When I last made a combat robot about 20 years ago, the thick cardboard and duct tape armour I used was borderline indestructible. Things have come on quite a long way, I see.
@@jerkwagon It feels like a lot of my TPU print designs have benefited more from being designed to print flat and then fold into shape than being printed 'in pose' as it were, this allows for many long extrudes within the slice, giving more strength than lots of short extrudes within the slice, particularly regarding the perimeters of the object's print.
@@jerkwagon you can change the support offset distance to get supports to peel off like in the video. I also print you really hot to get more homogeneous structure.
@@connorhoward895 Seconded. My preferred TPU print temp is 240c, sticks really well to the E3 S1 Pro's PEI coated sprint steel bed even with it only at ~45c. Almost too well on bigger / more rigid TPU prints, but plenty of homogeneity & strength. Blows my mind how strong just some small diameter strands of TPU can be, and it's a pretty rugged/hard-wearing material too.
A suggestion from my side would be to actually make your build even more flexible by not having a single stiff PLA plate but rather have like 2 or 3 plates. They could give stiffness to the different components, but let the chassi as a whole be more resilient to crushing damage. If the rotation of your weapon causes too much flex by itself, you could stiffen it more by having a joint connecting the plates. Also it might be woth to replace the metal screws with plastic ones to save some weight.
I'm wondering if the torque stresses would cause it to be difficult to control, but like you said, one could just have a support skeleton instead of a whole plate.
Try to use one wheel in the center (like the ball that a computer mouse used to have) and in the case of an inversion - if the body is thin enough and the ball is protruding, it can still drive upside down
So glad you're doing a video about the wonders of TPU, it's my favourite filament to design parts for due to its flexibility. Over the past couple of years I came up with designs to securely hold radio receivers & antennas on drones in ways nobody had thought of before which are very easy to print, those designs are now used by thousands of people. "Hoopin TPU Cable Tie" is another popular design of mine I created 8 months ago which is also very easy to print, it's very strong as demonstrated by holding up my 3D printer with a 100% scale print in the photos, and 150% scale print easily held 12kg weight.
Cool build. My suggestions are rather than having the strengthening plate a solid piece it should have holes cut out where there's no board or anything to protect. It will keep the same strength but be lighter. Likewise skeletonised wheel covers could do the same. Looking forward to seeing the next version as this is a really restrictive category but that forces creativity.
@@HMPirates yeah i think that was the idea behind this bodies design being so round, you can see most of the damage is where it stuck out a bit and an opponents weapon was able to get leverage and shear the tpu
@@HMPirates Fair enough. That makes sense for some parts but the strengthening plate could definitely be skeletonised without offering openings for attack as there looks like there's material behind it in a lot of places. I was also thinking if the part was strong enough their weapons might get caught in the armor but I guess that wouldn't work for 3d printed plastic
I remember seeing him commenting in the NHRL tournament live stream and waited in angst for this video! Love to see more combat robot videos in the future
I make the Jolt! Kits for the 3lb weight class in the USA, but for that I use 7075 aluminum plates to give the needed stiffness to mount the spinny bits and TPU for the armor and squishy mounting to protect everything, so TPU can definitely work as both armor and a main structural component even at the larger insect weight classes
Just so the size difference is easy to understand. It takes slightly more than nine (9) of these 150 gram bots to make 3 lbs... 150 grams is just shy of 5.3 ounces, and it really isn't much when you have to have motors, a battery, and receiver. Weight quickly get to be an issue.
This seems like a great use for TPU. I recently did my first TPU print. What I was most surprised of is how hard it gets when you give it a few layers. I also was super surprised about how resistant it is to fatigue cracks. I've been using a print as a fidget toy. and even trying to get it to break, but it keeps holding its shape, with only the mildest deformation.
It’s such an interesting material. One of the most satisfying things I’ve printed was just a 0.2mm, single layer of TPU across my entire print bed haha
PP is also a neat material, somewhere halfway in between PLA and TPU, with better layer adhesion then either, extremely fatigue resistant, harder than TPU, softer than PLA
It can vary between different TPU filaments/manufacturers, but generally speaking, it is super strong stuff. It's very difficult to stretch most TPU to the point of deformation by hand. Even harder to actually snap/break it - even something printed like a skinny rubber-band. Good material for coasters, bit holders, soft jaws, bumpers, soft feet/pads for furniture, etc.
From personal experience with TPU combat robots, a few notes I have are to use plastites with undersized holes rather than heat-set inserts, use thin UHMW for top plate (I use 1/16” for a beetle), and print hot and slow to increase the strength of the TPU prints. Also, if you build a beetle or larger, you should reach out to NHRL and they might be able to sponsor you competing there
I would recommend plain Nylon instead of CF Nylon for that top part. The CF makes the part stiffer but also more brittle. Cool bot! Here in the US we have been using TPU for all sorts of things on combat bots for years now.
This is funny, some drone frame builders, build a protective pod with TPU (as opposed to a top and bottom plate made of carbon fiber with standoffs in the middle). It is often heavier, but much more protective of your camera and other electronics inside. OpenRACER is an open source design that has a TPU thicker option, or a thinner PCTPE (nylon+TPE) blend. The squishy properties help it not to shatter in 90+mph hits. One tip we have is always bring multiple spare pods or covers for your robots to swap between rounds. For races we often bring spare pods, but also spare drones, i will usually take 5-6 for each race, more competitive racers bring 10+ drones. Take a look at week points and beefen all of those up as mentioned, and yes, motor protection for robots and drones is important, as nothing will keep you from taking off quicker, so beef that us as well. For some more rigid parts, we often get them cut out of carbon fiber. It may cost you a little in weight, but should also make it a bit stronger, im thinking 2.5-3mm for y our application.
Funny to see you here Jon . :) For others, we actually know each other from the local FPV/drone community and the OpenRacer project. I started experimenting with different types of Nylon for the OpenRacer pods we use to protect our electronics. After printing literally hundreds for our local pilots to crash I found that when it comes to impact resistance and protecting electronics, you want something rigid enough to protect internals, but the least rigid possible to achieve that. In short, the most rubbery you can get away with while still functioning properly. That's what brought me to PCTPE, because I needed a little more rigidity from Nylon; but dialed back just a little for better durability. I was watching this video because I have a pretty big interested in combat robotics, and it popped up in my subscriptions. I even went to Battlebots in Vegas this year, which was a pretty cool experience. I'm planning to start printing some antweights and get my feet wet. The knowledge gained from working on the OpenRacer project and other FPV/Drone design work should have a lot of crossover, as @Maker's Muse said. Pretty cool to see TPU making its way into use here.
