Lego Racer Day(s): 2/20/2015 - 2/27/2015

2/20/2015: Leg-go! 

After the presentation of our windlass, we got started right away on our lego cars! As the title suggests, our objective for this project was to build a lego race car fueled by a PicoCricket and a single DC motor as it carried a 1kg weight down a 3-meter track.

At first, we were given a small assignment to complete before starting our race car. My new partner, Tiffany, and I finished it by the end of the first class.

Gears! Getting them to fit together, using bushings to hold the gears in place not too loosely, but not too tightly...soon, these minor accommodations in this small assignment, such as fitting gear to gear together, motor to gear, axes to vehicle, would become some of the largest challenges of our project.

2/22/15: First Iterations

Tiffany and I came in late on Sunday night to start our first iteration. Grabbing a handful of various lego piece from the cabinet, we got started with a gear ratio of .5 to 1. As we were assembling the pieces together, our primary concern was getting the gears to mesh. We knew we wanted to increase torque, but we also wanted a bit of speed in the end, thus our 1:5 gear ratio on the wheel axle. We should have written out our calculations first, because our overall gear ratio of .5 to 1 wasn't enough to move our car on the ground, with just the battery as weight, at all. We decided on these larger wheels because they were light, and once they were rolling, they would cover more ground than smaller wheels.

Our second iteration of the night focused on increasing torque as much as possible. Since the motor we were required to use only had speed, we thought that more torque was necessary in order to move the 1kg weight. Using almost exclusively 24- and 8-toothed gears to create a 3:1 ratio, we knew this gear arrangement would slow the overall speed of the vehicle, but our first mission was create a product that could carry the 1kg weight across the finish line. We also used 3 wheels this time to reduce the friction between the floor and the car.

It worked! Though...very slowly. Building this iteration took much time that night, because not only were we trying to get the gears to mesh and to prevent the axes from scraping the wheels, the overall structure needed a firmer base. The width was only determined by one piece of lego across, and keeping the width in place while we were attempting to work with the gears was a bit of a challenge. 

This was 27:1 gear ratio, and it was definitely too slow. In the classroom floor that we were testing our racer on, the car seemed to move a bit more freely, but because the real track is on carpet, the friction would slow our car even more, making it as if our car was inching its way to the finish line. We would continue with our iterations, breaking down the entire car to start over with a hopefully firmer base.

In hindsight, after our second iteration, we should have started from the base and gone out further than shoot out far and then reel ourselves back in to the base. In other words, if we had started with a 3:1 ratio gear, tested that out, and then continued to increase torque until we have reached the maximum speed and torque gear ratio, that would have been a more clever approach than putting out as much torque as possible in our product, then reducing it little by little. Because we took the latter approach, our results, calculations, misconceptions threw us off along the way.

2/24/2015 & 2/25/2015: Gears, Gears, Gears 

The bulk of our iterations occurred during these two days. In class, we spent time redesigning our car to increase the speed. However, because we were still too focused on increasing torque, the racer was slow. The gear ratio for this third iteration was 27:1 and after a few seconds of contemplation, we realized we had an extra set of gears that was only contributing to more friction. Our car ran very slowly, more than 25 seconds. We also should not have tried to increase speed in any of the gear ratios, such as in the 40:14 (1:2.85) below, because it will be cancelled out in the overall gear ratio anyways, only to add friction in the process. This design, again, took a bulk of our time because we kept running into problems of meshing the gears attached to the motor to the gears of the vehicle. Meshing different gears to one another across the length of the car was not trivial either, because besides the 24- and 8-toothed gears that fit each other every 2 axle lengths apart, the other combinations required more creativity. It also took us a while to discover we could make our main axle (the one connecting the two back wheels), longer by connecting them another lego piece designed to accomplish the task. For a while, we had been stuck on how to make our base a little wider in order to accommodate the diameter of the 1kg weight, since making it wider without rod extensions meant the axles in the front would bump into the wheels. I wish we could have built these initial models faster and failed faster, because it definitely would have saved us time to work on our final racer.

Taking a more cautious approach, we broke apart our whole car and tried to make a sturdier base again. We realized from our previous iterations that there was unnecessary length in our car, and wanted a shorter base that was a little wider to accommodate more space for the gears. Lessons from our previous iterations, we discovered that it wasn't necessary for us to put the gears on the center or to the side of the axle as well, and that we could use bushings on the inside section of the axle to make the gears stable. This fourth iteration, a small car below, has a gear ratio of 44:1, and finishes the track in 15 seconds.

This model is simpler than the first iterations, and we changed the front wheel so that the weight could be distributed downwards towards the front wheel. By changing the incline, we hoped that some of the weight could push the car forward itself as it increased momentum. The extra lego pieces in the photo of this iteration was to secure the motor and to allow space for the 1kg weight to sit. We also increased the torque, since we thought that it was because we didn't have enough in the last iteration that kept the speed from increasing. Lightening the load, friction, and increasing torque definitely allowed our racer to be faster than its previous iterations, but we still needed to make improvements in order for our car to be competitive in the race.

From this model, I started keeping notes (as scrawly as they are) of the gear ratio in order to maximize our power (and speed!).

