SciBorg Day 1

If anyone has taken CS111, they may have been struck by the similarity between the Pico-software and the Buggle lessons that we first learn and program with.  I find that this type of very simple chain of commands is often used to teach children and get them interested in computer science and engineering.  It probably should not have been a surprise that the Pico-software was made for children.  Still, the software teaches important basic skills that perhaps children understand better than I do.

A. The first couple of tasks were obviously very simple.  We struggled slightly with making it spin right and left because we couldn't get the blocks to work how we wanted them to (my fundamental struggle in CS as well).  We understood that we had to get one wheel to move forward while the other moved backward.  I became hooked on the idea of using an if else statement but the blocks wouldn't allow it.  Eventually we figured out that we could instruct the car to move one wheel forward, move the other wheel backward, and then move both at the same time.  With this framework, we were able to make the SciBorg bear-right or left by setting the power of the opposite wheel to be slightly less than the other.

B. The most difficult part of this task was perhaps identifying how to use counta with the blocks (another fundamental CS struggle).  After we found the piece, everything else was fairly easy with the knowledge that we had gathered from the previous tasks.  Our borg was able to travel back to its original starting place after the code had run its course.  We did notice that using the counta code made our borg turn more to the left while the countb code was slightly straighter.

All and all, I'm excited to see how we'll be using these SciBorgs in the next week or so.  I've been interested in sensors and am excited to see how we'll be able to use them.  I hope we'll be able to use other types of code than this Pico-software but perhaps that will be much, much later...

Lego Racer

Prompt:
Build and design a vehicle with a single motor, powered by a PicoCricket, that can carry a 1.0 kg weight as fast as possible on a 4 meter course.

Engineering:
We began with learning about how torque and speed are inversely related: the faster the motor spins, the less torque it has which means the car won't be able to move when placed on the ground due to the insufficient amount of torque.  However, if we increase the torque but decrease the speed, it may mean that we will build a car that is too slow to win the race.  Friction was an additional factor that we had to take into consideration since it could also slow down our car.  Possible sources of friction included the wheels against the carpet and the number of gears meshing against each other.

Ultimately, the key to succeeding was to create the perfect gear reduction that would balance enough speed and torque that would move the car as quickly as possible without creating too much friction.  In order to create an optimal gear reduction by increasing torque and decreasing speed, the most logical decision was to use small gears to move larger gears.  This would cause the smaller gears to turn faster while the larger gears spun at a slower rate.  With this goal in mind, we set off to iterate several different gear reductions before we designed the car.

First Iteration:
After playing around with the gears to determine what meshed together better and observe what was physically possible, we decided that our baseline gear reduction would be the one provided in our packet.  This consisted of using three different pairs of 8-tooth gears to turn 24-tooth gears and yielded a gear reduction of 1:27.  We decided to add on the motor to see how it would work and made the newbie mistake of attaching the motor's 8-tooth gear to the gear train's 24-tooth gear.  This effectively reduced the gear reduction to 9:1 which produces an insufficient amount of torque.

We eventually figured it out

Second Iteration:
At this point we realized that using an 8-tooth gear to turn a 40-tooth gear provided the maximum amount of torque (1:5) and started to calculate a number of different ratios that would produce the most torque.  Modeling our design after our first iteration, we connected three pairs of 8-tooth gears with 40-tooth gears for a gear reduction of 1:125.   Getting this gear chain to run took an immense amount of power with just our hands but once we could get it to spin, the 8-tooth gear spun impressively fast while the 40-tooth gear spun extraordinarily slow.  Unfortunately we couldn't get this gear train to work reliably or at all but in the very least we were starting to build an understanding of what our gear reduction range was.

Calculated Iterations:

To make a smaller gear reduction and narrow our range down, we began to mix the combinations of gears.  At this point, we determined that it would be smarter to calculate the gear reductions first before building anything.  The following is a list of our varying calculations:

(8/40)(8/24)(8/40) = 1:75

(8/40)(8/40)(8/16) = 1:50

(16/8)(8/24)(8/40) = 2:15

(8/24)(8/40) = 1:15

(16/24)(8/40) = 2:15

(16/40)(8/40) = 2:25

(16/24)(8/24)(8/24)(8/24) = 2:81

We tested the first two and found that we couldn't get the gear train to move with the gear reduction of 1:75 but could move it from 1:50.  We then found that the 2:15 gear train lacked the torque to move and calculated a number of hypothetical situations thereafter to try to get it back within our range of 1:27 and 1:50.  Out of time, we quit for the day and sat on the ideas for a while.