@JonE5FPV @d3c0de I recognise those FPV names :) I saw another pilot, metaldirtnskin, on MakersMuse's Prusa i3 Mk4 unboxing livestream recently. Designing drone parts is a big reason I got my Prusa i3 Mk3, having fun learning how to design small prints which work with the 'grain' of how it's printed to get the most out of the minimal filament used, Bardwell's shown & recommend my prints for quads on his channel about a dozen times the past couple of years which is nice. His reaction to my 3D printed TPU props was hilarious, he couldn't quite believe they worked so well, neither can I actually, specially as they were the first print of the first iteration of the design!
This is so serendipitous. I was just at a Maker Faire on the weekend and I watched a beetle weight competition. This would be so cool, because you could have multiple chassis ready to swap out.
So some tips for making weight easier, carbon fiber top plate it adds the rigidity and is not heavy at all. you can probably get away with a 1mm thick one. ditch the heat set inserts, they weigh a ton you can use #4 plastice screws. the self cutting type. you can get really fancy with prusa slicer settings. like removing the top layers in the battery compartment, or changing wall line counts in areas you need more or less. also the more compact you make your bot the less it weighs. it looks like you could have reduced the length of pancake a good bit since you had a small diameter blade look at getting a blade made from 1mm grade 5 Ti its a lot lighter but very strong.
That thing about the inserts is really worth thinking about. Now the problem with self tapping screws is mostly when you have to disassemble and assemble things over and over, but with some care they can be used without having them chew up the material every time you assemble it again. But I wonder if anyone makes aluminium inserts? If you don't want to use self tappers alu inserts should be good enough to replace brass inserts and probably at half the weight. Also the lighter blade is a good idea. At this small size the momentum of the blade won't be massive, so the motor spinning it is more important than having a heavy blade. Or at least I think that's how it works.
suggestion: paint/print it green and add googly eyes! the little guy has so much personality leaping around the arena that he DESERVES to be a cute little frog!
You may want to try encasing everything inside a TPU case: print the bottom half, install the electronics, print the top half. This way you'll save weight on screws and inserts. Design your robot with disposability in mind, these tiny robots get so banged up in a weekend you'll end up printing another one anyway.
TPU is awesome. use the hardest shore rating you can for firmness. usually the cheap stuff. that's been a go to for quadcopters for years and we crash it hard. arm covers, antenna mounts, gopro mounts and cases.
0:52. Same reason rock slide barriers are nets instead of concrete walls. Shock absorption and increasing deceleration time result in less force being translated to the rest of the device. Modern cars cringle in crashes for the same reason. They are designed to collapse in order achieve the same effect.
Been flying racing drones since 2015....TPU is used heavily to protect our GoPro's, different parts of carbon fiber and to hold our antennas. So using it in smaller battle bots makes perfect sense
I imagine TPU-based Armor will hit the point of inversion of effectiveness once it's taken into Featherweights or heavier, since flywheel-style weapons ramp up in power significantly (probably exponentially) the higher up the weight classes you go, especially once you get to the full size arena classes. Once a flywheel gets a solid bite on a material, it will try to shear it away
Quite a cool innovation, possiply with the too plate have it sceltonized in places to increase flex, reduce weight, add airflow to help with cooling. Possibly add a vent next to the weapon as its spunning fast enough to push air around.
You should give TPE a go for the frame, its a lot better with repeated stress and shearing. I've used this as a stopper for airgun gas rams which experience a lot of abuse and its held up great. Its basically a rubber block that the ram crashes into so it doesn't damage the metal. Its better to use raw and uncolored materials as I noticed ones with certain colors seem to be weaker or tear easier.
@@conorstewart2214 you would design a channel for it in the model and then add a pause to the print before it would enclose it so you can add the wire. Or alternatively you could just design a hole through the length of the chassis and insert it when it's done printing.
I love it! An evolution on your 1-wheeled idea: If your single driven wheel was mounted as close to centre as possible, and the chassis sat low enough to the ground as to not be teeter-tottering, you could consider having that single wheel given a 180-degree sweep of steering articulation. Think of the handling characteristics like you would a bumper car, or even a ship's azipod. With a slip ring, you might even be able to achieve an infinite sweep of steering articulation. I imagine the slip ring's commutation would have to be on the outermost periphery of the wheel "pod", though.
Try printing the tpu a few degrees hotter for better layer adhesion. It has worked well for me in 1lb antweights. A few degrees makes a big difference. I also do like 4-5 walls and that seems to work well. I have also done just a tpu bumper around a pla+ chassis and that worked fairly well, but if it gets hit just right, it tends to be pulled off. I actually just printed a very similar 150g undercutter but I am attempting to use a much harder tpu. I think its rigid enough I won't need to use any pla, but we will see what happens once it goes into combat.
I have been making lightweight panels for my rc car. I havent tried the foaming epoxy yet, i think it'll make a huge difference: 1 - secure aliexpress 0.2mm fibre sheet to print bed. 2 - print with minimal gyroid infill, 0 top or bottom layers. 3 - using a syringe, measure out an appropriate volume of foaming epoxy, taking model volume and the advertised expansion ratio into account. 4 - place another prepared sheet on top. I believe the process could be used to make simple curved parts too - you can flex the panels into place, temporarily secure with superglue then inject the foaming mixture. I think this could make a great baseplate alternative for this combat beast.
I would try removing the ridgid plate all together. Use TPU for the whole assemble, this avoids creating stress concentration around the joins between the flexible and rigid parts.
Seems like the perfect use case for Ninjatek Armadillo, I love that stuff when printed right it's indestructible like other TPU but, much much stiffer.
Something that is worth looking into is the different TPU types/blends. Some TPU can be quite rigid, which might be a better choice for these combat robots, since you might be able to make the whole robot out of it. What I don't know is whether a more rigid TPU would also be a lot worse at absorbing impacts, though it would still be a little flexible. AS for the top plate, CF PA is a good option but I would also be interested to see PETG, since that can be relatively flexible (compared to PLA) in my experience and that might help it survive.
We are exploring the use of TPU and TPE for one off gaskets. The TPE seems to have better layer to layer adhesion. One idea I'd recommend is using a stiff material inside and only where absolutely necessary. Think of the stiff material as a spine and the TPU as muscle. Print the TPU and press the very small amounts of PLA/PLA+/PLA-CF into the TPU. No need for a wide piece, and keeping it narrow, the stiff part can be thicker to be more rugged.
I literally said “holy f*****g s**t” when you said the weight limit was 150g. What an incredibly interesting weight class, I’m honestly astounded by the capabilities possible with such little allocation towards total weight.
Rubber robots can absorb impacts but the compliant surface allows a spinner's claws to get a better grip and throw them harder. What I'd want to try is having a multiple hard shells and the electronics mounted to a soft inner frame. You get the best of both worlds: hard surface to glance blows from spinners and soft inner frame resistant to bending which cushions the electronics.