As Amy had shown us, we would maximize our power at a balance between torque and speed. The DC motor given to us only had speed, so our goal in terms of the gear ratio would be to find the right amount of torque. Since we had been rebuilding all of our iterations so far, we did not allow ourselves the chance to discover this balance. Not only did the total load of our racer change with each iteration, but so did the friction on the wheels, the number of axles, and the number of gears. In our last three iterations before our final racer, I built a solid base for the first of the three, and afterwards we concerned ourselves with switching the gear ratios to achieve a balance. We did not break down the entire racer as we did in our previous cars, and to make the base sturdier and to eliminate the #1 challenge of connecting the motor gear to the other gears, we firmly set the the motor gear and the wheel gear positions and worked our way inwards, filling it with different pairs of gears that were more readily changeable. We used the fourth iteration's gear ratio as our standard in the next few iterations, and actually built an iteration with the exact same gear ratio with new base.

We took off the 40-toothed gear to try if a 24-toothed could fit, but it just wouldn't. This was our 5th iteration, a little modified already. It ran in 15 seconds with the 40-toothed gear attached.

What happens if we reduced the torque to gain more speed?

The torque of this iteration of the racer was too small. The ratio of the car was 14.6:1, finishing the finish line in more than 25 seconds. The gear attached to the motor here was a 24-tooth gear.

What happens if we increased the torque to be more than our starting iteration (the fourth one)? In our next iteration, we were running our car with a gear ratio of approximately 77:1. Again, the car was finishing the finish line more than 25 seconds, the torque was too large.

Since we were using our fourth iteration as the standard to compare the models afterwards, it was difficult to find the right amount of torque to increase/decrease in order to achieve a faster racer. Sadly, our eighth iteration, the one after the the model shown above, ended up finishing the line around fifteen seconds like the fourth iteration...and because the gear ratio was the exactly the same as the ratio in the fourth iteration. We used time in building this seventh iteration, and I wish I had been more careful with the ratios. The largest challenge we faced with these last iterations, which distracted us from concentrating on the gear ratios, was obtaining the right meshing of gears. A 14-toothed gear did not fit the same way with the 24-toothed gear as it did with a 8-toothed gear, and although we found more flexible options for our car by using the axles diagonal of the other, it was still difficult to achieve the correct fit. Most of our time was spent on figuring out how to mesh one gear into the other in the middle, because although it was more flexible in terms of which mix of pair was allowed, it took a bit of creativity to figure out how to get them to mesh in the limited space. In addition, for our starting gear in the motor, we were limited to a 24-toothed gear to 40-toothed gear pair, or 8-toothed gear to 40-toothed gear pair, from its positioning, because no other mix of pairs would fit perfectly in the space allotted.

Finale! To achieve our final racer, we modified the gear pairs one-by-one instead of multiple at a time. I realized that when we made our 6th iteration, beginning the motor with a 24 tooth to 40 tooth gear made the racer smoothly glide along the floor. Thus, we tested out this product and it made it across the finish line around 10 seconds!

2/27/2015: Presentation Day 

Engineering Analysis:
Gears of Our Final Product From the Motor:
1) 24 tooth to 40 tooth 
2) 8 tooth to 24 tooth (right alongside each other)
3) 8 tooth to 24 tooth (right alongside each other)
4) 14 tooth to 24 tooth (diagonally connecting)
(40/24) * (24/8) * (24/8) * (24/14) = 25.7
Gear Ratio: Approximately 25.7:1
Torque and speed are inversely proportional, and the gears we used provided torque to allow the DC motor to move the car, slowing the DC motor. The DC motor, with only speed, would not have enough power to move the 1kg weight, which is why torque was increased, decreasing the overall speed. Since we had four pairs of gears, there was also friction that hindered the speed of our racer. The carpet, an fixed variable, and the wheels also increased friction as our vehicle made its way to the finish line. We were able to reduce friction in this situation by only using 3 wheels instead of the typical four, and by testing out different wheels. Since we are also working with Lego, and a 1kg weight needs to be carried, the axles are susceptible to slight bending, which may cause more friction as they touch the top of the axle holder.

Concluding Reflection: If we had more time, we could have reduced more torque or an even better balance of speed and torque. I think our approach to building the lego car was a little sloppy at the beginning, which ate up a chunk of our time, and we weren't able to finish quick enough to realize that our racer wasn't going to be competitively fast. In addition, we could have rethought the inclination of our vehicle, which gave too much stress at the front wheel and the two lego pieces holding it to the main body. If we wanted it to remain inclined, we could have added extra support so that the lego pieces connecting the front wheel were sturdier and secured to the body. If given a lego or building project again, it would be a more clever approach to start off with as little torque as possible and then to increase. Although torque is necessary to allow the vehicle to move, finding a balance of it would have been more efficient than leaping in with a huge amount of it. The beginning obstacles we faced from connecting our racer together helped us move through later iterations at a more rapid pace near the end though, so although it was a more time-consuming process, it highlighted the importance to me of failing quicker to not only achieve a successful product, but to learn the necessary techniques to efficiently continue future iterations.