After agreeing on a final design, we started making the pieces on SolidWorks. Here are some pictures of some important parts:
This is the front and back piece of the walking module. It has two "legs" that will help the robot balance. The height of this piece determines the height of the robot. We decided to connect the side pieces on to the front piece with heat staking. This process consists of melting the plastic. In order for it to work, we had to make this square holes where the side pieces would connect on to the end pieces. As you can see in the picture of the side piece below, there are small bits of plastic sticking out.They are intended to fit into and stick out of the holes in the front piece. Then, when using the heat staking machine, we would melt the extra plastic sticking out of the holes so that it would secure the two pieces together. The piece below is the bottom. We forgot to add this piece in our original design and then we realized that we had to include the motor in our design, because without it, it would simply fall out. We also made sure that the motion module would be big enough (but in width and length) for the motor to fit in. A The bottom piece is designed to fit tightly through the "legs" protruding from the front piece. This would allow us to connect the pieces easily without worrying about heat staking or other methods of attaching two pieces of plastic. This pieces below are what we replaced the gears in our previous design with. We made the circle in the middle a tight fit, so that the circle would rotate with the rod. We made the circle on the outside a loose fit so that the bar could rotate freely. The bar connects two of these circle pieces. The bar has corresponding holes where a small rod would connect it with the circles. This circle piece is slightly different from the other one, in that the circle in the middle is smaller. The motor that we are using has a rotating metal rod that would power the whole robot. Since this is what we are attaching our circles to, we had to adjust the size of the middle hole so that it would tightly fit with the metal rod.
We decided that we would work on the "box ski dude" idea instead of the "frogbot" because we didn't really think the elastic concept would work properly. We also were more interested in a motion that was most similar to walking.
Like we previously found out, we had to make a couple important adjustments to our original design. 1. We had to put the legs off center from the gear and we had to add extra legs to keep the robot balanced.
We were inspired by a lego robot that was in the engineering room. It has a motor rotating an axle with two small gears attached. These small gears rotate a four larger gears where the legs are attached. The legs are similar to our design because, they are not individual legs, but skids. Another important aspect of this lego robot that is the hidden skids that keep the robot balanced.
Here is the picture of the lego robot:
And this is a picture of skids that keep the robot in balance:
We decided to incorporate this by having two front "legs" extend from the front of the robot. They would look like large teeth, making the robot look more like a walrus. But this would be a creative solution to issue of balance.
Another issue that we came up with was whether we wanted to make gears on SolidWorks or to find another way to create the same motion without gears. The problem with plastic gears is that they don't work properly. So it would be better to figure out a way just to make circles that would rotate simultaneously. In order to come up with a solution we played with legos and Professor Turbak helped us. This was very tricky because when we attached a bar to two circle and rotated one circle, the bar did not follow a constant rotating motion because it would often rotate the other way. This occurred because the motion wasn't constrained enough. We found, however, that if you attached another set of circle pieces with a bar connecting them to the original circles, giving you a total of four circles, the motion was constrained and the bars followed a circular motion. However, there was one condition: the bars had to be less than 180 degrees out of phase. So we made the separation between the bars as close as possible to 180 degrees. Hopefully that would be enough to imitate the act of walking. Here is a picture of the lego motion module that we will base our robot on:
We made a couple models for the frog concept. We first made models that followed our original design, in that a motor would rotate the robot's legs with feet that would lift up the body. We made two models for this concept:
The problem with these two models was that they were two similar to a wheel-like motion and that we weren't sure that the feet would be able to push the rest of the body up. We would definitely need to have a powerful motor for this robot. With these models would also have to change the shape of the body so that it doesn't obstruct the motion but just slides as the legs are pushing off the ground. The other models focus on different ways to get a hopping motion without relying on a circular motion.
This is the first one we came up with. Its legs are attached to the body but they are free to rotate. The leg has a foot that would push off the ground and make the frog jump. To achieve this, we need a motor that pushes the feet at the same time towards the ground, until the whole body is lifted and then brings them back to their original position. Here is a picture of the foam model: The problem with this model is that the motor moves at the same rate and it would preferable for it to be more powerful when it pushes the legs downward. Professor Berg told us incorporate rubber bands in order to get this "burst of energy" that is needed for a hopping motion. He helped us to come up with this motion module with a rubber band. So as you rotate the oblong piece, it stretches the rubber band and at the point where the maximum distance has been reached, the rubber band helps the bar snap back to its original place. It is this snapping motion that is important for us and we decided to use it to create a hopping motion. The frog's leg could be the bar that snaps back and after the motor pushes it down towards the floor and quickly moves back towards the body, it would make the frog leap forward. The motor needed for this motion would need to be very powerful so that it can resist against the rubber band.