I would try to replace the PLA plate with something like carbon rods as skeleton. Then put a relatively soft TPU body around it to absorb the energy of hits. Lastly (if possible because of the weight) add thin metal scales to the convex parts of the body to prevent cuts.
I use weird TPUs quite a lot to make surgical simulators. I found that a good way to make it stronger is to run a soldering iron on the outer surface of the print in areas of high stress. This connects the layers together and prevents delamination… hope this helps 👍🏻
Brilliant. I've switched to TPU for many of my functional parts that need to be durable on my ebikes, including a massive frame mount battery enclosure. I should try to build my own and share, I have ideas on how to make a good TPU frame for such a combat robot.
don't use CF nylon. The fibers primarily add stiffness, not strength. If you're going max toughness, use pure nylon (actually, that might be even better than TPU for durability too. You should try it)
idk if it will be of any help, but for squishy materials ive had the best luck in terms of weight:strength ratios using circular infills. try to set a infill % in which the corners are well supported by having the smallest fillet and largest infill circles possible. the circles tend to absorb the shock by folding/squishing rather than pushing the infill walls back rather than crumpling then bouncing out. back when i was testing this, i ended up making egg cradles and doing a classic egg drop test with the only change being the infill type. although bouncy, the circle bois had the highest drop height and survival rate.
My thoughts include not using PLA as support for the weapon. Instead, let it flex, but to avoid cutting into itself add slight PLA or similar to the exact part that would bend for it to hit itself. Potentially is a rebar kind of fashion, with thin bars of PLA threaded inside the TPU to offer rigidity without adding much weight (Currently assuming PLA weighs more, otherwise printing sections in PLA (melted to connect to TPU) might serve even better). Or potentially even thread actual metal nails or mesh as rebar, if light enough. Instead of 1 wheel+1 weapon, use 2 weapons so it compensates for itself trying to spin like a reaction wheel. Then either have the weapons be the wheels, or add 1 wheel for "forwards/backwards" (with Yaw being set through modulating the weapons spin-ratio so one overpowers the other, turning the robot as desired as a reaction-wheel).
I would recommend designing it so the expensive parts have their own area and are isolated. This way if the battery or the ESC blows up you have a much lower chance of destroying the main board which is most expensive. Also I feel like you could lower the size of the front motor and still have most of the power.
A bit late here, but three thoughts. 1, look at floor buffers/scrubbers for motion inspiration. When operating one you lift/lower the handle to get it to swing left/right and experienced users can give it a slight twist to get ti to move itself forwards and backwards. 2, as much as possible remove anything rigid. That plate, by virtue of it being rigid, was taking on the strain that would otherwise have been dissipated through the TPU. Instead do something like printing a motor mount and electronics shell only, and insert them into the TPU like you would a captive nut in a print. 3, to minimize any corners enemy weapons could catch on, consider making the entire perimeter a continuous circle. To summarize: if I were going to do an iteration on your design I'd make something that looks even more like a pancake, fully circular where the only external moving part was the weapon (can't loose a wheel you don't have). The part that didn't overlap would be thicker to act like a "foot", which would act similar to something like soft robotics and use servos to flex the foot and manipulate how the weapon spindle end touches the ground to effect rotation and forward/reverse motion. Even more interestingly, if the weapon spindle came out both sides you could make it agnostic to which way it was flipped. Being upside down it'd reverse your flexing inputs, but it would also reverse the rotation direction in relation to contact reversing the effects as well and making your inputs have the same result no matter which way up it was.
Great video and great robot, I love the design. I love the idea of a one wheeled one in the future. I've gotta get into this stuff in the UK :). I've attended a few events a very long time ago. Now I've got a 3D printer ;) .
I forget the full story, but somewhere around the 50's (or maybe pre-war) an engineer at MG took a drill to the chassis of their race cars to lower the weights. I think he also did it to few wrenches too. So put some "lightening" holes in the rigid plate, and maybe the body.
I know nothing about combat robots, but I *love* Colorfabb's Varioshore TPU. I have printed so many different types of functional and high stress parts with it. For example, I printed custom 2-in-1 doorstop/catches for a pair of double-swing cafe doors. These things take a beating daily as they stop the doors from hitting the wall as they are swung [and sometimes kicked] open, while also catching them in place to keep from swinging back shut. A simple flick of the latch and they release/close. A singular piece of printed TPU attached to the wall with sufficient wall anchors and they have lasted over a year so far now with no signs of deformation or wear. I do have a thought about the robot, tho. Think of all those living hinges out there for other materials, then think of how much better they would be in TPU. Perhaps instead of normal slicer designed infill, you design the infill into the actual model using a thin accordion-type leafing that would compress in exactly the ways you want, while being rigid where you need it. Slicer infill is fine for most prints, but I find sometimes it is necessary to design the infill yourself as part of the part and leave the slicer out of the equation so maximal control over placement is possible.
Have you tried using polycarbonate for frames or reinforcement plates? It is supposed to have a very high impact strength. Another flexible material you could try is polypropylene, it is pretty flexible and very durable with excellent layer adhesion. For the most part if you don’t need it to be very very soft and flexible then polypropylene is probably better than TPU. It is excellent for things like living hinges because it just doesn’t break. I think I have seen some with an elongation at break of something like 600 %.
between the heavy weapon motor and the heavy electronics, the relatively thin connecting piece is always going to flex, no matter what you try. I'd design a certain amount of flex into the robot and give it an articulated top plate.
@@mifiamigahna No, the "articulated" idea was god awful and stupid. Flex is always an issue with horizontal spinners, but you don't see anyone else make a fucking ARTICULATED arm to hold the weapon because if you do that, it literally makes the entire weapon useless. Source: have actually participated in one of these events
You might try a polycarbonate frame, surrounded by a plain nylon bumper that you soak in water for a few days. Poly is really rigid but can still take a hit, and soaking nylon makes it gain a lot of flexibility. String trimmer line often recommends doing that because it takes the hits better.
I run a TPU chassis for my 3lb. Can whole-heartedly endorse using thread-forming screws (McMaster has some specially for plastic that I like). I find they're lighter and hold better than heat inserts. I’ve only ever had one get ripped out in almost 3 years of using them (they do get bent when they get hit)
My step brother blew the esc in his crawler a couple of weeks ago. The smell wouldn’t go away as he was changing it out outdoors, the day after it blew.
I've been printing with TPU for a little bit in my 3lb robot, Radix. TPU is fantastic for absorbing hits. My drive has never gone down before. The biggest challenge I've faced is hitting my own weapon motor with my weapon when my frame flexes. I think I've fixed this issue, but we will see at NHRL in June.
