Event Horizon Part 2: Machining and Electronics

The machining of Event Horizon started with a large amount of work on the CNC router. I cut all of the UHMW pieces first, then cut the aluminum side plates and the bottom plate. UHMW cuts very easily on the router. All UHMW was cut with a single flute 1/8" endmill. Aluminum pieces were cut with an aluminum cutting router bit (one of these).

Preparing to cut the side pieces

Some of the pieces completed.

I had not actually cut 7075 aluminum on the router before, and it basically consists of dumping lots of flood cooling onto the piece continuously and hoping that nothing breaks.

The bottom plate after machining

I assembled all of the router pieces, which generally fit well. The grooves in the back piece were not quite wide enough and I had to file them out slightly. This is because the UHMW stock is actually slightly thicker than 3/8", which is a common problem with UHMW.

Initial assembly of the frame

Next was the machining of the drum pieces. I did most of this with the big lathe, which I attached a quick change tool post to. The first drum piece was the drum itself.

I started by cutting the stock roughly to length using the horizontal bandsaw, and then faced off the ends to the correct length.

Drum end being faced down

Drum stock cut to length and polished with fine grain sand paper

I initially planned to drill the holes on the mill with the small rotary table, however, after fixturing the piece initially, I realized that the small rotary table does not even come close to holding large parts straight. Additionally, the locking mechanism that allows the chuck to tilt is not strong enough to allow me to drill the holes farthest from the chuck jaws without slipping. I thought of three possible solutions to this: 1) put a block under the drum piece at the end 2) buy a lathe chuck for the big rotary table 3) drill the holes with the drum piece held in the vise. I decided to go with option 3 since blocking the piece up would not solve the issue of the chuck being off center, and buying a lathe chuck for the big rotary table would be expensive. I used a height gauge and the small rotary table to mark the end of the drum every 60 degrees and then clamped the drum piece in the vise, using a parallel block as a stop. The parallel block doubles as a point of alignment for the marks on the drum, allowing me to drill the holes at 60 degree intervals with good accuracy.

Drum piece clamped in the vise

After the completion of the drum, I made the two bearing rings and the motor mounting hub. This required several hours of machining on the lathe since all three pieces have the majority of material removed from the original piece of stock. I also machined the axle piece for the free-spinning side of the drum.

Completed drum piece and motor mounted to motor hub

I managed to break not one, but two 4-40 taps off in the bearing ring for the 3/4" bearing, which means that I have to leave two of the six screws out. This is fine for now, but I will probably re-make the piece at some later point.

Frame with drum attached

At this point, the majority of work was done on the bot and I just had to finish up a few minor parts. The first of these parts were the armor support blocks, which back the front half of each aluminum armor plate. I cut these pieces from some leftover EH 1 frame stock and beveled them on the mill.

My solution to putting beveled edges on things

Next, was finishing off the drive train. I finally got around to generating G-code for cutting the pulley blanks (the pulleys have a hex bore and I don't have a hex broach), and CNCed four, which I then added grooves to on the lathe. I also machined hubs for the front wheels from some pieces of aluminum hex shaft. The hubs were a little rough while spinning on the shoulder bolts, so I put a little spindle oil on the shoulder bolts, which solved the issue. I also made the disk for connecting the motor to the drum, which I cut from the same piece of UHMW sheet that the frame came from.

EH with completed drum, side armor supports, and front wheels attached

Aside from the top plate, the bot was mechanically done. I weighed it with electronics, and found that it was ~5-6 oz. over weight, which is a lot.  I removed the side armor braces, drilled holes in the frame side plates and enlarged some of the holes on the drum (also useful as vent holes for the motor). The side armor pieces should be unnecessary against all robots but powerful horizontal spinners. Hopefully, the angle of the armor will be sufficient to deal with these types of robots. After looking over the electronics system, I realized that the drive motor controller that I weighed the robot with is actually an ounce heavier than the two Vex Motor Controller 29's that I was planning to use, which helped as well. These changes brought the weight down to a more manageable 3 lb. 2 oz, so I started working on the electronics.