Gear Worm Iteration:

It was at this point that I was partnered with Sam since both of our partners dropped from the class (they were both wonderful people).  Sam had been working on trying to get her worm gear to work.  We spent the majority of that class piecing together her worm gear with my previous gear train (8/24)(8/24)(8/40) = (1:45)

However, a number of problems began to occur.  For one, we couldn't get the gear reduction down because most of our time was consumed trying to get the structure holding the gears to work and gears to fit together.  Another problem was that the worm gear and connecting gear produced too much friction to the point where the gear was spinning at a much slower rate than it should.  After many frustrating attempts trying to get the structure of the car to allow the gears to mesh together and trying to reduce the friction between gears, we abandoned this design for my original and much simpler design.

Final Iteration:

We decided to go back to the very basics.  We were aware that using an 8-tooth gear with a 40-tooth gear provided the most efficient gear reduction.  We were also aware that 1:27 was our minimum baseline.  Therefore, we decided to use a pair of 8-tooth gears with 40-tooth gears to create a ratio of 1:25.  To set our car apart from others and reduce the amount of friction, we also decided to build a lightweight frame.

The result was a rather precarious car that ran backwards but also one that had the minimum amount of torque needed to move the car while maximizing speed and minimizing friction.  Our resulting time was 10.8 seconds which put us in first place for the class race.

Reflection and further iterations:

I have always known that rebuilding and testing was an inefficient method of trying to perfect a project.  However, I learn by observing so I could never fully grasp certain scientific concepts that required me to believe in the principles that they stated.  There is no doubt in my mind that they are correct; I just couldn't comprehend how it was possible because I was unable to observe it.  This is perhaps the first project I have encountered in which I relied more upon my calculations than sight.  I realized at some point that there was no way to distinguish between different gear ratios with my eyes which meant I had to trust that my calculations would reflect reality.  It seems like a simple logical line of thought but this project was just my epiphany into that world.

I think if I had been given more time, I would have tried to redesign the frame to be even lighter but sturdier.  Even though we tried to use the bare minimum amount of pieces to piece our car together, there are still some pieces that could have been redesigned and allow for a lighter car.  I would also be interested in trying to make an even slightly smaller gear reduction to try to see what is possible.

Sienko Lecture

I have stated previously but wish to reiterate my interest in user design.  I'm originally a Psychology major and so, it should probably come to no one's surprise that I'm highly interested in this field.  However, I also enjoy building, designing, or creating something so it should also come to no one's surprise that I also enjoy engineering.  I highlight this point because after I attended a talk by Kathleen Sienko on her experiences designing medical devices for those in global health, I was struck by how complicated  and careful the design process must be for her and others in her profession.

From my very limited experience, the design process usually starts with interviewing the client to understand their needs, drafting prototypes, perhaps gathering more feedback from the client, producing several more prototypes, perhaps talking with the client again, and repeating the cycle until a final product is created.  Although her work generally follow the same model, she highlighted the complex issue of her work in which they are only able to test their prototypes once when they visit their global clients since such trips are so expensive. In essence, her prototypes must be finished by the time they reach their clients.  However, being able to fulfill that expectation is almost impossible since there will always be a need to create another prototype to address issues that were overlooked or not foreseen before.

The complexity and pressure that Sienko and her team face when they design their products for their clients represents the barrier that many others face when endeavoring to create medical devices for global health.  However, I think their process makes them better engineers and teaches them valuable concepts that I hope can be used to help both their clients and other engineers learn how to make their design process more efficient.

Well Windlass: The Build

Back in middle school, I was part of a Science Olympiad event called Tower.  We were required to build a tower about a foot high that could support as much weight as possible.  However, the tie breaker for the event was that the lightest tower would win.  Hence, this resulted in some very thin wooden sticks being used to build fragile towers that supported pounds of sand.  So when we were told to build a windlass that had a similar functionality (minus the weight limit), I was a bit excited to get my hands back into building something.

Sketches

At first it seemed like a logical step to make the base square and the sides trapazoids.  However, since we were only allowed to use 500 cm2 of material, my partner and I had to iterate our designs over and over again to find the perfect shape that would be able to bear the weight.  We finally decided upon hollowed out triangles for the sides to solve this problem.

