/robowaifu/ - DIY Robot Wives

Advancing robotics to a point where anime catgrill meidos in tiny miniskirts are a reality.

LynxChan updated to 2.5.7, let me know whether there are any issues (admin at j dot w).

Reports of my death have been greatly overestimiste.

Still trying to get done with some IRL work, but should be able to update some stuff soon.


Max message length: 6144

Drag files to upload or
click here to select them

Maximum 5 files / Maximum size: 20.00 MB


(used to delete files and postings)

Welcome to /robowaifu/, the exotic AI tavern where intrepid adventurers gather to swap loot & old war stories...

CNC Machine Thread Robowaifu Technician 05/12/2020 (Tue) 02:13:40 No.2991
Many of the parts needed to build our robowaifus will need to be custom made and they will need to be metal. For parts that have a high tolerance for imperfections a 3d printer can print a mold and then a small scale foundry can be used to cast the piece with metal (probably copper or aluminum). BUT there will be pieces that need a higher degree of precision (such as joints). For these pieces a CNC machine would be useful. CNC machines can widely range in size, price, and accuracy and I would like to find models suitable for our purposes. I know there are CNC machines available that can cut up to copper for under $300, but I don't know if that will be enough for our purposes. (https://www.sainsmart.com/products/sainsmart-genmitsu-cnc-router-pro-diy-kit?variant=15941458296905&currency=USD&utm_campaign=gs-2018-08-06&utm_source=google&utm_medium=smart_campaign) Some CNC machines can be used to engrave printed circuit boards and that may prove useful for our purposes as well. Are there any anons that know more about CNC machines? Anons looking to buy one ask your questions here.
>>2991 You're definitely right about this OP. Now once someone steps up and starts offering kits for sale, then small factory runs of such parts would be included obviously. But for those of us who are the true pioneers here blazing this trail, a decent CNC machine is practically a given. Thanks.
Open file (62.84 KB 1151x825 rip economy.png)
>>2991 I've seen videos of people building their own CNC routers that can cut aluminum gears precisely. So long as you're not using cheap chink aluminum it should suffice. Low grade aluminum should not be used for parts under any load since it will easily break in surprising ways no matter how thick it is, but the high grade stuff is durable and strong enough to take a beating. Using pure steel will make your robowaifu weigh a ton. Copper corrodes easily and is not a good idea for joints. Brass might be a useful material but it's slightly heavier than steel and I imagine there are nuances to working with different materials. I've only worked with aluminum, silver and steel before. Don't forget to factor in the cost of replacement drill bits, buying metal to machine and shipping if necessary, although the price of metal has fallen considerably, kek. The last time I checked, where I live it's cheaper to prototype parts in plastic and order them in steel via CNC machining services than to invest in some equipment myself to make my own aluminum parts. Unless my trading bot strikes it rich I'm gonna go with what's cheapest and support local business.
>>3035 Are duraluminum or aluminum bronze usable for joints?
>>3036 No idea, but I know aluminum is corrosion resistant. Wikipedia says aluminum bronze is pretty good. It might be worth looking more into: >Aluminium bronzes are most valued for their higher strength and corrosion resistance as compared to other bronze alloys. These alloys are tarnish-resistant and show low rates of corrosion in atmospheric conditions, low oxidation rates at high temperatures, and low reactivity with sulfurous compounds and other exhaust products of combustion. They are also resistant to corrosion in sea water. Aluminium bronzes' resistance to corrosion results from the aluminium in the alloys, which reacts with atmospheric oxygen to form a thin, tough surface layer of alumina (aluminium oxide) which acts as a barrier to corrosion of the copper-rich alloy. The addition of tin can improve corrosion resistance. High temperature should also be considered selecting material. I cook with cast iron pans and it corrodes very quickly in contact with water at high temperatures if it's not properly seasoned. >Aluminium bronzes are most commonly used in applications where their resistance to corrosion makes them preferable to other engineering materials. These applications include plain bearings and landing gear components on aircraft, guitar strings, valve components, engine components (especially for seagoing ships), underwater fastenings in naval architecture, and ship propellers. https://en.wikipedia.org/wiki/Aluminium_bronze