After coming up with five initial concepts, we decided to pick two to focus on. We really liked the frog idea because we wanted to find a way to create a "hopping" motion. We also thought that the "box ski dude" would be fun to work with and that it was our only concept that really imitated the act of walking. We started making foam models for these two ideas to better grasp their motions and complexity.
Here is our foam model of the "box ski dude":
After making the model, we realized that we had to make some adjustments to the design. There is the problem of balance and our foam model lacks stability. Because it only has two legs, the robot would just fall as one leg moved forward. This shows us that we need to add elements that would support the robot. For example, we could add two more legs (not attached to the motors) that would keep robot stable.
Another adjustment we have to make is where we have to attach the legs according to the gear. On the model, the legs are attached on the same axle of the motor and it is concentric to the gear. To ensure that the robot actually moves, we have to connect the legs off center to the gear. This would lift the legs off the ground and move them forward without having them slide on the ground. If we decide to make this adjustment, then the supporting elements would be very important, because as the robot 's weight would shift left to right, it would lose its balance.
This challenge consists of making a device that moves from one place to another by some means other than wheels. The device will be powered by a continuous DC motor or by a hobby servo motor. Connections can be made with screws and threaded standoffs , press-fit music wire, or a “heat staking” process.
We began this project by analyzing the motions of various animals. We were inspired by previous projects that were made in this course and by Youtube videos of "walking" robots. We found the movements of snakes, frogs, and monkeys particularly interesting. We quickly started developing ideas for different kinds robots. Many of our ideas were very complex and required more than two motors, for example with our snake-like robot. We also had to remember to avoid using a wheel-like motion and we also wanted to create a robot with a unique motion, for example like a springing motion. This was a fun process, but also a difficult one because even if we could visualize a specific motion, it was hard trying to design a robot that could do it. This is why we played with legos, to experiment with the different kind of motions we could create.
After playing with legos and searching for ideas on the web, we came up with 5 main concepts:
The Ostrich
This concept consists of two separate movements. The "neck" rotates down to the ground and the "head" shifts right to left as the head reaches the ground. The head's motion was inspired by the motion module presented in the last challenge. A picture of it is attached below. This is a useful motion that would enable the head to extend and lengthen the neck and it would be powered by a dc motor. The whole robot would require two motors, one to move the neck down and another to move the head. They would have to be synchronized so that as the neck reaches the floor, the head is extended farthest right. Then, once the beak (attached to the head) touches the floor, the motor powering the motion module would bring the head back in to the left. This would cause a dragging motion caused by the beak unto the rest of the body. By repeating this motion, the robot would move and appear like an ostrich continuously planting its head unto the ground, hence its name.
The Ski Walking Box Dude
For this concept, we tried to create a robot that would imitate the motion of actually walking. The main challenge with this is balance. Separate legs would have a hard time keeping the robot balanced, so we decided that flat platforms (like skis) would keep the robot stable. This robot would have a total six legs and they would be attached two platforms. The robot would walk forward, one leg at a time. Each set of legs would be powered by a servo motor which would create a shuffling motion. This would be a complex design and would require a lot of gear, but the final result should look pretty cool and work very well.
The Inchworm
Like with the The Ostrich, for this concept we were inspired by a motion module. This motion module was made of legos and it was made up of many rod-like pieces attached to each other so that whenthey rotate, the whole module would either contract or extend. This reminded us of an inchworm and we designed a robot where a servo motor would extend the front of the robot (where the head would be located), so that as the head moved forward, the rest of the body followed. The head would have a beak-like piece attached to it so that it could better grip unto to the floor. The rest of the "legs" would be attached to platforms, to better slide unto the floor. This design would look pretty neat, but the only downside was that a previous group had designed a inchworm-like robot and we wanted to create something completely unique.
