the angle brackets. After they were secured onto the base they were drilled into place from the bottom using 3.5-inch-long wood screws. After securing the base, the timing belt was affixed to the pulleys. The valves and valve cap were then placed on-top of the cylinder. To finish off the assembly the crank handle was press fit onto the input side of the crankshaft.
Things to Improve:
Tolerances:
We had a lot of trouble with tolerances during our manufacturing process. 3D printed parts had a lot of error in dimensioning values because of the mesh size, printing quality, and the feed rate of the end effector. For example, if we were to make the cams and cam followers out of machined aluminum instead of 3D printing, we likely would have seen a much better valve actuation. With the 3D printed parts, we had trouble creating large valve actuations. While the aluminum parts would likely also have to be lubricated, they would have provided a better example of valve movement.
Increased Friction:
In addition, we also had a lot of friction between the piston and cylinder wall. This was caused by a few factors. Once the belt was placed onto the pulleys, the camshaft was pulled up while the crankshaft was forced downwards. This adjustment in the rotating shafts cause the piston to traverse upwards in a slightly different path than we expected. Because of this, the piston would hit the cylinder wall on upwards motion and try to fall out of the cylinder on the downwards motion. Another source of friction was the linkages being misaligned and creating contact friction. This was a factor of very slightly misaligned crankshaft input holes. Because these holes were where the crankshaft was press-fit into, the linkages did not move together at the exact same time. This also contributed to the friction between the cylinder wall and piston. The solution to this would be to CNC these parts instead of manually making holes.
Material Choice:
Another issue that contributed to the misalignment of parts was the material selection. The wooden base was subject to movement and misalignment with the metal shafts, as well as poor press-fit with the hex bearings. If the base was made from aluminum, it is likely that we could have avoided these poor fitting issues. Another issue was the PLA used throughout the machine. Often, we would have to modify these parts after being printed to work the way they were designed. PLA also has rough edges that must be sanded down to use for heavy contact parts.
In summary, these three factors combined and did not allow our robot to work completely as intended. It was interesting to see the difference between designing in theory and manufacturing, every seemingly small issue was magnified in the final operation. In future attempts, the robot parts should be CNC machined from aluminum to eliminate tolerance error and misalignments. We realized this mistake during the manufacturing process and made the linkages from aluminum, but our project would have benefited if most components were also made from aluminum.
Further Improvements and Advancements:
Adding a motor input with feedback control:
If we could add a motor input with an encoder to sense the crank angle, we could project how the valves actuate at different RPMs, as well as show how an Internal Combustion engine works at rated speed. Furthermore, implementing a closed-loop PID control would be an interesting way to see how the motor speed adjusts for the different cams, and give insight into how variable valve timing works within the engine assembly.
Variable Compression:
Automakers are just starting to implement motors with variable compression. This technology basically allows for an engine to change compression ratio based on engine load. This can be achieved via controlling crank motion as well as valve timing. It would be interesting to see if this sort of variable compression could be achieved mechanically through linkage analysis or feedback control.
Final Assembly shown below (Various Viewpoints)