I have not yet mentioned the control system for this bot, so here it is:

• Drive controllers: modified Vex Motor Controller 29 (2x)
• Weapon: HobbyKing 40A ESC
• Radio: HobbyKing 6 channel radio
• Battery: E-Flite 3S 1300 mAh LiPo
• Power Switch: FingerTech Mini Power Switch

I started by removing the casing of the Motor Controller 29's and cutting off the connectors. I then removed the casing from the ESC and soldered the V_in wires of the 29's to the ESC board where the power wires are attached. This allows me to eliminate the ground wires running to the receiver from the drive motor controllers, saving some weight and reducing the number of wires cluttering the inside of the bot. I cut down the output wires of the motor controller 29's, and soldered on the bullet connectors, then taped everything together in a small block. Before competing, I will make a second block for a quick replacement should anything in the electronics system decide to spontaneously combust.

Electronics block soldered together and mostly completed

I finished the electronics system by adding the battery connector, PWM connectors, and power switch, then stuffing everything into the rather limited space between the motor and the inside frame member.

Electronics system mostly installed

With the electronics system now complete, I finished the top plate and assembled the full bot.

Completed bot

In all, the weight came out to 3 lb. 1.3 oz. I should be able to cut the remaining excess weight by changing the top and bottom plates to garolite and polycarbonate, and if necessary, lightening the middle armor support blocks substantially.

And, of course, here is some testing footage:

EH is more maneuverable than it looks in this video. I am just bad at filming and driving at the same time.

I noticed a few things while testing today that could be problematic:

1. The weapon can spin up fast enough to flip the bot over on startup, then send it shooting across the floor. This is why you see me spinning it up slowly in the video. I might be able to use this for self righting, which would be very useful.
2. The turning speed is very limited when the weapon is at full speed because the gyroscopic forces cause the bot to tip and ground out on the side armor.
3. One of the front wheel axle bolts came loose. I will need to loctite everything in place before competing.
4. I now need to make a testing box.

Event Horizon Beetle Weight Combat Robot Part 1: Design

Some Background:

My interest in combat robotics goes back many years. When I was in middle school, my father brought home a few books on the subject and I was enthralled. I thought up a few ideas, and had some vague plans to some day build a robot. And then I joined FRC 3467 and my combat robotics ambitions were superseded by FRC. A little over a year ago, after becoming frustrated with the state of the FRC team, I began looking for alternative robotics competitions. At some point, I started designing combat robots again (probably had something to do with the renew of Battlebots). Originally, my robot was supposed to be a 30 pounder, but as I started filling out those designs, I realized that it would be too expensive for my limited budget. Eventually I settled on a UHMW unibody 3 lb. drum spinner and created the first iterations of Event Horizon (EH). Since then, EH has gone through numerous revisions and two of those revisions have reached the point of being made.

CAD of the entire EH line may be found in this GrabCad folder: https://workbench.grabcad.com/workbench/projects/gcvuOsckFDIWLJdoowpnEofy4OMchZ2VH8DeKmusFdmL7P#/space/gcvl1rrfAIgrpm1Xv3alz6UXAZ9_w34lScfgpi5XAR6gS5
(Note: I re-arranged my GrabCAD projects and this link is broken now. I'll update it when it's fixed)

Event Horizon V1:

I quickly abandoned the unibody aspect of EH after concluding that my manual milling skills were not sufficient to actually make the frame. The design progressed to finger-jointed UHMW plates. After a fairly lengthy design process, I finally bought parts and machined EH 1 over winter break last year.

The frame of EH 1

Here is the completed frame of EH 1. The large chunk missing from the front of the right side plate was a milling mistake on my part. I can only imagine how horrible the results would have been had I tried the unibody.

Here is EH 1 fully wired and functional. I added wheel guards on at some point after these pictures were taken.

This is about as far as the EH 1 design got. I intended to compete with it at Motorama 2016 but Motorama 2016 fell on the last weekend of the 2016 FRC build season and as the design lead, I could not justify being gone at such a critical point in the season. EH 1 floated around my room waiting for me to find it a competition. But that didn't happen before I decided to re-design it.