FPV freestyle enthusiast here, our crowd & the drone racing scene would say more walls but honestly no idea how to manage that well given the weight requirement. I suppose you could just go 1-2mm laser cut carbon fiber top plate with a more "open" cage like design like we do in the drone scene. Then you might be able to get away with a couple of strategically placed spots with more TPU walls.
Very interesting, would love to see more on the subject. Don't have any tips but I am always amazed how tough TPU is. I am using single wall TPU end caps for the darts of my big bore blow gun. The only time they get busted is when I hit the cap with another dart. I have never had the layers separate. This stuff is crazy.
PETG & TPU will stick together very well in a multi-filament 3D print I discovered, changing filament at different layers. I once tried printing thing-3565827, "3D-Printable Filament! -Print Your Own Filament for Multi-Color!" out of PETG & TPU, the resulting print from that dual filament sort of worked, I think the % blend of the two needs to be adjusted & tested more to see what mechanical properties you can get out of the final blend.
Just cut a lot of triangles in the chassis. Keep the wheel protection and improve the connection between the esc and the motor. There are some pretty good three pin connector out there that can prevent these things from breaking apart.
I think a dome like shape made of this stuff would have a really high defence stat on it's own. Add a flipping over mechanism with a cutter and it should be awesome
It's worth experimenting with modeling thin voids in your parts to force 'infill' generation selectively, reducing mass and facilitating optimized deformation characteristics.
Idea: 2 weapons motors side by side and a wheel in the middle. Rotate by diffing the weapons speed, forward/back with the single wheel. A bit like you do now, but with 2 you'll have control before the weapon is spun up, and the gyro is a bit more manageable.
Any material can shatter with a high enough impact energy. A measure of how resistant it is, is the 'glass transition temperature' - effectively it's freezing point. For TPU this is around -40C. One of the best options in Polymers is Polyethylene. HDPE for example is around -95C. This means it can survive a higher energy impact and still behave as a plastic, deforming rather than like glass, shattering. HDPE Filament has fallen from favour, but you can still buy it. Just a thought. I have first hand experience of making (heavyweight) robots from HDPE Plastic (Barberous 2). While we had loads of other problems, the shell was only once ever pierced. The previous version was polypropylene - and it did indeed shatter like glass!
I know part of the benefit of that PLA bottom plate is protection when it's flipped, but since its mostly for stiffness I wonder if you can get away with a couple of thin titanium round bar stringers? You can get really thin bar stock online in Australia so it's worth having some on hand anyway
If you want to make your armor lighter, create holes in the bottom and the top, somewhere you don't expect to be hit. Holes are a great way to remove material to balance weight.
wow, that damage in the tpu really surprised me. If you're cleanability regarding supports has gotten that much better, I would give it a try to print it in a 45° angle (since split layers were part of the damage and weapons seem to hit horizontally and vertically mostly)
I know weight is at a premium, but I'm wondering if potting the electronics and\or some of the wiring (maybe in a silicon) would provide some strain relief.
I haven't made a combat robot, but my experience with TPU and Foam disc replacements for nerf blasters has been really positive. I've noted that using small air holes in the print can make the entire structure more durable as air pockets will just sort of... fart out. Fewer walls tends to be better anyways to get the most flex, and infill as low as you can possibly get it will also help that. More flex means less tearing, though I'm not entirely sure that'd apply to getting whacked at high speed by another robot. If you can, having a 3 dimensional infill also helps keep it structurally sound, but keep in mind all the pockets need holes so you'll be drilling/cutting more of them. In the end if your TPU looks a bit like swiss cheese it should hold up to some pretty ludicrous abuse.
Before watching the video yet, I can see how TPU would be great for impact and could be good (especially combined with other stuffer materials as well), but I would worry about it being grabbed and thrown more easily by weapons. Seems like it could be great to have a combination of TPU and some PA6 to provide a similar effect to some phone cases to get the best of both worlds which could be good.
Maybe a composite armor, made of alternating layers of TPU and PLA, could withstand more impacts and cuts. Some bulletproof glasses are made of tempered glass and polycarbonate. It should be possible if your printer can handle multiple materials.
When I last made a combat robot about 20 years ago, the thick cardboard and duct tape armour I used was borderline indestructible. Things have come on quite a long way, I see.
I guess that wouldn't hold up these days, as batteries and weapons have gotten so much more power
@@ManIkWeet Yep, even just the blade on this thing sounds pretty terrifying if that thing were to come at you on foot.
you're cheating if you're using duct tape though, they use that stuff to fix space shuttles
@@7sev7nsev7n7 No one is using duct tape to fix space shuttles because they are no longer in service.
Will you shutup
The biggest tip for printing TPU for combat robots is more walls. Infill you can set to like 2%, and then just bump up the walls as high as you can.
any tips for supports with TPU, like angus mentioned, before they were welded..
@@jerkwagon It feels like a lot of my TPU print designs have benefited more from being designed to print flat and then fold into shape than being printed 'in pose' as it were, this allows for many long extrudes within the slice, giving more strength than lots of short extrudes within the slice, particularly regarding the perimeters of the object's print.
@@jerkwagon you can change the support offset distance to get supports to peel off like in the video. I also print you really hot to get more homogeneous structure.
The new Cura has settings for multimaterial now, I wonder how well you could add a CF Nylon endoskeleton to a TPU print? That would be sweet!
@@connorhoward895 Seconded. My preferred TPU print temp is 240c, sticks really well to the E3 S1 Pro's PEI coated sprint steel bed even with it only at ~45c. Almost too well on bigger / more rigid TPU prints, but plenty of homogeneity & strength. Blows my mind how strong just some small diameter strands of TPU can be, and it's a pretty rugged/hard-wearing material too.
A suggestion from my side would be to actually make your build even more flexible by not having a single stiff PLA plate but rather have like 2 or 3 plates. They could give stiffness to the different components, but let the chassi as a whole be more resilient to crushing damage. If the rotation of your weapon causes too much flex by itself, you could stiffen it more by having a joint connecting the plates. Also it might be woth to replace the metal screws with plastic ones to save some weight.
I'm wondering if the torque stresses would cause it to be difficult to control, but like you said, one could just have a support skeleton instead of a whole plate.
This was my thought as well. Maybe no ridget plates at all? Or maybe just right where it is needed. Like maybe to support the motor-axle connections
Try to use one wheel in the center (like the ball that a computer mouse used to have) and in the case of an inversion - if the body is thin enough and the ball is protruding, it can still drive upside down
That is not a good idea. You also cant turn effectively if you do that.
That allows for lots of drift and little control or grip
@@user-cg3em4cw4f maybe 3 balls... Hmmmmm
@@superslimanoniem4712 Still not used vs wheels just because the complexity leads to horrible breakages in battle.
@@user-cg3em4cw4f more complexity doesn't allways meanbless durability you just need a good design
So glad you're doing a video about the wonders of TPU, it's my favourite filament to design parts for due to its flexibility.
