Thinking about building a small cnc mill for small steel and aluminum parts. I’d like to make the thing mainly out of aluminum extrusions. I was thinking that a fixed gantry design like picrel would be the most rigid, except I’d like to mirror it so there are diagonal braces on both sides and to increase the depth of my work area.Is this the most rigidity possible given the size and format constraints?
>>2950082I'd be concerned about the rigidity if you're trying to cut steel.
>>2950082dont most edamamebois that build 80/20 routers get disgusted with the lack of rigidity pretty quick and move on to something else? and theyre just trying to cut interwoven cellulose
>>2950082Those things don't even have enough rigidity for cutting aluminum.
>>2950082this thing might cut plastic, maybe some soft wood.you will spend more on end mills trying to cut steel on that than just buying a grizzly g0704 or something and converting it to cncactually is that a taig headstock on that thing in your model?just buy a taig or sherline imo.one of those will be infinitely better than this
/emt/ retards begone
>>2950082>Thinking about building a small cnc mill for small steel and aluminum parts. I’d like to make the thing mainly out of aluminum extrusions.You'll need to replace that with plate aluminum parts for steel or aluminum. you can start off with aluminum extrusions to cut / drill aluminum plate. or just buy a chink cnc machine and modify it with a chink 110V 1.5KW ER11 spindle.
>>2950082>steellolMilling steel on a router like this is going to be unbelievably slow. Like you can get a tiny feature into a tiny part over 20 minutes, while holding the spindle by hand the entire time to reduce vibrations and add preload. And it's still going to break at least one endmill per operation. Aluminium is much more forgiving but still painful. If you do not particularly NEED large parts, it's a very good idea to build the CNC around a machine vise, since it immediately resolves all your workholding woes. And then you should adjust the Z axis so it's the smallest reasonable distance away from the vise. Make sure to use beefy linear rails. The open bottom SBR12/SBR16-UU round slides have tension adjustment, the MGN types don't, but larger MGN rails are very solid.
>>2950082Your design looks like it is 8 times as stiff as my cheap wood-aluminium thing, but the table looks like it’s exactly the same as mine so that would be a weak point unless those two rails are some super tight tolerance hard steel kind of slides I don’t know about. And don’t you need some plates on the diagonal braces as well
>>2950082I’ve mostly just roughed out steel with these table top things then finished with one of those belt sander kits you can buy a hand file and a countersink to deburr
i've seen some mexicans make diy sliding tables out of slabs bolted together. do you guys think this would be rigid enough for machining on it?
if i bought two of these alibaba sliding tables and bolted them on an L shaped steel frame, wouldn't I technically have a milling machine? the top slide would have a router with a spindle under it. the spindle strapped in place by two bolted flange bearings. the bearings bolted on plates secured to the vertical sliding table. voila
>>2950584You would technically have a milling machine and practically have a big pile of junk metal.
>>2950584like so
>>2950592do they have ball screws or do they have threaded rod?if they have threaded rod, they are junk and not worth your time.What you should do is source 3 linear stages with rails and ball screws and build your own.
>>2950082Diagonal braces do surprisingly little. First of all, the base isn't perfectly rigid and, unless ridiculously overbuilt, twists under its own load by about the same amount. Second, by the time the gantry legs are beefy enough to not twist into a pretzel (a major weak point in many designs), they're also beefy enough to transfer the forces into the base without much need for additional bracing. For pic related the difference is less than 7% which is not worth the trouble imo.
>>2950082I was watching a weird chinese guy build a 4 axis like this in his 1 bedroom apartment, he had to keep it small so it would fit in soundproof cupboard (lol)Realistically i think the Not An Engineer / Kris Temmerman design is the only one that would work but what would i know, fucking nothing
Get a DMC2 mini.
anybody in the 1st world try out one of the chinkshit ddcs controllers? i see them on lots of ruskie builds but it seems like they have no look ahead so they slam on the brakes and creep around every g02/03 circular move. that seems like it would completely defeat any high speed roughing programming
>>2950898Need to make it out of steel if you want it to be good the stiffness of steel is around 2-3x stiffer than aluminum. Your best bet would to design it out of an epoxy granite base and 1/2" -1" thick plate weldments to hold the table and spindle assemblies. It needs to be super duper stiff if you want to be machining steel with any kind of efficiency, look at how much mass people add to their Bridgeport and things to adjust the vibrational frequent of the machine. Look at how much mass and stiffness there is on a modern consumer mill, and that is the weakest you should go
>>2953374>look at how much mass people add to their Bridgeportwtf are you talking about? poast proofive been in and around west coast job shops for 40 years and never once seen mass added to a bp mill. with the exception of insurance mandated bolt downs otherwise all of them were simply plopped on the floor and plugged in without complaint. and anybody that claims a bp will chew away with a 3/4 endmill buried 2" deep in steel is full of shit. theyre not a #5 k&t
>>2953374>Need to make it out of steel>Your best bet would to design it out of an epoxy graniteSo which one is it? FYI elastic modulus of EG is about 15% of steel, and while it's a useful material, it's quite a bit more of a meme than a lot of people love to imagine.
