Embedded Files
About
About
Triton Robotics is an interdisciplinary project team that designs and builds robots to compete in the international DJI Robomaster tournament. Its members go through every step of the design process, from designing the robots in CAD, to manufacturing the parts and assembling them, to programming them to prepare for the competition. In doing so, the members gain valuable hands-on experience applying what they learn in a classroom, to a real-life problem. My experience with the club is where I gained most of my experience working on large-scale projects with a team.
As seen in the video, the Robomaster competition is very large in scale both based on the breadth of the expertise required, and the physical playing field itself. As a result of this large scale, I worked with smaller subteams on individual projects at a time. For my first three months on the team, I was on the Infantry subteam where I worked on redesigning the ammo box to make the turret more balanced. After working on this project, I then moved on to working on the Hero subteam for the next year-and-a-half where I started on redesigned the yaw mechanism for the turret and proceeded to work on the suspension, turret, ammo box, chassis, and feeding mechanisms. For my last six months on the team, I was on the administration board of the team and was the VP Internal where my job was to primarily run the recruitment for the team.
We competed in the North American RMUL competition in the summer of 2021 and received Third Prize from the competition.
To preface this page, I would like to say that I will be using a lot of jargon related to the competition and mechanical design, so if there is anything you have questions about feel free to reach out to me at my email that's listed on the bottom of the page!
Infantry
Infantry
My work on the Infantry robot was the least involved out of the other work I did for the team, but was a nice transition between high school robotics and Robomaster which was a lot more involved and had a much larger learning curve to get used to. For example, in high school robotics, most of our design happened as we went and had minimal pre-planning and minimal custom manufacturing as most of our robot parts were stock Tetrix parts. In Triton Robotics on the other hand, almost all of our parts were custom manufactured using 3D printing (both FDM and SLA printing), planar manufacturing using waterjets and aluminum sheets, and basic machining of metal tubing and metal rods. What's more is that in high school robotics, we only used CAD with the 3D printed parts we made, while with TR, all of our robots were completely designed in CAD before any manufacturing was started.
The requirements for the redesigned infantry ammo box were to be cheaper and easier to manufacture because the previous iteration was a single 3D printed box which was made with SLA, making it quite expensive to manufacture. So the new design used sheets of aluminum or acrylic that were cheap to manufacture and were put together with L-brackets that we had plenty of already. It was also slanted backward in order to shift the center of mass backward on the turret to help the balancing of the turret and help prevent the gimbal motor from overheating and potentially being damaged in competition.
The other part of the infantry I worked on was changing the design of the ammunition revolver to be more jam-resistant and more accessible for maintenance.
While these were two of my earliest designs for the team, they still ended up making it to the final rendition of the infantry robot for that year with only a few modifications.
Hero
Hero
The hero robot acts as a big brother to the infantry robot but has its own quirks and specifications that made it very unique compared to the infantry. The main difference is that it uses a 42 mm projectile launcher instead of a 17 mm launcher the infantry uses. What's more, is that you can add an additional 17 mm launcher to use in tandem with the 42 mm launcher. When I first started on the hero, there was no real design to start with and modify like there was for the infantry. Although there was a previous year's rendition of the hero, the only part of it that was reused was the launcher barrel and launcher mounting plates, which both ended up being heavily modified to meet our needs.
For the new design, we wanted to use a bottom-fed projectile launch mechanism in order to bring the center of mass of the entire robot down, to prevent it from tipping during sharp turns in a competition. The other benefit is that we can put the 17 mm launcher on top of the 42 mm launcher giving us twice the damage output potential. Since the 17 mm launcher uses a top-fed mechanism, Using a top-fed mechanism for the 42 mm launcher was not possible.
My first job while working on the hero was to design the yaw mechanism that would allow for the 42 mm projectiles to be fed through the center of rotation of the yaw, which required a large turntable for a bearing and was driven via chain and sprocket as opposed to a direct drive that was used on the previous version of the hero. The initial design for the yaw involved a stationary drive motor with sliding idlers that would be used to tension the chain.
This original design did not stick however because we had decided to make the entire robot modular to be able to easily take it apart for access to the internals and to make transportation easier. So instead of the idlers sliding around, we decided to make the entire front mounting plate be able to slide in and out to tension the chain. The plate would also have slots that would allow us to completely remove the front of the robot to access the electronics inside.
Once the yaw mechanism was completed, my next task was to make as much of the robot easily accessible as possible, and one addition was the mount for the front armor plate. I decided to mount the front armor plate to a hinge instead of the chassis in order to allow for the front armor plate to be rotated outwards to be able to further access the electronics on the inside. What's more, is that the 3D printed mount for the armor plate was made so that the back of it would be able to mount electronics onto for easy access as well.
Another aspect of the modular design was the ammunition box that would hold the 42 mm projectiles and funnel them into the serializer (the revolving mechanism another student made used to feed the projectiles up to the launcher). Similar to the front plate, the Hero ammo box would be mounted to the chassis using slots that would allow us to slide the box out and up and completely remove it by only loosening two bolts. Designing the ammo box and incorporating it with the rest of the parts of the robot was another one of my responsibilities.