Over the past couple of years I came up with designs to securely hold radio receivers & antennas on drones in ways nobody had thought of before which are very easy to print, those designs are now used by thousands of people.
"Hoopin TPU Cable Tie" is another popular design of mine I created 8 months ago which is also very easy to print, it's very strong as demonstrated by holding up my 3D printer with a 100% scale print in the photos, and 150% scale print easily held 12kg weight.
Have used this one and many of your other great designs. :)
Cool build. My suggestions are rather than having the strengthening plate a solid piece it should have holes cut out where there's no board or anything to protect. It will keep the same strength but be lighter. Likewise skeletonised wheel covers could do the same. Looking forward to seeing the next version as this is a really restrictive category but that forces creativity.
Skeletonised armour is a bad idea as it gives the opponents weapons more bite.
@@HMPirates yeah i think that was the idea behind this bodies design being so round, you can see most of the damage is where it stuck out a bit and an opponents weapon was able to get leverage and shear the tpu
@@HMPirates Fair enough. That makes sense for some parts but the strengthening plate could definitely be skeletonised without offering openings for attack as there looks like there's material behind it in a lot of places.
I was also thinking if the part was strong enough their weapons might get caught in the armor but I guess that wouldn't work for 3d printed plastic
I remember seeing him commenting in the NHRL tournament live stream and waited in angst for this video! Love to see more combat robot videos in the future
I make the Jolt! Kits for the 3lb weight class in the USA, but for that I use 7075 aluminum plates to give the needed stiffness to mount the spinny bits and TPU for the armor and squishy mounting to protect everything, so TPU can definitely work as both armor and a main structural component even at the larger insect weight classes
Just so the size difference is easy to understand. It takes slightly more than nine (9) of these 150 gram bots to make 3 lbs... 150 grams is just shy of 5.3 ounces, and it really isn't much when you have to have motors, a battery, and receiver. Weight quickly get to be an issue.
That strafing idea sound great as a way to make it less predictable as where you'll move. Plus, less weight from the wheel will help in other areas
This seems like a great use for TPU. I recently did my first TPU print. What I was most surprised of is how hard it gets when you give it a few layers. I also was super surprised about how resistant it is to fatigue cracks. I've been using a print as a fidget toy. and even trying to get it to break, but it keeps holding its shape, with only the mildest deformation.
It’s such an interesting material.
One of the most satisfying things I’ve printed was just a 0.2mm, single layer of TPU across my entire print bed haha
PP is also a neat material, somewhere halfway in between PLA and TPU, with better layer adhesion then either, extremely fatigue resistant, harder than TPU, softer than PLA
It can vary between different TPU filaments/manufacturers, but generally speaking, it is super strong stuff. It's very difficult to stretch most TPU to the point of deformation by hand. Even harder to actually snap/break it - even something printed like a skinny rubber-band. Good material for coasters, bit holders, soft jaws, bumpers, soft feet/pads for furniture, etc.
This is an amazing example of the difference between ultimate tensile strength and toughness.
This is so brilliant!! I loved the original battle bots show as a kid and seeing how far things have come in the past few decades is lovely.
This is awesome. Such a cool design. Can’t wait to see version 2
From personal experience with TPU combat robots, a few notes I have are to use plastites with undersized holes rather than heat-set inserts, use thin UHMW for top plate (I use 1/16” for a beetle), and print hot and slow to increase the strength of the TPU prints. Also, if you build a beetle or larger, you should reach out to NHRL and they might be able to sponsor you competing there
I would recommend plain Nylon instead of CF Nylon for that top part. The CF makes the part stiffer but also more brittle. Cool bot! Here in the US we have been using TPU for all sorts of things on combat bots for years now.
This is funny, some drone frame builders, build a protective pod with TPU (as opposed to a top and bottom plate made of carbon fiber with standoffs in the middle). It is often heavier, but much more protective of your camera and other electronics inside. OpenRACER is an open source design that has a TPU thicker option, or a thinner PCTPE (nylon+TPE) blend. The squishy properties help it not to shatter in 90+mph hits.
One tip we have is always bring multiple spare pods or covers for your robots to swap between rounds. For races we often bring spare pods, but also spare drones, i will usually take 5-6 for each race, more competitive racers bring 10+ drones.
Take a look at week points and beefen all of those up as mentioned, and yes, motor protection for robots and drones is important, as nothing will keep you from taking off quicker, so beef that us as well. For some more rigid parts, we often get them cut out of carbon fiber. It may cost you a little in weight, but should also make it a bit stronger, im thinking 2.5-3mm for y our application.
Neat! There's a lot of crossover honestly, these bots all use drone parts as well. The motors are *incredibly* hardy.
Funny to see you here Jon . :) For others, we actually know each other from the local FPV/drone community and the OpenRacer project. I started experimenting with different types of Nylon for the OpenRacer pods we use to protect our electronics. After printing literally hundreds for our local pilots to crash I found that when it comes to impact resistance and protecting electronics, you want something rigid enough to protect internals, but the least rigid possible to achieve that. In short, the most rubbery you can get away with while still functioning properly. That's what brought me to PCTPE, because I needed a little more rigidity from Nylon; but dialed back just a little for better durability. I was watching this video because I have a pretty big interested in combat robotics, and it popped up in my subscriptions. I even went to Battlebots in Vegas this year, which was a pretty cool experience. I'm planning to start printing some antweights and get my feet wet. The knowledge gained from working on the OpenRacer project and other FPV/Drone design work should have a lot of crossover, as @Maker's Muse said. Pretty cool to see TPU making its way into use here.
@JonE5FPV @d3c0de I recognise those FPV names :) I saw another pilot, metaldirtnskin, on MakersMuse's Prusa i3 Mk4 unboxing livestream recently.
Designing drone parts is a big reason I got my Prusa i3 Mk3, having fun learning how to design small prints which work with the 'grain' of how it's printed to get the most out of the minimal filament used, Bardwell's shown & recommend my prints for quads on his channel about a dozen times the past couple of years which is nice.
His reaction to my 3D printed TPU props was hilarious, he couldn't quite believe they worked so well, neither can I actually, specially as they were the first print of the first iteration of the design!
This is so serendipitous. I was just at a Maker Faire on the weekend and I watched a beetle weight competition. This would be so cool, because you could have multiple chassis ready to swap out.
@5:59 when it flips out like that it brings my mind to a startled kitten ;D
I have to say you did really well for your first battle bot competition
So some tips for making weight easier, carbon fiber top plate it adds the rigidity and is not heavy at all. you can probably get away with a 1mm thick one.
ditch the heat set inserts, they weigh a ton you can use #4 plastice screws. the self cutting type.
you can get really fancy with prusa slicer settings. like removing the top layers in the battery compartment, or changing wall line counts in areas you need more or less.
also the more compact you make your bot the less it weighs. it looks like you could have reduced the length of pancake a good bit since you had a small diameter blade
look at getting a blade made from 1mm grade 5 Ti its a lot lighter but very strong.
