BIT 1.36 V3 Draft Design

    Though the second beetleweight robot I built has not been tested in any events, I have started to design the third version. The design of the previous two robots are relatively conservative. Frames are made out of aluminum sheets, weapons use belts to transmit power and the wheel guards are made out of TPU. But nowadays more and more robots started to use hub motor weapons, use one-piece aluminum frames and run on a higher voltage. Those changes result in much beefier armor, much smaller robot size and much higher kinetic energy. I hope I can try some of those radical designs in this robot. 

    Design expectation:

    Drum spinner

    Two wheel drive

    Make the size as small as possible

    Hub motor weapon

    One-piece 7075 aluminum alloy weapon uprights

    TPU wheel guard

    Run on 6s voltage

    Able to drive inverted

    Self-right by hitting the ground or using gyroscopic effect

    Have a wedge configuration

    Have a fork configuration

    Have a wedgeless configuration

    In Chinese beetleweight competitions, hub motor drums are very common and there are some reliable designs, rarely showing any problems in matches, especially after Aiden from Team Gladiator came up with a much simpler hub motor design used in his beetleweight robots. He explain his design thoroughly in this blog, and I learnt a lot from it: https://mc.dfrobot.com.cn/thread-36190-1-1.html

    I have strong doubt about whether I have the ability to make hub motor weapon or not, since all I know about hub motor designs are merely some observation of other teams' robots and some information from the blogs that other builders wrote. And I absolutely don't have the resource to take the risk of failing and build a robot just to test everything out. So I have to consider about as many problems as I can think of, and the design must have a high fault tolerance, in order to be able to make some upgrades after finding problems. Also, I need to put as much protection as I can provide to the brittle parts like motors in the design. 

    We can first list all the known problems and problems that I suspect will happen: 

    Known problems:

    Magnets on motor crack

    Magnets come off from the motor

    Cocentricity issues between the motor and the drum

    Motor mounting bolts come off due to vibration from the weapon

    Motor or ESC burn up due to motor stalling

    Problems that I suspect will happen:

    The stator and rotoron the motor rub against each other when making weapon exchange

    Motor Bearing get stuck due to the load

    draining the battery due to motor stalling

    Hard to take the rotor out of the drum once it is pressed in

    The problems mentioned above might be impossible to happen or too subtle to be worried about, but I don't have enough experience and evidence to prove that they don't exist, so it is quite neccesary to keep an eye on them. 

    The most common approach of making hub motor is to add huge bearings at the sides of the drum, and add a layer of aluminum alloy to absorb the vibration. But this approach has a lot of uncertainties: how thick does the aluminum layer need to be? Does it provide enough protection to the bearings on the motor? What can I do if there is cocentricity issues? Those questions makes me have doubt about the plausibility of using an aluminum layer to protect the motor, and it would be better to use something softer, so that it can tolerant some cocentricity issue, absorb more shock, and slip a little when sudden load is imposed. Polyurethane can probably do that. And it is way lighter than aluminum so I can put more weight on the steel drum that is further away from the axis to gain more moment of inertia. This also makes repairing so much easier, since you can just pull the rotor away from the drum. 

    The motor I tentatively choose is the sunnysky V3 2216 1000kv. The previous two beetleweights had no problem with this motor and it is very powerful. And I have some motor with the same size right now so I can take measurements when needed. 

    It is hard to choose the size of the bearing since I have literally no I idea how much load the bearing need to withstand in actual combat and I can't even estimate. The only reference I have is the bearings that other teams chose. Aiden used two ball bearings with maximum dynamic load of 8kN in his 2kg robot, and I think half of that would be sufficient for beetleweight. I found a ball bearing with a inner diameter of 25mm, a outer diameter of 37mm and a thickness of 7mm, its maximum dynamic load is 4.3kN, and the size is appropriate, and provide enough space for the motor and PU belts. 

    About the design of the frame, in China, most of the robots are aluminum and more and more robots are made out of one single chunk of aluminum. But outside China, plastic frames become more common and HDPE or UHMW frames are have shown great reliability. At first I planned to make an one-piece aluminum frame, but later I found multiple problems with it:

    1. That big of an aluminum part would cost a lot. 

    2. Because of the width of the drive motor, the side panel connecting the front half and the rear half of the frame is really thin and making the whole frame with one big CNC part barely has any structural advantage. If I increase the side panel thickness, the cost and the weight of the part will increase significantly. I would rather reinforce the frame by make the bottom and top panel thicker. 