My next project for the hero robot had to do with designing the overall chassis, which would have to accommodate the ammo box and the yaw mechanism and also have room to mount the suspension that had yet to be designed. The chassis would be assembled out of 3 mm aluminum sheets and aluminum square tubing and I ended up using simple rectangular segments attached together with shear plates and L-brackets. After running some FEA simulations, we realized that the design was more than enough to hold up the weight of the robot and would be able to withstand considerable impulse forces.
As an extension of the chassis, another aspect of the robot I worked on along with another team member was the suspension. Originally we were going to use the suspension that the infantry used and retrofit it onto the Hero by making some small changes in order to account for the higher weight of the hero as well as the larger physical size of the robot. Because the hero was quite a bit larger and heavier than the infantry, as well as issues with the cost of the infantry's suspension, we decided to make our own four-bar linkage suspension mechanism that would be able to hold the weight of the hero and would also be cheaper to manufacture since it would use the same 3mm aluminum plates that the rest of the robot used, so we wouldn't need to pay for additional manufacturing costs.
The team member I worked with made the base design, and I altered it to make it interface with the rest of the robot and fit on the chassis that was designed. I also performed some analysis on the suspension to see how strong the springs on the shock absorbers had to be to keep the robot from sinking too low and also so that the hero could survive a 1-meter drop test that we would have had to do as part of the inspection at the tournament.
The biggest issue we ended up having with this design was that the robot would rock back and forth when it would accelerate and decelerate quickly, which would impact our aim at the competition. Videos from the completed prototype are shown below.
Epilepsy Warning: Video is in slow motion under flickering lights.
One of the last crucial aspects of the Hero that was left was the "trigger" mechanism. The launcher of the robot worked by squeezing the 42 mm projectiles between two large flywheels. For the infantry robot, the projectiles were fed into the flywheels with the ammo revolver that was shown above. There were multiple issues preventing us from using the ammo revolver on the hero to accomplish this same task. The first issue was that the fire rate was capped at only 1 projectile per 5 seconds which was much much lower than the fire rate for the 17 mm projectile launcher, so we needed a much more controlled firing mechanism. The other issue was since we were using the bottom-fed system, the projectile would be fed a lot more slowly into the launcher which would slow down the exit velocity of the projectiles. In order to counteract both of these issues, I designed a double-action trigger mechanism that would push a single projectile through the flywheels with a cam-shaped design that is visualized below. The trigger mechanism would use a limit switch placed in the barrel of the 42 mm launcher to detect if there is a projectile in the barrel. If there is a projectile, it would tell the ammo revolver to stop spinning and then launched the projectile when it was ready.
The trigger mechanism ended up working as we had hoped; however, issues with the ammo revolver prevented the trigger from being used properly at the competition.
The trigger cam would rotate backwards to a specified location.
The ammo revolver would then push a projectile into the barrel until it would activate the limit switch on the wall closest to you (not pictured).
When ready to fire it would then push the projectile forward through the flywheels with an initial velocity to give an extra boost to the exit velocity (located in the cutouts on the left side of the barrel). The trigger cam would also prevent another projectile from entering the barrel and jamming the mechanism.
Then it would rotate back to its initial position ready for another projectile.
A part of the robot that I also worked on that did not make it to the final prototype was the integration of the 17 mm launcher to the turret for the hero. For this design, the 17 mm launcher from the infantry turret was taken and modified to fit onto the turret of the hero. For the modifications, the ammo box holder and mounting plates were altered significantly to have the infantry turret interface with the hero turret without mechanical parts and wires running into each other.
A derivative of this project was going back through the rest of the hero design to make sure that all of the bolt locations had proper clearances and making sure they did not run into other parts accidentally.
Takeaways
Takeaways
There's much I learned through Triton Robotics that had to do with the practical application of engineering theory. I learned many aspects of SolidWorks I would not have known otherwise, like configurations, large assembly design, using the SolidWorks toolbox, and other aspects of top-down assembly and multi-body part design. Another aspect of engineering I learned about was machining tolerances and designing them into parts so that we wouldn't have to machine parts again if their tolerances were too loose or tight. This was an especially important skill when designing 3D printed parts as they would usually have the highest amount of machining inaccuracy.
I also gained much experience with the logistical side of design and making sure all of our cad models were made in a way that allowed for easy changes later on and would not break if a single dimension was changed. There were many parts I worked on early on in TR that I regretted not putting the extra time into as I was making them because I would later have to go back and fix every error when a slight change was made to those parts. Additionally, making sure all sourced parts were well documented and added to the BOM was important to prevent having to backtrack when ordering parts.
One of the biggest takeaways had to be learning to properly work and communicate with team members working on the same robot. It was important to communicate design constraints and progress to teammates to make sure that everyone was on the same page so we wouldn't have to waste time redoing work that was done because of miscommunication.
Page updated
Google Sites
Report abuse