That thing about the inserts is really worth thinking about. Now the problem with self tapping screws is mostly when you have to disassemble and assemble things over and over, but with some care they can be used without having them chew up the material every time you assemble it again. But I wonder if anyone makes aluminium inserts? If you don't want to use self tappers alu inserts should be good enough to replace brass inserts and probably at half the weight.
Also the lighter blade is a good idea. At this small size the momentum of the blade won't be massive, so the motor spinning it is more important than having a heavy blade. Or at least I think that's how it works.
suggestion: paint/print it green and add googly eyes! the little guy has so much personality leaping around the arena that he DESERVES to be a cute little frog!
You may want to try encasing everything inside a TPU case: print the bottom half, install the electronics, print the top half. This way you'll save weight on screws and inserts. Design your robot with disposability in mind, these tiny robots get so banged up in a weekend you'll end up printing another one anyway.
TPU is awesome. use the hardest shore rating you can for firmness. usually the cheap stuff. that's been a go to for quadcopters for years and we crash it hard. arm covers, antenna mounts, gopro mounts and cases.
This was the video that got me hooked on combat robots :)
This is SO cool! I made a combat robot in college in the early 90's and the competition was on ESPN2.
0:52. Same reason rock slide barriers are nets instead of concrete walls. Shock absorption and increasing deceleration time result in less force being translated to the rest of the device.
Modern cars cringle in crashes for the same reason. They are designed to collapse in order achieve the same effect.
I LOVE the Straubing idea ... It just drifting so we'll looked crazy good
Been flying racing drones since 2015....TPU is used heavily to protect our GoPro's, different parts of carbon fiber and to hold our antennas. So using it in smaller battle bots makes perfect sense
I imagine TPU-based Armor will hit the point of inversion of effectiveness once it's taken into Featherweights or heavier, since flywheel-style weapons ramp up in power significantly (probably exponentially) the higher up the weight classes you go, especially once you get to the full size arena classes. Once a flywheel gets a solid bite on a material, it will try to shear it away
Honestly, I like seeing the really well-built micro BattleBots it brings her unique challenge to the table that the large ones don’t
Pancake must evolve & become the ultimate bot.
Quite a cool innovation, possiply with the too plate have it sceltonized in places to increase flex, reduce weight, add airflow to help with cooling. Possibly add a vent next to the weapon as its spunning fast enough to push air around.
You should give TPE a go for the frame, its a lot better with repeated stress and shearing. I've used this as a stopper for airgun gas rams which experience a lot of abuse and its held up great. Its basically a rubber block that the ram crashes into so it doesn't damage the metal.
Its better to use raw and uncolored materials as I noticed ones with certain colors seem to be weaker or tear easier.
First 3d printing oriented channel to hit 1 million subs!! Well done mate :D
Angus - What about running a 1mm steel wire embedded in the TPU for added strength around the perimeter of the chassis?
Steel wire is heavy and how do you plan to embed it in the TPU?
@@conorstewart2214 you would design a channel for it in the model and then add a pause to the print before it would enclose it so you can add the wire. Or alternatively you could just design a hole through the length of the chassis and insert it when it's done printing.
@@conorstewart2214 pause mid print I've done carbon fiber strip implants in my prints that way.
@@conorstewart2214 heat the wire with electrical current and just put it into the tpu wall
I love these! now I have so many more ideas to print with TPU.
Awesome video. I love this content so much
Excited to see your take on a single wheel drive robot. Seemed like losing a wheel helped improve control.
I love it! An evolution on your 1-wheeled idea: If your single driven wheel was mounted as close to centre as possible, and the chassis sat low enough to the ground as to not be teeter-tottering, you could consider having that single wheel given a 180-degree sweep of steering articulation. Think of the handling characteristics like you would a bumper car, or even a ship's azipod. With a slip ring, you might even be able to achieve an infinite sweep of steering articulation. I imagine the slip ring's commutation would have to be on the outermost periphery of the wheel "pod", though.
I'm really excited for this robot. Can't wait to see the version with one wheel.
Try printing the tpu a few degrees hotter for better layer adhesion. It has worked well for me in 1lb antweights. A few degrees makes a big difference. I also do like 4-5 walls and that seems to work well. I have also done just a tpu bumper around a pla+ chassis and that worked fairly well, but if it gets hit just right, it tends to be pulled off. I actually just printed a very similar 150g undercutter but I am attempting to use a much harder tpu. I think its rigid enough I won't need to use any pla, but we will see what happens once it goes into combat.
Builder here and I say tpu is God's gift to combat robots. I recommend a carbon fiber tpu sandwich for construction. Metal is only for weapons.
I have been making lightweight panels for my rc car. I havent tried the foaming epoxy yet, i think it'll make a huge difference:
1 - secure aliexpress 0.2mm fibre sheet to print bed.
2 - print with minimal gyroid infill, 0 top or bottom layers.
3 - using a syringe, measure out an appropriate volume of foaming epoxy, taking model volume and the advertised expansion ratio into account.
4 - place another prepared sheet on top.
I believe the process could be used to make simple curved parts too - you can flex the panels into place, temporarily secure with superglue then inject the foaming mixture. I think this could make a great baseplate alternative for this combat beast.
I would try removing the ridgid plate all together. Use TPU for the whole assemble, this avoids creating stress concentration around the joins between the flexible and rigid parts.
Not watching a lot of your videos, but combat robots are awesome! Would love to see more of it!
Seems like the perfect use case for Ninjatek Armadillo, I love that stuff when printed right it's indestructible like other TPU but, much much stiffer.
Armadillo is mostly too stiff, and usually cracks or breaks in combat. 95A Cheetah with high wall-count is the sweet spot in TPU for combat robots.
Something that is worth looking into is the different TPU types/blends. Some TPU can be quite rigid, which might be a better choice for these combat robots, since you might be able to make the whole robot out of it. What I don't know is whether a more rigid TPU would also be a lot worse at absorbing impacts, though it would still be a little flexible. AS for the top plate, CF PA is a good option but I would also be interested to see PETG, since that can be relatively flexible (compared to PLA) in my experience and that might help it survive.
We are exploring the use of TPU and TPE for one off gaskets. The TPE seems to have better layer to layer adhesion.
One idea I'd recommend is using a stiff material inside and only where absolutely necessary. Think of the stiff material as a spine and the TPU as muscle. Print the TPU and press the very small amounts of PLA/PLA+/PLA-CF into the TPU. No need for a wide piece, and keeping it narrow, the stiff part can be thicker to be more rugged.