A further but small problem was the fact that we wanted to speed up the process of pulling up the bottle without turning the crack a few hundred times.  However, designing a way to increase the surface area of the rod that the string wrapped around while allowing the sides to turn proved to be quite a challenge.  In the end, we decided up on a 2.5cm rectangle that would span across the middle of the rod to increase the surface area that each turn of the crank would pull in.

An addition challenge with our design was how to fasten each part.  In my previous post, I listed a few different methods on how I planned on attaching various things. However, after examining the physical prototypes and the sketched out designs on the paper, it appeared that most of my plans were either not feasible or excessive.

Designing

Although I was initially frustrated with Solidworks, I have to admit that the program is starting to grow on me as I gain a better control over the software.  I'm still somewhat jaded that I can't type in the commands and am forced to search for each one in the toolbar but once I learn what the shortcuts are, I predict that I'll be more favorable to it.  I think the hardest part of designing this time was simply creating all the notches and hinges since the measurements kept changing on us.  We went through so many iterations trying to adjust the widths and heights of everything to account for the differences in Delrin thicknesses.

Engineering Analysis

We were initially concerned that the structure would collapse within itself since the weight of the bottle will pull the structure down and ultimately cause it to push outwards.  Hence, we designed support beams in the middle of the triangles to keep them from moving apart from each other.  The triangle sides were also thoughtfully designed such that the force would not cause one side of the structure to fail (ie collapse inwards as a square might do).  The equation variable was the most under our control was most likely the force of the bottle on the structure.  The 2.5cm rectangle that the string wrapped itself around distributed the force a little more equally in relation to the structure since the force would then not be concentrated in a single spot and cause a structural failure. 

Final thoughts:

In the end, our structure was probably not the most structurally sound thing on the face of the earth.  I believe this is most likely due to our obsession with the material constraints.  If I could iterate the design again, I probably would have made the triangles solid instead of hollowed out to improve its sturdiness.  Additionally, I wish there could have been some way to make the notches better but this was a factor controlled by the laser cutter.

Bottle Opener: Dragon Chronicles

Design:
The main concerns I had about the material we were given to make our bottle openers, Delrin, were that the force of prying off the bottle cap would cause the bottle opener to break and that since we were only laser cutting the material out, I had to design the opener in such a way to fit under the rim of the cap and over the top of the cap.  Given these constraints, the most logical strategy was to use the material length-wise rather than width-wise since the thickest piece we could use was only 1/4", a thickness that would most certainly break under any sort of pressure.  However, I did attempt to design some openers that used the width rather than the length of the piece to fully explore all options.

My next steps were to figure out the optimal method of taking the cap off.  It seemed readily apparent to me that there needed to be a leveraging part that held down the bottle cap while the bottom of the cap was pushed upward to create a malfunction in the cap.  Since this part of the design seemed set, I set out to figure out a design for the bottle cap which eventually led to the concept of dragons or dinosaurs.

Between my partner and I, we originally created a dragon that was a simple, horizontal opener.  However, we decided to curve the dragon's back in order to provide the user with a more ergonomic fit and add an additional artistic element to it.  We were slightly concerned that the curve would interfere with the ability of the user to open the bottle if it were too curved and the mouth/bottle opener function could not fit the cap.  However, it became clear that it wasn't an issue for our design.

Physics:
We were briefly introduced to the physics of levers.  It became apparent that there were certain factors that we could control - length and material geometry.  We also recognized that there were other factors that we could not - material strength, how the material moves under stress, and force since there was a minimal amount of force that needed to be applied to get the cap off.  As such, we tried to maximize the strength of the material by using its length rather than width and keeping the length relatively short.

3D Software:
I took an IAP course at MIT in January on product design that taught us how to use Rhino to build our products.  At first I wasn't completely sold on the program - there's some learning curve to figuring out where everything is in the program first.  However, my favorite function of it became the ability to type in whatever command I wanted and its hints as to what it needed to complete the function.  As someone who's interested in user interface as well, I thought that this feature skillfully allowed more advanced uses to quickly access whatever they wanted without having to search for it in the toolbars or drag their mouse to the other side to complete the action.  The silent prompts for what information it needed to complete the function, however, catered to more intermediate and new users such as myself and ultimately helped me figure out how to use many of the different functions.
As of now, I'm not completely sold on SolidWorks.  I think I'm just not familiar enough with it to like it but I do have a friend who has promised to help me learn how to use it.  Apparently there are keyboard shortcuts that will probably mimic the type-in function so I am excited to learn how to use it in a more efficient way.