Event Horizon V2/2.1:

Once the FRC season ended, several things had changed: 1) My father bought a CNC router, 2) I had spent a lot more time using the Bridgeport, 3) I had an internship at Foxx Life Sciences. These factors meant that I now had vastly expanded machining resources and money to throw at expensive robotics projects, which obviously meant that I was going to take on another robotics project.

After completing and testing EH 1, I was immediately somewhat dissatisfied with it. The weapon did not store enough energy. It was light, small diameter, and low speed. And so, a new robotics adventure began: make a robot with a really big weapon.

EH 2 began as a completely separate design from EH 1 and heavily utilized the CNC router. It used a strange drive train setup involving offset wheels for the ability to drive upside-down. It also used inverted motors, which I eventually deemed too space inefficient and changed. The front was a wedge, which was primarily designed for fighting wedge bots. (This is the Event Horizon v2 and Event Horizon v2.1 folders in GrabCad) I was about to order parts for EH 2.1 when I started having second thoughts. The odd wheel configuration was a liability as a spinner could easily take off the belts, the wedge probably did not have sufficient down force because of the wheel position, and the weapon motor was exposed and asking to be forcibly removed. Instead of buying parts for EH 2.1, I briefly tried to remedy the issues in EH v2.2, then scrapped the design and started over.

Event Horizon V2.3:

Frame:

For the next version of the Event Horizon design, I started with a more conventional wheel setup, using four 1 5/8" Bane Bots wheels, two on each side of the bot. The back set is direct driven off of two KitBots 1000 rpm gear motors (leftover from EH 1). The front set is driven via pulleys off of the back set. The smaller wheels and four wheel setup should make EH 2.3 slightly slower and easier to control than EH 1. Unlike EH 2.1, 2.3 has angled 7075 aluminum side armor, which is supported by UHMW blocks, and a UHMW plate in the front section to absorb spinner hits. The frame also has two ears at the top to prevent the drum from touching the ground should the bot be flipped over. Rather than the offset wheel arrangement of EH 2.1, 2.3 has an angled section on its top plate that exposes the wheels when the robot is flipped over. The front of the EH 2.3 frame is flat, which is not good for wedges. If I have weight left over after everything is assembled, I will add removable wedgelets of some sort to the front and may also make a set of removable ablative armor blocks for spinners.

Weapon:

For this version, I finally put the weapon motor inside of the drum (a result of getting the Hendey Lathe). Originally, I was going to disassemble the motor and machine a new housing for it to interface with the drum, but after receiving the motor, I realized that it would be easier to simply mount the entire motor  inside of the drum without modification. This has the additional benefit that it will make the motor much easier to replace if I manage to break it. The motor has a free speed of around 15,000 rpm at 11.1 V, which is fast enough that I don't yet feel the need to re-wind it (a boost converter may happen at some point, though). Initially, the drum had a titanium shaft running through it. This change in motor setup makes a through-shaft impossible and forced me to use separate hubs on each side of the drum. Hopefully this will not cause the frame to bend in after repeated hits. If it does, I will have to switch back to the original design.

When I first designed EH 2.3, it used a steel drum with steel bars welded on as teeth. I actually ordered the stock for this design, but since I don't know how to weld, I switched to an aluminum drum with flat head screws as teeth. I may go back to the welded steel drum at some point in the future, but for now, I want to get the bot up and running. You can see the various drum and motor designs in the EH 2.3 GrabCad folder. Common to all of the drum designs is an inset hub on each side of the frame. This is because the frame is made from UHMW, which is very soft. Repeated impacts on a small shaft would likely dent in the UHMW and cause issues over time. 

To protect the drive base, the frame pieces have some nubs on the bottom, which should bottom out as the wheels compress during a hard hit.

That's it for my summary of the Event Horizon design process. EH 2.3 is now in its final stages of machining. I will post a writeup of that when it is completed.