I literally said “holy f*****g s**t” when you said the weight limit was 150g. What an incredibly interesting weight class, I’m honestly astounded by the capabilities possible with such little allocation towards total weight.
Good video! I like battlebots with the spinning disc's for weapons.
They're a lot of fun and very chaotic!
Rubber robots can absorb impacts but the compliant surface allows a spinner's claws to get a better grip and throw them harder. What I'd want to try is having a multiple hard shells and the electronics mounted to a soft inner frame. You get the best of both worlds: hard surface to glance blows from spinners and soft inner frame resistant to bending which cushions the electronics.
I would try to replace the PLA plate with something like carbon rods as skeleton. Then put a relatively soft TPU body around it to absorb the energy of hits. Lastly (if possible because of the weight) add thin metal scales to the convex parts of the body to prevent cuts.
I use weird TPUs quite a lot to make surgical simulators. I found that a good way to make it stronger is to run a soldering iron on the outer surface of the print in areas of high stress. This connects the layers together and prevents delamination… hope this helps 👍🏻
Brilliant. I've switched to TPU for many of my functional parts that need to be durable on my ebikes, including a massive frame mount battery enclosure. I should try to build my own and share, I have ideas on how to make a good TPU frame for such a combat robot.
don't use CF nylon. The fibers primarily add stiffness, not strength. If you're going max toughness, use pure nylon (actually, that might be even better than TPU for durability too. You should try it)
idk if it will be of any help, but for squishy materials ive had the best luck in terms of weight:strength ratios using circular infills. try to set a infill % in which the corners are well supported by having the smallest fillet and largest infill circles possible.
the circles tend to absorb the shock by folding/squishing rather than pushing the infill walls back rather than crumpling then bouncing out.
back when i was testing this, i ended up making egg cradles and doing a classic egg drop test with the only change being the infill type. although bouncy, the circle bois had the highest drop height and survival rate.
My thoughts include not using PLA as support for the weapon. Instead, let it flex, but to avoid cutting into itself add slight PLA or similar to the exact part that would bend for it to hit itself. Potentially is a rebar kind of fashion, with thin bars of PLA threaded inside the TPU to offer rigidity without adding much weight (Currently assuming PLA weighs more, otherwise printing sections in PLA (melted to connect to TPU) might serve even better). Or potentially even thread actual metal nails or mesh as rebar, if light enough.
Instead of 1 wheel+1 weapon, use 2 weapons so it compensates for itself trying to spin like a reaction wheel. Then either have the weapons be the wheels, or add 1 wheel for "forwards/backwards" (with Yaw being set through modulating the weapons spin-ratio so one overpowers the other, turning the robot as desired as a reaction-wheel).
I would recommend designing it so the expensive parts have their own area and are isolated. This way if the battery or the ESC blows up you have a much lower chance of destroying the main board which is most expensive. Also I feel like you could lower the size of the front motor and still have most of the power.
A bit late here, but three thoughts. 1, look at floor buffers/scrubbers for motion inspiration. When operating one you lift/lower the handle to get it to swing left/right and experienced users can give it a slight twist to get ti to move itself forwards and backwards. 2, as much as possible remove anything rigid. That plate, by virtue of it being rigid, was taking on the strain that would otherwise have been dissipated through the TPU. Instead do something like printing a motor mount and electronics shell only, and insert them into the TPU like you would a captive nut in a print. 3, to minimize any corners enemy weapons could catch on, consider making the entire perimeter a continuous circle.
To summarize: if I were going to do an iteration on your design I'd make something that looks even more like a pancake, fully circular where the only external moving part was the weapon (can't loose a wheel you don't have). The part that didn't overlap would be thicker to act like a "foot", which would act similar to something like soft robotics and use servos to flex the foot and manipulate how the weapon spindle end touches the ground to effect rotation and forward/reverse motion.
Even more interestingly, if the weapon spindle came out both sides you could make it agnostic to which way it was flipped. Being upside down it'd reverse your flexing inputs, but it would also reverse the rotation direction in relation to contact reversing the effects as well and making your inputs have the same result no matter which way up it was.
What an awesome video. I would love to see more about robotics from you.
And could you please make a video about TPU and support? ❤
Great video and great robot, I love the design. I love the idea of a one wheeled one in the future. I've gotta get into this stuff in the UK :). I've attended a few events a very long time ago. Now I've got a 3D printer ;) .
I forget the full story, but somewhere around the 50's (or maybe pre-war) an engineer at MG took a drill to the chassis of their race cars to lower the weights. I think he also did it to few wrenches too. So put some "lightening" holes in the rigid plate, and maybe the body.
Speed holes
I know nothing about combat robots, but I *love* Colorfabb's Varioshore TPU. I have printed so many different types of functional and high stress parts with it. For example, I printed custom 2-in-1 doorstop/catches for a pair of double-swing cafe doors. These things take a beating daily as they stop the doors from hitting the wall as they are swung [and sometimes kicked] open, while also catching them in place to keep from swinging back shut. A simple flick of the latch and they release/close. A singular piece of printed TPU attached to the wall with sufficient wall anchors and they have lasted over a year so far now with no signs of deformation or wear.
I do have a thought about the robot, tho. Think of all those living hinges out there for other materials, then think of how much better they would be in TPU. Perhaps instead of normal slicer designed infill, you design the infill into the actual model using a thin accordion-type leafing that would compress in exactly the ways you want, while being rigid where you need it. Slicer infill is fine for most prints, but I find sometimes it is necessary to design the infill yourself as part of the part and leave the slicer out of the equation so maximal control over placement is possible.
How about a partial duct around the weapon disc and adding a propeller shape to its spokes in order to create suction for better "ground adhesion"?
Have you tried using polycarbonate for frames or reinforcement plates? It is supposed to have a very high impact strength.
Another flexible material you could try is polypropylene, it is pretty flexible and very durable with excellent layer adhesion. For the most part if you don’t need it to be very very soft and flexible then polypropylene is probably better than TPU. It is excellent for things like living hinges because it just doesn’t break. I think I have seen some with an elongation at break of something like 600 %.
Such a fun weekend and had a tonne of fun in that fight with Mr Obvious. Cant wait for the next one
Wow, Great design.
Love the little fight5
between the heavy weapon motor and the heavy electronics, the relatively thin connecting piece is always going to flex, no matter what you try. I'd design a certain amount of flex into the robot and give it an articulated top plate.
Why is your solution “make the flex worse”
@@user-cg3em4cw4f You mean better?
The bot wasn't good enough at flexing, so it broke. Increasing the flex makes perfect sense.