>>3061 Very interesting. I'm assuming it's more expenisive to create the alloys. Do you obtain them as cast stocks? The use in guitar strings intrigued me, since one of the mechanical actuator approaches we will probably need for torso==>major_joints (shoulders, hips) probably needs to be some type of a pulley/cable system approach for lightness, and for keeping the heavy drive system well-inside the torso for kinematics optimizations & balance needs.
>>3062 If you have a good enough foundry you could probably just cast your own stock.
>I'm going to go ahead and repost the bunker comments here as well.
>>3067 So, Anon was saying in the waifu CNC thread that if you have your own foundry, you could make your own brass/aluminum alloy stocks. The question I had was; what kind of up-front setup costs would be involved in the micro-foundry to smelt these alloys? Could anyone possibly produce small runs of this stuff for machining that would be cheaper than you can just buy the alloy itself?
>>3080 You need a steel bucket, a pipe, and fireclay. The fire clay you need will need to be able to withstand temperatures of up to 1300 C and your foundry will use charcoal.
>>3081 OK, now we're talking. We need to keep things as simple and 'primitive' as is feasible for each discipline if we're going to successfully see this in the hands of Anons everywhere. If we can literally meet metallurgy needs in so straightforward a fashion then that's one off the list, so to speak. I'm assuming the fireclay is both fairly accessible and form-able by Anon? Also, 1'300C is pretty effin high. Just charcoal is sufficient for that?
>minor bump for anon to see the moved comments
Charcoal will do it and you can find bags of fire clay on amazon for like $50. >/// I should add that it needs to be real wood charcoal. Not the briquettes; they have limestone added to them so they won't get too hot.
>>2991 Here is a hobby CNC that can do aluminum and in theory up to 4mm steel (not recommended, convert a manual mill instead) https://www.cnc-step.com/cnc-router-1000x600-s-1000t-ballscrew/ Expect to pay an additional $1000 on tooling on top of the machine price itself. Here is an detailed review of the machine in German: https://www.precifast.de/mechanik-stabilitaet-und-genauigkeit-meiner-high-z-s-400t/ inb4 muh unsupported rails are shit yes unsupported rails are worse than profiled rails of the same size but the unsupported rails on this machine are extra large to compensate for that https://www.youtube.com/watch?v=JoI-nnipoxo https://www.youtube.com/watch?v=sISfwGV1VLU https://youtu.be/HYoNGTt67lk https://youtu.be/C30nRmlP4ak https://youtu.be/G3Mcw8HrC6s If this is too expensive then let yourself get chinked: https://www.omiocnc.com/products/x8m-cnc-desktop-engraver.html No guarantee that this machine can do anything at all but hey it's cheap. (find reviews/cutting videos yourself)
>>4797 Thanks for all the video links Anon they really help finding important things.
>>4797 Thanks. I saw several sources over the last few years claiming that at least wood can be handled with cheap hobbyist machines. I won't go there currently, no space, noise, other things to do and no real usecase for working with wood. But good to know that this exists and metal working might also be possible. >Tools for 1k Is this for metal or same for wood?
I have basically no knowledge on this topic, but this here strikes me as informative: https://youtu.be/EaGFQ7M04Wo and its exactly about what I'm going to need at some point: Cutting metal sheets of copper and aluminum for combining it with plastics, to make parts stronger or run current through them. Fazit: It's a build, not a machine to buy, he modding it with a new motor and some changes to the electronics. 400$ and quite some work, limited usability, professional software necessary for better usability bc toolpath. Maybe for thin parts it's good enough, he didn't elaborate on that. His conclusion is 1000$ are necessary to get something useful. Sanladerer also build something which might be useful for my usecase: https://youtu.be/ovrtvLFvFpk - I watched it a while ago, but don't remember the content. After all, this here >>3035 might still be valid: >where I live it's cheaper to prototype parts in plastic and order them in steel via CNC machining services
>>8422 On second thought, and after going through the comments: To me the video was interesting, but the headline is misleading. Also, he never says at what thickness of metal the machine failed. Others pointed out, that a "single flute bit" would have been required, I don't know if that's true same goes for other advices.