    3. If one part of the frame breaks, the whole frame needs to be replaced. 

    I feel like an one-piece aluminum frame might not be the best solution in this circumstance. Then how about switching to a plastic frame? Using an one piece plastic frame might have some problems as well: 

    1. Are bolts strong enough when they are screwed into plastic threads? Are they strong enough to hold the wheel guards and top panel? 

    2. Is it cheaper or more expensive than aluminum?

    3. The whole frame needs to be replaced if one part of the frame breaks. 

    If I use plastic sheets to make the frame instead of an one-piece plastic part to reduce cost, the structural strength of the frame would be weakened even more. Since both aluminum and plastic frame have their problem, I think combining the two idea might work well. I can make an aluminum weapon module, and use TPU for the rest of the frame. I can connect the top and bottom panels with aluminum threaded standoffs and those standoffs can be anchors for the TPU parts to be mounted on. The major advantage of this structure is its low cost, since the weapon module is the only relatively expensive part that requires CNC, and all other parts are either standard parts or 3D printed parts. I think the structural strength would be okay if properly designed, and the TPU wheel guards can absorb tons of shocks. 

    Next let's talk about electronics. The reason why I want to use 6s instead of 4s is I want to achieve the same speed and torque using a motor with lower kv, so the current will be smaller and the motor can be run at a higher efficiency. But the battery size is a problem. 6s batteries are usually thick and hard to fit into the frame vertically. If I place it horizontally, the frame needs to be so much longer. And there aren't really a lot of companies that makes 6s battery at such a small size, so I don't really have many choice. I want to try to assemble my own battery pack using six 1s battery, so the layout of the battery can be more flexible. The battery I chose is Tattu 1s 500mah 95c, and stacking six of them would be 36mm thick, 18.5mm wide and 58mm long which should fit into the frame. The 6s500mah battery should have approximately the same watt hour as the 4s750mah used in previous robots, so it should be enough. 

    The drivetrain I decided to use is once again the Heiyue Drivetrain V2 and cyclone 35a ESC. As I do more and more driving practice, Heiyue Drivetrain under 4s feels a little too slow for me. If 6s is too fast, I can always adjust the mixing settings and throttle curve. A higher maximum speed can help me escape in emergency situation and when the robot is trying to self-right, a higher turning speed can speed up the process. 

    On my second beetleweight, I tried to use a dys 70a ESC and use an extra bec to power the receiver. Though not tested in actual fights, this configuration did survive some tough tests, so I want to stick with it. In theory, running 6s voltage can reduce the current, so using a 50a skywalker ESC should also be fine. both configuration would fit into the frame and you can replace them at any time, so it is not an urgent decision that needs to be settled. 

    Next its the wheel guard design, in the design expectation I mentioned we need a wedge configuration for fighting against horizontal spinners, which is usually pretty heavy. So when I'm using the wedge, I nedd to get rid of some armor on the other parts of the robot to compensate for the extra weight. After some weight estimation work, I came to the three following configurations:

    No wedge+fully surrounded wheel guard

    Fork+fully surrounded wheel guard

    wedge+half surrounded wheel guard

    Among the three configurations, the wedge configuration is probably the most tricky one. I planned to use a TPU part to connect the frame and the wedge, so the TPU part can absorb some shock. This configuration is designed to fight against horizontal spinner, so the sides of the weapon uprights are super likely to get hit. If the weapon uprights deform, not only the weapon will get stuck, but the repairing work will also get much harder, since parts may get stuck, so the whole weapon uprights must be covered up, so horizontal spinners can't get bite on the robot except the drum. 

    I'm not sure when I can start building this design, but it would be something after the retirement of my second beetleweight. I will keep perfecting this design in the future. 

CAD link: 

https://cad.onshape.com/documents/a2eeb8333a2269c4bd53f085/w/933dc104d214b3d48bb6ee1d/e/f46be0bddf53ee4565faa34a