@@mifiamigahna No, the "articulated" idea was god awful and stupid. Flex is always an issue with horizontal spinners, but you don't see anyone else make a fucking ARTICULATED arm to hold the weapon because if you do that, it literally makes the entire weapon useless. Source: have actually participated in one of these events
You might try a polycarbonate frame, surrounded by a plain nylon bumper that you soak in water for a few days. Poly is really rigid but can still take a hit, and soaking nylon makes it gain a lot of flexibility. String trimmer line often recommends doing that because it takes the hits better.
I run a TPU chassis for my 3lb. Can whole-heartedly endorse using thread-forming screws (McMaster has some specially for plastic that I like). I find they're lighter and hold better than heat inserts. I’ve only ever had one get ripped out in almost 3 years of using them (they do get bent when they get hit)
My step brother blew the esc in his crawler a couple of weeks ago. The smell wouldn’t go away as he was changing it out outdoors, the day after it blew.
Let the combat robot content come man, i like it alot.
I've been printing with TPU for a little bit in my 3lb robot, Radix. TPU is fantastic for absorbing hits. My drive has never gone down before. The biggest challenge I've faced is hitting my own weapon motor with my weapon when my frame flexes. I think I've fixed this issue, but we will see at NHRL in June.
I got so tired of dealing with the rigidity issues that I made a hub motor!
FPV freestyle enthusiast here, our crowd & the drone racing scene would say more walls but honestly no idea how to manage that well given the weight requirement. I suppose you could just go 1-2mm laser cut carbon fiber top plate with a more "open" cage like design like we do in the drone scene. Then you might be able to get away with a couple of strategically placed spots with more TPU walls.
Very interesting, would love to see more on the subject.
Don't have any tips but I am always amazed how tough TPU is. I am using single wall TPU end caps for the darts of my big bore blow gun.
The only time they get busted is when I hit the cap with another dart. I have never had the layers separate. This stuff is crazy.
PETG & TPU will stick together very well in a multi-filament 3D print I discovered, changing filament at different layers.
I once tried printing thing-3565827, "3D-Printable Filament! -Print Your Own Filament for Multi-Color!" out of PETG & TPU, the resulting print from that dual filament sort of worked, I think the % blend of the two needs to be adjusted & tested more to see what mechanical properties you can get out of the final blend.
Just cut a lot of triangles in the chassis. Keep the wheel protection and improve the connection between the esc and the motor. There are some pretty good three pin connector out there that can prevent these things from breaking apart.
You should do a medium or heavy-weight next, it would be epic!
Haha 😅 would love to if I had the space and budget! Those machines are pricy!
it would be cool to see you cover it in the knife-proof fabric on top of the TPU
Like a kevlar or similar? That's a really interesting concept!
Love the casual Furdu merch
I think a dome like shape made of this stuff would have a really high defence stat on it's own. Add a flipping over mechanism with a cutter and it should be awesome
It's worth experimenting with modeling thin voids in your parts to force 'infill' generation selectively, reducing mass and facilitating optimized deformation characteristics.
PP has 75% of density of may have some weight saving there Also doesn't crack tends to just stretch and layer adhesion is the top of the class
why even have walls in places? open gyroid infill would be a nice cushion.
That's a mad idea, i'm gonna try it haha
Could you use a multi material printer to add a PLA core to the TPU?
For sure. A multi material print would be a really clever way of adding stiffness only where you want it.
@@MakersMuse zero thickness voids where you need more layers too (easier than using modifiers?)
@@MakersMuse enjoy that rabbit hole buddy 🤣
Also could spring steel within the print. Make a pocket for it, stop the print, drop the spring steer, and resume the print.
@@iteerrex8166 multi material + carbon fibre plates.
Idea: 2 weapons motors side by side and a wheel in the middle. Rotate by diffing the weapons speed, forward/back with the single wheel. A bit like you do now, but with 2 you'll have control before the weapon is spun up, and the gyro is a bit more manageable.
Any material can shatter with a high enough impact energy. A measure of how resistant it is, is the 'glass transition temperature' - effectively it's freezing point.
For TPU this is around -40C. One of the best options in Polymers is Polyethylene. HDPE for example is around -95C.
This means it can survive a higher energy impact and still behave as a plastic, deforming rather than like glass, shattering.
HDPE Filament has fallen from favour, but you can still buy it. Just a thought.
I have first hand experience of making (heavyweight) robots from HDPE Plastic (Barberous 2). While we had loads of other problems, the shell was only once ever pierced. The previous version was polypropylene - and it did indeed shatter like glass!
I also have an 150g horizontal spinner but I use a 3mm carbon fiber plate machined in the top, to secure the weapon and motors.
I know part of the benefit of that PLA bottom plate is protection when it's flipped, but since its mostly for stiffness I wonder if you can get away with a couple of thin titanium round bar stringers? You can get really thin bar stock online in Australia so it's worth having some on hand anyway
Been using TPU as airsoft armor and it's absolute beast tanking BB hits up close.
I loved these micro combat robot format due to its affordability.
If you want to make your armor lighter, create holes in the bottom and the top, somewhere you don't expect to be hit.
Holes are a great way to remove material to balance weight.
wow, that damage in the tpu really surprised me.
If you're cleanability regarding supports has gotten that much better, I would give it a try to print it in a 45° angle (since split layers were part of the damage and weapons seem to hit horizontally and vertically mostly)
I know weight is at a premium, but I'm wondering if potting the electronics and\or some of the wiring (maybe in a silicon) would provide some strain relief.
I haven't made a combat robot, but my experience with TPU and Foam disc replacements for nerf blasters has been really positive.
I've noted that using small air holes in the print can make the entire structure more durable as air pockets will just sort of... fart out.
Fewer walls tends to be better anyways to get the most flex, and infill as low as you can possibly get it will also help that. More flex means less tearing, though I'm not entirely sure that'd apply to getting whacked at high speed by another robot.
If you can, having a 3 dimensional infill also helps keep it structurally sound, but keep in mind all the pockets need holes so you'll be drilling/cutting more of them.
In the end if your TPU looks a bit like swiss cheese it should hold up to some pretty ludicrous abuse.
Before watching the video yet, I can see how TPU would be great for impact and could be good (especially combined with other stuffer materials as well), but I would worry about it being grabbed and thrown more easily by weapons. Seems like it could be great to have a combination of TPU and some PA6 to provide a similar effect to some phone cases to get the best of both worlds which could be good.
Mini combat robotics is awesome coz it's relatively cheap and accessible, and you can do more iterations of a design in shorter time!
More combat robot content, I love it!
Maybe a composite armor, made of alternating layers of TPU and PLA, could withstand more impacts and cuts. Some bulletproof glasses are made of tempered glass and polycarbonate. It should be possible if your printer can handle multiple materials.
TPU is just fun to use, trial and error is all I have for you depending on your use case. Your use case is, needless to say, much different than mine.