Anon studying advance manufacturing. At is best if you spend the money to get a bigger and better cnc. Unless you are going to make small parts out of easy to cut materials then it would be best if you spend the money to get a better cnc than cheaping out.
>>9183 Agreed. Any kind of machine tool is a capital investment. Hopefully it will last you a lifetime.
>>9185 At my college we have two lathes that have been there for over a decade and they run better than the new lathes. They are forgiving for newbies and have more power. Spend the money in a good machine and its will take whatever you throw it it.
>>9183 I'm probably only going to buy a small one, for cutting sheets of metal. Don't know about the best price/gain ratio yet. Thin layers in plastic parts could conduct electricity, thicker parts could add a bit strength. It's rather about 1-2 or maybe 3 mm thickness. It isn't urgent anyways, I'll look further into it when I buy one. For bigger machined parts there seem to be services where they can be ordered, similar to those companies which are etching PCBs.
>>9194 As some of the industry guys have told me, don't buy anything Chinese or overseas unoess you cam vouch for their quality. And before I came to study cnc I was a tech at a injection molding factory and the saying holds true. Chinese products are too finicky. Also there is https://www.voltera.io/ if you want to male homemade PCB.
>>9197 >I was a tech at a injection molding factory That's interesting. Would you mind having a look at our threads Plastic Production >>108 and Waifu Materials >>154 , and giving us all the perspectives of a tech at a injection molding factory Anon? I'm sure you would have some insights that those of us completely inexperienced in this field would benefit from. Creating injection-molded external shell parts, as well as mass-produced plastic frames, connectors, fittings, etc., are all likely needs for us in the future. Small-scale production techniques will need to be worked out in these beginning times ofc. One thing I actually did already was work on a heat-lamp + dry-vac vacuum-molding rig set up with ventilation for our attic at home. Little ideas like that would be very helpful at this stage.
>>9199 I found another gem. A Kickstarter 5 axis mill. Honestly I have higher hopes for it since it would bring more power to us and let us have more flexibility. I have a few concerns like how sturdy it is or how hard of a material can it cut? But honestly it it cuts aluminum then it is good enough. Biggest problem would be the part or the mill shifting mid run. So i want to hear your guy's thoughts. https://www.kickstarter.com/projects/2003668803/5axismaker-first-ever-affordable-5axis-multi-fabri?ref=discovery_category
>>9314 That is a very neat project Anon. However, it's possible there may be some issues with the project management itself. Did everyone receive their units and everything they paid for? I would closely examine everything before I gave them any money. BTW, since this kickstarter was a few years ago, maybe there's a similar project out there by now?
>>9197 Yeah, people working in some industry with expensive tools will always say things like that. Of course I would first look at some YouTube channel I trust before I'd buy anything. In 3D printing, the ratio between price and quality from China is mostly quite okay for non-industrial use. If I had to pay expensive engineers per hour, I might buy something else of course. Big companies think different than smaller shops or hobbyists. >homemade PCB I think I saw a review of this printer and the conclusion was that it's not so interesting for most people, since PCBs are so cheap to order in small series nowadays.
>>9328 "Don't Build a Metal Foundry Until you See This First" https://www.youtube.com/watch?v=l2FuvKTyRMQ
>>9328 That's a very nice looking metal. >>9329 That guy makes it look easy. So, it looks like it's possible to craft custom actual metal parts on a small scale at home then?
>>9330 >custom actual metal parts on a small scale at home Yes, didn't watch the vids here yet, but a few years ago there was a talk on that on Chaos Communication Congress. They melted Aluminium in some foundry, I think. Printed molds, put them into sand if I recall correctly and put it into a microwave for some reason. The microwave was used outside. Quote ~"If something goes wrong, you don't want to deal with molten Aluminium on your kitchen floor." - So it appeared to be more interesting for people living a bit remote (fumes?) or at least in a house. I think the microwave also had to run for some hours. This whole thing is also more interesting in USA than in EU, because their soda cans are actually made out of Aluminium while in EU it's Tin or something. Old bicycle wheels were also mentioned as a potential source for the Aluminium.