Build:
Our first laser cut of our design was surprisingly satisfying.  My partner and I were pleased with how everything came out (the laser cutter accidentally scorching the body of it was actually a bonus that we weren't expecting).  However, I think our design came out slightly smaller than what we thought it would be.  More specifically, the body of the dragon was much smaller than what we were expecting but the mouth piece/the bottle opener portion of it was accurate to our initial measurements.  Because he was so tiny but adorably fierce, we decided to name him Smaug.

 Smaug in action

Perfect fit!

The mouth dimensions worked perfectly since the bottom portion hooked under the bottle cap perfectly while the top portion of the snout provided enough force downward to allow the cap to bend.  However, Smaug's snout snapped off after a few successful attempts at opening caps.  I later determined that the triangular shape of the snout distributed the force rather poorly in that area since it placed a huge strain on the peak of the nose and ultimately caused it to snap.

Iteration #1:

To address that problem, we created a rounded snout to distribute the force.  We decided to name this iteration Toothless since he was more adorable than fierce.

However, the dimensions of the mouth and entire bottle opener were somehow altered in our iteration which yielded an imperfect fit in the mouth and twice as large body.  He ironically started losing his bottom jaw because the mouth was actually slightly too small to fit the bottle cap comfortably between its jaws.  However, the larger size yielded a more ergonomic and comfortable fit.

Iteration #2:

We carefully adjusted the mouth measurements to accurately reflect Smaug's snug fit and inadvertently made the opener about the same size as Smaug .  As a result, Special was born and named because the laser cutter had problems and melted its back.

Iteration #3?

Although Special works pretty well, if I could reiterate it again, I would make the bottom jaw a little thicker to provide a more stable-feeling opener.  I would have also increased the body size to that of Toothless to make it more ergonomic

Fastening and Attaching

We learned of three different fastening and attaching methods in class a few days ago.  The first was through a method called heat stake.  Essentially, a plastic stud piece that is fitted through a plastic ring is melted down by the heat stake machine into what appears to be a button.  Through our group's trial and errors, we learned that giving the pieces ample time to cool and prying the final piece off the machine slowly yielded the best results.  I think this method yields a very strong and sturdy attachment for two pieces.  However, there may be some difficulties designing the pieces such that this method can be used.  The added benefit in its resilient immobility consequently limits it from being used for hinges.  I think in the next windlass project that we are about to design, this piece will most likely be used to build a solid base but not for any of the moving parts.  Just as a thought, I might try to combine this method with another fastening method below to make an even stronger base.

The next method we used was the bushings.  I find this method the most difficult since it requires such precise measurements on behalf of the designer and the slightest thousandths of a millimeter makes a difference.  We noted that three different rings yielded three different fits.  A diameter of 0.2600 yielded a loose fit which allowed the pieces to spin around each other, 0.2545 yielded a snug fit which could be used to isolate sections in a piece if need be, and 0.2500 yielded a tight fit which would be used as an end piece.  I was amazed by how the smallest difference in diameter could yield such drastic differences in how the pieces moved together.  I am definitely daunted by the possibility that I will have to use this method, but it will ultimately be a good experience in the end.  I envision myself using this method to work on the turning pieces of the well but I may use the next method instead since I find it so interesting.

We next learned of using piano wires as hinges.  To be honest, I was a little apprehensive about using such a powerful machine despite the fact that my engineering friends use these machines and even more dangerous ones all the time.  However, upon learning how to use them, I was surprisingly excited to know how to use one and this might be my favorite fastening method if only because it's rather fun to use the machine.  In this method, we used a drill bit to drill a hole into a piece of plastic and the machine to again push the wire through the hole.  Again, the nuances of just a thousandths of a millimeter makes a huge difference in how the pieces move but I find it to be much easier than bushings since the drill bit is really the only part that determines how loose the hinge will be.  I think using piano wire instead of bushings will make my life easier (and quite honestly more entertaining) when I design the well lass so there will be a clear bias in my design (unless the bushings work better of course).

Our last fastening method, and one that I could use in conjunction with the heat stake method, was designing two separate pieces in Solidworks to fit together.  The idea is that the dimensions will be so exact, to the thousandths of measurements again, such that trying to separate the pieces will be almost impossible without the help of pliers.  I think this method would best be used to design the base of the well lass but it also means that it cannot be used in any of the hinge designs.  Since the laser cutting machine was malfunctioning that day, we were unable to make the pieces.  However, I would be curious to attempt this method later when I am designing the well lass.