The Jumping Frog
For this concept we were inspired by a youtube video. Here is the link: http://www.youtube.com/watch?v=_o-bvoHt7p4. We found this motion very interesting because even though this motion was entirely rotational, it was able to create a "hopping" motion. This was also a relatively simple concept because it only required a DC motor that would rotate an axle where the legs would be attached. The legs would be made of two parts. It is important for the legs to located far back so they would be able to propel the rest of the body forward. Our only concern with this design is that it was too similar to a wheel-like motion, but we were definitely intrigued with motion
Modified R2D2
This concept consisted of a motion powered by a DC motor where R2D2's front legs would come forward and its last leg would follow up. This design would be really fun especially to make a realistic R2D2. It would allow us to learn new laser cutting techniques, such as engraving. And since this robot would be relatively thin, we would use new connecting techniques, such as wire pressing and heat staking. The whole downside for this concept is that it is too similar to the "inchworm" movement from the previous robot that served as an example. As we have seen, it is an effective motion, but it would better if we could come up with a completgely unique one.
After handing in our motion module, our professors reviewed it and gave us some advice to make it better. One of the things they said was that we could have used less delrin to make our motion module. For example, the base could have been a lot smaller. Instead of having a large flat surface, we could have used two stands to hold up the module. The stands would have used half the amount of delrin that was needed for our original flat base. Another way to save delrin was to change the top piece. Instead of having the top and the side pieces fit loosely, we could have connected the two pieces with a tab-like lip. Not only would this save plastic, it would also make the motion module more stable. With this "tab-like" connection, the sides wouldn't slide along the top piece. So we remade the top piece and this is what it looked like:
In fact the motion module was more stable and the pieces just fit nicely in their proper place. It also gave the module a cleaner look.
Another small change we made was on the bridges. Even though we already made the hole inside the bridges smaller, they were still a bit large (in the width). By making them smaller, it would stop the bridges from moving sideways and just move up and down. We fixed the old bridges on solidworks and printed the new pieces out. This is what they looked like:
Even though this was a small dimension change, it did make a significant change in the way the drawbridges moved. Now our motion module was effective and efficient.
In order for our motion module to work we made many adjustments. We adjusted the dimension of the bridges, the middle circle, and of the circle at the end of the rod that moves vertically and that comes into contact with the middle circle.
The bridges: The problem with the bridges was that they fell off the circle platform once the rod pushed them up. So we made the bridges longer and wider. We made the bridges longer by .33 inches and wider by .1 inches. This would make the bridges more stable and prevent them from falling off the circle platform. We also made the openings within the bridges smaller by .25 inches so that they would not move too much.
Here is a picture of the new bridge design (bottom) with the old bridge design (top): And this is a picture of the bridges on the motion module: The middle cirlce: The problem with the middle circle was that it did not yield a large enough range of motion. So we made its diameter from 1 inch to 1.5 inch. This was very effective in raising the bridges higher.
The other circle: The problem with this circle was that it was too small of an area for the middle circle to push off of, making it hard to rotate the rod smoothly. So we replaced this piece with a circle of a new diameter of 1.4 inches instead of 1 inch.
This is a picture of the new circle pieces:After all our adjustments, our motion module worked a lot better. The bridges didn't fall off the platform and they could be raised higher than with the previous design. It was also easier to rotate the rod, making the motion module much more efficient. Here is a picture of the final product: This is a video of our motion module:
After printing out all the pieces from a delrin plastic sheet, we assembled our motion module. We quickly found that the holes for the rods in the circle pieces were too big. The middle circle piece (the one where the hole is off center) did not stay on the rod and it was very important that it would be tight enough not move in order to push the other rod upwards. It was also important for the other circles to fit in tightly with the rods because otherwise they would simply fall off. Because of that, the motion module could not work. Therefore, we changed the diameter of the new circles to .242 inches, which was the perfect dimension for the circles to stay on the rods.
After printing these parts out, we assembled our improved motion module. This is what it looked like:
With the new pieces, the motion module worked and the rotating motion was successfully transferred into a translational motion. However, the drawbridge aspect did not work as well as we expected. The bridges just fell off the circle platform that is attached to the rod that moves vertically. The problem was that the drawbridge or the platform was too small. This is a picture of what happened: Now, we have to figure how to fix this aspect of the motion module...
