Phase 4 - Rotating Mechanism

Overview

This phase involved creating the rotation mechanism (“shoulder”) to move the lifting mechanism, grasping mechanism, and payload designed earlier in the project. This design will incorporate gear train calculations to allow the servo to rotate the required amount to get to the desired drop off spot. 

Constraints

• Your rotary table must be gear-driven by the third servo, and the servo must travel through at least 150° of rotation during the pickup/drop-off sequence.

•  Each team will receive a 10” diameter circle of 0.75” thick MDF and two drop zone boards (also 0.75” thick MDF). The circle has a 1/8” diameter hole in the center, added during the manufacturing process, that your rotary table can pivot on should you choose to use it. This piece serves as the ground and you may modify as needed drilling additional holes to anchor gears, servos, etc.

• The MDF circles and drop-off boards are designed to be arranged in a polygon pattern, with the number of sides depending upon the number of groups in the class. A sample arrangement for five groups is shown the top right of Figure 3 but you must modify the interior angle based on your section.

• Each mechanism must rotate 180*(N-2)/N degrees, where N is the number of groups in the class.

• Previous phase deliverables must integrate with current phase.

• Final design should assemble simply.

Sketches

• Paul's Design consists of a simple 7:6 rotations gear ratio. The gear rests and rotates on plastic feet glued to the base. The servo is placed above the 6 in diameter gear to keep the design thin. A pro of this design is that it is very simple, but this poses difficulty when mounting the claw and gears. The gear ratio is designed so that the servo rotates 150 degrees to meet our constraint, while reaching our drop zone.  A con to this design is that it may be less stable because of the lack of thickness in the gears. 

• Jorge's Design takes less space on the base and everything looks more simplistic. The bracket for gear driving servo also took less space which made the design more aesthetic. However, it involves three gears or "gear train" which makes the process a little more calculated in terms of how it would work with the gear ratio. The calculations for the teeth will be a little bit different as it would a set of three driving gears but also one  thicker gear that offsets to the base that could cause friction within the movement of the gears. If there was a way to avoid the friction from the middle gear to the gear of the center of the base then the design would demonstrate more promise.

• Andrew's Design has the benefit that it stores the servo underneath and does not need a large assembly to hold it above the gear.  While a bracket is needed, it is not as complex as if the servo was above, and the servo fits nicely underneath the gear train. However, the gears are very large and mounted very high due to the servo's height.

• James's Design is largely 3D printed. First, it featured two "containers" that functioned as bearings for the two gears to easily rotate in place via minimizing friction (instead of rotating on the MDF test bed). While this was a good idea, it was quite clunky, and the design could have been streamlined to minimize the size and amount of filament used. James's design also utilizes a 3D printed driven gear. The purpose of this is that, by being able to 3D print this gear, the team would be able to also 3D print a piece sticking out of the gear that could function as a screwing point for the entire arm mechanism. By having this joint being made with the same material as the gear, the whole mechanism would theoretically be much more rigid. Lastly,  James's design had a very small driving gear and a very large driven gear. The thought behind this was that, through the use of a gear ratio, the arm would be able to easily move. However, the team ultimately decided that they did not think the arm mechanism would be difficult enough to rotate to justify such a dramatic gear ratio. 

Decision Matrix

Upon doing the decision matrix we decided to continue forward primarily with Andrew's design, as it seemed to be both the simplest and most effective. In particular, we wanted to continue with the design aspect that involved creating a 3D printed bracket for the servo to lie underneath. We also really liked James's idea of incorporating the servo bracket into the driven gear, so  we will continue with that. In addition, looking at Paul's sketch also gave us the idea of having a "partial gear," where the driven gear only has as many teeth as necessary to conserve print time and material as well as allowing the arm mechanism to reach further down. 

For our rotation mechanism, the aspect that required the most consideration was bringing the design to 3D space and making sure everything could easily be assembled. Before designing the mechanism we decided on the gear ratio. The constraints stated that our servo needed to rotate 150 degrees,  and we also need the actual tower to rotate 130 degrees so we could reach the drop off zone. This gave us a starting ratio a 150:130 ratio is approximately equal to 7:6. This ratio allows us to not only meet the constraint but reach our required drop zone as well. We decided to 3D print the driven gear because we wanted a recession in which we could put a bracket for our lifter arm. The raised teeth of the driven gear allow us to place the servo under the driver gear so we did not have to make an overhanging bracket. We also wanted to make the mechanism very rigid with little play. We did this by using large diameter gears, we also printed a slider for the driven gear to reduce friction. The servo and Arduino are attached using brackets fastened to the given base.  Overall, our priorities were to make the mechanism strong and cohesive. 

Video

Reflection

When it came time to test our claw's performance, we were overall quite happy with it. The mechanism preformed its task adequately and efficiently like we expected. The arm was robust and did not wobble while turning. It took quite a bit of time to tune the base so that it would pick up and drop the payload right on the center of the drop zones, but after a while of messing with it we were able to land the payload exactly where we wanted it to. 

However, our grasper consistently dropped the payload about a centimeter in front of the drop zone. So while we were perfect along the X-axis of the drop zone, we were always a little off on the Y-axis. 

In addition, when it came time for testing we quickly realized that we made the tolerances for payload pickup a little too small. Whenever a neighboring claw dropped the payload onto our dropzone, it was usually a little off the mark. If it was placed a little too far back, for example, our claw would not be able to pick it up. However, if it was ever dropped perfectly onto our pickup zone, we were able to efficiently and accurately drop the payload onto the neighboring claw. 

The main thing to change about the mechanized arm that could use some improvements is the grasping mechanism. First, we can change it so that it will not drop the payload a centimeter too forwards. Second, we can also increase the "play" the claw has so that it can successfully pick up payloads that were dropped a little off center.