>>9331 >because their soda cans are actually made out of Aluminium while It's true. While there's not much there at all (maybe a gram or two), it's probably close to be 100% pure aluminum. There's an entire industry surrounding recycling these things here. >t. Burger
The best all around metal casting book is by C W Ammen. "The complete handbook of sand casting" Here's a link for it. http://libgen.rs/book/index.php?md5=044722E0EE1D11BFEF4E4F654FAAAF30 He has several other books you can find at Amazon that are also very good. Now this is old school tech going all the way back to B.C. so it doesn't take some unobtainable technology to do this but it does take some practice. Aluminum would probably be the easiest metal to use. You can melt aluminum in steel pots.You could use old car parts like aluminum intakes and stuff like that for metal. Aluminum cans can work also but there's a lot of oxidized metal in the cans so you would have to be careful about skimming off this oxidation. One thing he doesn't cover is lost foam. Lost foam may be the easiest way to may highly intricate parts easily. They use this in the auto industry and many other places for high volume but it's just as good for a few parts. Basically you make whatever you want cast in foam, pack sand with a little clay to hold it together and then pour the metal into the foam. The foam instantly vaporizes and the metal replaces it. There's also lost wax which is much older. You make whatever you want cast from wax, melt it out then pour in metal into the open spaces. All the ancient statues you see were made this way. For making furnaces David J. Gingery makes a great book. Several of his books and his sons are here including the charcoal foundry to melt metal. http://libgen.lc/search.php?req=David%20J.%20Gingery&column[]=author There are several propane furnaces on Youtube and the web that might be more suitable. He also has great make your own shop machines series which is good but it may be the open source multimachine would be better. here, http://opensourcemachine.org/
>>9455 Thanks, great info Anon. We may need this kind of thing for more than just robowaifus some day heh. Just in case, like me, others can't get to the site you linked for us: https://web.archive.org/web/20210325152108/http://opensourcemachine.org/
>>9461 >>9462 >>9463 >>9464 >>9465 >>9466 >>9467 >>9468 Apologies for stretching the dump out like this instead of making it more compact /robowaifu/. All of these are US Army machining training guides.
Since this thread is unofficially for everything related to metal parts and especially alternatives to the CNC machining as well: This https://youtu.be/v44bEsL3TCo shows casting of metal gears in some kind of gypsum after using gelatine and glyzerine and then wax to copy a gear with a silicone casting technique. Downside is, you need a furnace and work on it, also he was still using a machine towards the end, to clean it up. But it looks easier than machining the whole thing (except if you have a good industrial CNC machine). Then again, don't forget that it might be possible to order parts from online services doing that for you. Then again, we wan to be independent (as an option).
>>10515 Thanks Anon. That seems reasonably accessible with some experience and effort. >Then again, we wan to be independent (as an option). I would suppose that during the earlier years of robowaifus, hobbyists would rely quite heavily on 3rd party vendors for practically everything, one way or another. As entrepreneur-minded anons step out to begin their own manufacturing operations this would gradually form a different balance. Big industrial concerns like a Mitsubishi will very likely be producing most everything mechanical for a robowaifu themselves.
>>9469 No need for apologies anon, this is a lot of quality content! Very practical and in quite some depth and detail. I can think of several other groups on the internet such as survivalists, homesteader/permaculturalists and gunsmiths who would be thrilled to have these manuals. So thank you very much for sharing!
3D printed parts can be used for metal sheet forming: https://youtu.be/fxzqkhmcRlY
>>10588 really impressed, Anon. I admit I was a bit skeptical when I saw how he was going about it, but I was really impressed that he seemed to mostly pull it off. I think if he had worked on creating some sort of compression jig first to keep the dies rigidized as they cured it would have been a bit better, but my overall verdict: >really impressed. :^)