We finally finished making our design on SolidWorks. It took a lot of time making sure that the pieces were the right dimensions and to assemble them correctly. One of our concerns when working with SolidWorks was what dimension we should make the holes for the teflon rods to pass through. We didn't want them to be too big so that the circles would fall off, but we also wanted the rods to actually fit in the holes. It's such a small adjustment in the diameter that we really didn't what to expect. In the end we decided to make the diameter of the hole the exact diameter of the teflon rod, which is .25 inches. This is what our design looked like in the end:
In order to print out all our pieces for this assembly, we created a drawing on SolidWorks with all our parts. This is what it looked like:The printing process was not as easy as we expected. There was a problem with the drawing, (some of the outlines of certain parts were of a different shade of red than usual), which made the laser cut one of pieces incorrectly. So we spent quite a bit of time fixing the problem and adjusting the parts on our drawing so that we would use the least amount of plastic possible. We were finally able to print and all we had to do was assemble the parts and see if our design needed any adjustments. Here's a picture of the laser-cutting process:
This week's challenge is to build a mechanism that converts rotational motion to some other kind of motion. It will be constructed out of Delrin sheet and teflon rods.
My partner and I were inspired by a motion module that our professors showed us in class. By rotating a teflon rod in this module, another part of the mechanism moves up and down. It's a relatively simple motion, so we wanted to elaborate this design by adding another element and increasing its size. In this module only the rod moves up and down and we wanted something else to move as a result of that. So we came up with this idea of a drawbridge. As the rod moves upwards, it would push up two separate plastic boards.
Here is a sketch of our design: In fact, our design is very similar to the motion module shown by the professors. One our concerns was how to prevent the boards of the "drawbridge" from falling down as the rod would push them up. We first considered putting in place a rod that would stop the board from falling, but that would just add to the complexity of the module. Our other idea was to make a hole in one of the plastic boards where a piece a plastic would prevent it from falling off completely. Here is a sketch of that: After coming up with a general idea we made a foam board design with dowel rods. This really helped us realize which dimensions we had to adjust in order for the module to work efficiently. It also made us pay attention to how we would attach the plastic boards together. Here's a picture of our foam model: As you can see, the model doesn't function very well. The sides of the model are too small, therefore it didn't give enough room for the rotating circle to push up the vertical rod. So we recalculated the dimensions of the model and the location of where the rod should be. Making this model really showed us that careful, precise measurements had to be made. Here is a close up of the "draw bridge": For the "draw bridge", we have to make sure that the opening on the board to be long enough so that it would give it enough freedom to move, without falling of course.
After building the foam model and taking in consideration all the adjustments we had to make, we started transferring our design unto SolidWorks. The nice thing about SolidWorks is that you can make changes very and visualize your design very easily. However, since we are still learning how to use SolidWorks, it did take us a lot of time to construct our design. We learned how to create an assembly and it was a bit tricky to connect the pieces together. In order to do that, you have to create relationships between two pieces. For example, you can make certain edges coincide or make two faces parallel. We figured out that there is usually a specific order when forming these relationships and in doing so, it allows you to move a piece in a certain way. After making the general structure, Alex and I got the hang of making assemblies and now all we had to include was the rods, bushings and "drawbridges". Here is what we ended up with at the end of the day:
This is the front and back piece of the walking module. It has two "legs" that will help the robot balance. The height of this piece determines the height of the robot. We decided to connect the side pieces on to the front piece with heat staking. This process consists of melting the plastic. In order for it to work, we had to make this square holes where the side pieces would connect on to the end pieces. As you can see in the picture of the side piece below, there are small bits of plastic sticking out.They are intended to fit into and stick out of the holes in the front piece. Then, when using the heat staking machine, we would melt the extra plastic sticking out of the holes so that it would secure the two pieces together.
The piece below is the bottom. We forgot to add this piece in our original design and then we realized that we had to include the motor in our design, because without it, it would simply fall out. We also made sure that the motion module would be big enough (but in width and length) for the motor to fit in. A The bottom piece is designed to fit tightly through the "legs" protruding from the front piece. This would allow us to connect the pieces easily without worrying about heat staking or other methods of attaching two pieces of plastic. 
This pieces below are what we replaced the gears in our previous design with. We made the circle in the middle a tight fit, so that the circle would rotate with the rod. We made the circle on the outside a loose fit so that the bar could rotate freely. The bar connects two of these circle pieces. The bar has corresponding holes where a small rod would connect it with the circles.
This circle piece is slightly different from the other one, in that the circle in the middle is smaller. The motor that we are using has a rotating metal rod that would power the whole robot. Since this is what we are attaching our circles to, we had to adjust the size of the middle hole so that it would tightly fit with the metal rod.
