Since the concept descriptions generated in Assignment 2, more insight into the two primary design concepts has been obtained. In this section, the Go/No Go matrix was used to select the two concepts, and then to refine them (See Appendix B Go/No Go Matrix). The refined concept descriptions including new information are provided below.
Figure 1 below shows a conceptualization of Concept 1-Sonar before refining the concept. The current rendering of the robot can be located in the CAD Drawings section. Wind is collected through a turbine connected to a motor, which generates energy stored in a battery. The DC electric energy in the battery will be used to propel the device around the track with the assistance of two wheels and a castor wheel, as well as multiple motors to allow for turning. The device will be activated by a button and features an Arduino Uno for the brain, which we will program to use a sonar sensor to read data. For each component, the means by which it will be acquired is listed in Appendix C.
Figure 1: Concept 1-Sonar
Since Assignment 2, our team has mostly decided on exact parts, so we can now describe parts in exact terms, rather than general ideas. The turbine will be an assembly of 3D printed parts; a hub, 12 propeller blades, and a wooden motor bracket (See Appendix C Engineering Drawings). The turbine will not attach to the robot but will be connected to a funnel made of cardboard or another similar material, which will then be taped directly to the fan. The purpose of the funnel is to compress the air to give the turbine blades more momentum and generate a little extra energy. The wind will be converted to DC electricity and stored in 3.7V, 3000mAh rechargeable batteries. This design will use three rechargeable batteries in a battery clip to allow the robot to draw enough power to get moving. For the turbine, we elected to use a Hobbypower A2212 1000kv brushless motor, whereas each wheel will be moved by a mini 12V DC 35RPM 25N.cm gear box motor.
The wheel will be a plastic 3D printed rim, featuring an appropriately sized rubber band for a “tire” to give the robot more traction. In the interest of calculating the maximum speed of the device, given that the diameter of the tire is 95.92 inches, it follows that the circumference of the tire is
Where D is the diameter of the tire. Therefore, the maximum speed of the vehicle can be calculated using Equation 4.2:
Where is the angular speed of the motor shaft in revolutions/minute and C is the circumference of the tire. If we were to operate the robot at the maximum speed, it would likely not have enough traction to move very well and would definitely not turn well. Additionally, it would consume power at a very high rate. Therefore, we will operate the motors at a lower rpm which will be experimentally determined to allow for better torque, handling, and efficiency.
Though we originally designed the robot with two sonar sensors (ultrasonic sensors), we decided that our robot chassis is small relative to the sensor and would only need one to be able to turn accurately enough to avoid contacting the walls of the course. Using one sensor rather than two in this case lowers the overall cost of the design without compromising its performance, making the device less expensive and more marketable.
In Assignment 2, we noted that the parts for this design could mostly all be purchased, which makes the parts standardized and easy to acquire if one needs replacing. However, after creating a bill of materials and evaluating our budget, we decided to manufacture the chassis, the tires, the wheels, the turbine assembly, and the motor bracket ourselves. The tires, wheels, and turbine assembly will be 3D printed at the Tom Love Innovation Hub, and the wooden chassis and motor bracket will be manufactured in the Tom Love Innovation Hub.
Lastly, our team made the decision to put all operating machinery on the bottom of the robot, to allow for a large, flat surface to carry a payload on.
While this design is able to pass every criterion of the Go/No Go matrix (See Appendix B Go/No Go Matrix), we leveraged the results of the analysis to further refine the design concept. For the cost, we aimed to be significantly under the budget of $100, around $80, and came out at $76.92, leaving us almost 25% of our total budget! With respect to “Parts Available”, initially our team was expecting to purchase more components of the vehicle, such as the turbine assembly and wheels. However, we opted to fabricate the parts ourselves to gain valuable experience in manufacturing and to help exceed our budget goal of $80. This did of course affect “Build Time”, and while we still have two parts left to manufacture and assemble when we hoped to be finished, we are on schedule to complete the robot with almost a month to test the prototype and conduct a post-mortem analysis. “Standardized Parts”, was virtually unaffected by our decision to manufacture our own parts, because we still have the digital files used to create the parts. Therefore, in the event that one of our self-made parts fails, we will be able to create a replacement within a day. In order to optimize “Payload” we put the machinery on the bottom to create a surface to carry the payload.
Figure 2 below shows a conceptualization of Concept 2 before refinement. The current design of this concept can be found in the CAD Drawings section of the document. In the event that Team Breaking Wind needed to resort to our secondary design concept, there would be several components which would remain the same; the chassis, the tires and wheels, the castor wheel, the turbine assembly and funnel, Arduino Uno, and programming. The difference in the two concept lies in the wind conversion and navigation of the robot. Concept 2 utilizes an alternator rather than a generator, and infrared sensors in place of the sonar sensors. One other small difference is the use of a switch to activate the device rather than a button.
Figure 2: Concept 2-Alternator
Like Concept 1, we originally designed this concept two use two sensors, and decided on one infrared sensor for the sake of cost. Ultimately, the decision to use sonar on our primary design and infrared on our backup came down to marketability of our product. While sonar sensors work by emitting sound waves and measuring the time it takes for the wave to return, infrared sensors work by emitting infrared light waves (or IR) and measuring the time it takes for the light to return. Because the cost of both were comparable, we decided to buy sonar over infrared because sonar sensors work in the dark, whereas infrared sensors do not, making our robot more marketable (MaxBotix Inc., 2018).
Because this design would use the same motors we already possess to move the tires and wheels used in Concept 1, the maximum speed would remain 14.6 . Therefore, the same applies: we would experimentally determine the most effective rpm setting of the motor to achieve a balanced torque to speed ratio.
If Concept 2 is used, the alternator voltage output will need to be rectified due to high voltage output.
Once again, in Assignment 2 we were overly focused on standardization of parts and noted that the components of this design could mostly be purchased as a positive. Since Assignment 2 we have decided to focus more on budget and gaining the valuable experience of manufacturing parts ourselves. For this design concept, we would be manufacturing the chassis, wheels, turbine assembly (propeller blades, hub, and motor bracket). The chassis and motor bracket would be made of wood, and the other parts would be 3D printed at the Tom Love Innovation Hub.
Lastly, we decided this design would also benefit from putting the machinery and electronics on the underside of the carriage to provide a sizeable flat surface for carrying a payload.
This design concept was also able to satisfy each criterion of the Go/No Go Matrix (See Appendix B Go/No Go Matrix). We leveraged the Go/No Go matrix in a very similar way, since the two designs have little variance. Once again, with this design we aimed to spend approximately $80. Infrared sensors and ultrasonic sensors for Arduino Uno cost the same, so the only price difference is between the alternator and the generator. A generator suitable to our purposes would cost $18.99, bringing the price up by $2.00. Because of this, coupled with the alternator’s need to be rectified, we decided to go with a generator on the primary design. Similar to Concept 1, we chose to manufacture every part that we could for the cost and the experience. Once again, the build time on this design concept would be roughly the same as the first. The first design concept took three days to complete (taking into consideration the parts which have not been made yet), and parts could be acquired from RadioShack that day if necessary. Meaning this design concept could conceivably be built in 3-4 days, leaving two weeks for testing and post-mortem analysis even if we don’t find out for another week that our first design concept needs to be scrapped. Birch plywood has been used to cut “Weight” and “Cost”, and lastly the machinery for this design would also be affixed to the bottom of the vehicle to provide a space for carrying a payload.