>>Aluminium bronzes are most valued for their higher strength Trust me when I say that the more difficult to machine, the more cost prohibitive. You're best sticking to aluminum and no space age alloys. >>3080 It is not cost effective to be trying to create foundries to smelt materials especially alloys if you have no idea what you're doing. You can buy all of this material through material suppliers. There are many local and online based ones you can buy stock of all sizes and materials. >>4797 Shapeokos can cut aluminum but with hobby machines its much more about fine tuning tool paths, feeds and speeds, and the specific tooling you use. You have a lot less wiggle room. >>9314 By the look of that machine, trust me that thing won't cut dick besides foam and plastics. If you want a 5axis cnc that can cut aluminum at least you'll need a pocket.NC. Only super cheap available 5axis desktop mill I've seen so far.
>>11217 >If you want a 5axis cnc that can cut aluminum at least you'll need a pocket.NC. You seems to have some understanding of this field Anon. Would you mind expanding on that please? What is a pocket.NC, and why is it good at cutting aluminum? TIA.
>>11217 Thanks for that, but I didn't understand all of it. Are Shapeokos also hobby machines or did you mean it so, that these are more professional ones? (I found their website) Do I need something like that just for cutting some millimeters of sheet? >>11218 Not the same anon. YouTube: https://youtu.be/tV5lDZOUeyY https://youtu.be/kn34-LqDKQo
>>11222 >digits Ah, I understand better now. Thanks Anon! Now I'm curious how one goes about programming such a machine? Is it anything like the files you simply load into a 3D printer?
>>11225 Never used one, but for all I know: Yes. You can create files with a CAD program like Solvespace for example. Though the format might be different.
>>2991 >Many of the parts needed to build our robowaifus will need to be custom made and they will need to be metal Source?
>>11250 Not the OP, but it's obvious. Plastics are rather weak and can get brittle or wear of.
>>11250 >>11362 a graphene + epoxy resin compound was one idea being floated around, might be more useful as a type of internal "cartilage" than as structural support I posted a while back on Aluminum Bronze, which is low enough melting temp to be forged in a home foundry (several YT videos on this) IMO we'd want to avoid something ferro-magnetic. Also as light weight and corrosion resistant as possible (aluminum is a contender here), we're not at the point where it matters if our waifu can rescue us from a burning building or something.
>>11362 >but it's obvious. More x=x logic from "ur a troll" guy I think
>>11374 e.g., the plans for my waifu involve a build comprised solely of plastic printed parts, mainly polycarbonate for structural support. This is fairly easy and inexpensive to build an amateur 3d printer for
>>11375 >comprised solely of plastic printed parts, mainly polycarbonate for structural support It requires a better printer than other materials. It's not lighter than aluminium in relation to strength and parts like gears would wear down over time. Maybe you can do without metal parts or only using standardized metal parts, but it clearly has advantages to be able to use custom made parts. >>11374 You didn't address the problem that plastics are weaker and wearing off, so it was an appropriate answer, and yeah it is obvious that at least using some (custom made) metal parts has advantages. Maybe more so for the higher quality versions. Not necessarily many custom made ones in every case.
Open file (298.52 KB 963x1625 IMG_20210703_201416.jpg)
>>11410 >you may even be able to put some kind of metal plating for mechanical parts that interact That was one of ideas of using some custom made metal parts. Yes. >metals aren't very cost effective There are standardized parts, which are very cheap and companies that mill custom parts have been mentioned as a source in this this thread. >for the average user Some here want to build very cheap robowaifus, others might go for the more expensive version, which might be able to be build as a cheaper version with some drawbacks. >metal 3d printer That's only one option, there's milling, casting and ordering custom parts from companies. Also these printers might get cheaper at some point. >comparative stress-tests for materials CNC Kitchen on YouTube.
>>11417 What happened here? He deleted his posting I replied to?

Report/Delete/Moderation Forms

no cookies?