>>2950082There's videos of people shitting together 200usd builds capable of roughly cutting steel so shouldn't be impossible.
Use rectangular steel sections, bolt it together, and fill with either cement or sand.Aluminum extrusions are not rigid and are not vibration dampening, they're also too light.
>>2950898>Diagonal braces do surprisingly little.Not trying to be a jerk anon, but if you put diagonal braces where there is the least of distortion then you shouldn't expect much. Can you run another sim with braces protruding away from the bed?
>>2954728>with braces protruding away from the bedI don't get it
>>2954969Well the pillars show a ton of distortion to a side so maybe a brace there will to better than where there is less distortion
>>29501523020 or 3030? Genmitsu 3030 uses ball screws, not sure if it's worth the additional money.
>>2950898How did you set up your forces? It looks like you did it as a force in the +z direction, which for a drill press would be fine, but a mill will generally work more in the x and y then the z direction. It also appears most of the deflection is due to the z axis/tool head. Suggesting adding rigidity to the tool head/mount would be a good place to start rather then braces.
>>2955803It may help a bit but you'll need at least a similarly big tube or a stack of tubes in the base to resist the resultant forces and all the accompanying jiggamajig because the bottom of the base is not flat anymore. Depending on design it might also hurt the torsion stiffness of the base or introduce more welding and distortion or make aligning and keeping everything flat and square more complicated. Might as well just simply use a bigger tube for legs/pillars, which is sort of what already happened on that pic, was hoping to use the same tube for everything but turns out those legs are a major weak point. After seeing the simulations I now always kek @ chink machines with legs cut out of plate. Can't even begin to imagine how much they flex with sideways forces.>>2955954>How did you set up your forces?66% of total force on X axis, 56% on Y and 50% on Z. Yes it doesn't add up to 100% because it's a vector sum. Iirc got it from some sort of old standard on (manual) milling machines due to lack of better reference.>It also appears most of the deflection is due to the z axis/tool head.Yes X and especially Z carriages usually make a big part of total deflection due to mass and geometry constraints. There's no good way around it, you can only choose what kind of compromise to eat.
>>2955948You want the 3030 prover max with ball screws and linear rails.https://www.youtube.com/watch?v=f2c4BXAzRu43030prover max + something like a MAFELL Milling motor FM 1000 PV-ER or Sorotec SFM 1000 PV-ER spindle is literally the best thing you can buy for the price.OR go the super cheap route with 3030 prover max and chinese VFD/110V 1.5KW ER11 spindle.
this you?
>>2956457Thank you anon, same braincell
>>2950152>You'll need to replace that with plate aluminumWhile i am not a fan of aluminium profiles for cnc mills, these plates would be less stiff and more expensive and thus a terrible choice.The design of that entire machine is just bad and was done by someone clueless on stiffness, design for manufacture and economics.>>2956337>MAFELL Milling motor FM 1000 PV-ER or Sorotec SFM 1000 PV-ER spindleThese are terrible compared to chink ER spindles for 1/2 to 2/3rd the price since those have actual vector VFD torque regulation and are brushless instead of these garbage brushed universal motors with at best phase angle control.Please stop giving terrible advice.
>>2956579This is by far the most retarded post I've ever seen on 4chan
>>2956611He didn't say anything wrong though
>>2956457Was just about to comment the same.
>>2956321>but you'll need at least a similarly big tube or a stack of tubes in the baseI'm pretty sure you can minimize deformation with 4 extra square tubes per side. That being said, the original deformation seems pretty easy to model and predict so prevention or minimization is more expensive than a few extra lines of code.Also, the diagonal beams seem to be there to minimize the gantry's bottom screws' radial effort over the axial effort, as they have less strength on those axes
>>2958120There are countless ways to "minimize deformation with 4 extra square tubes per side" so just please draw a sketch next time.In any case here's what the above machine ended up looking like. There are some things that still need polishing but the frame is more or less final. It's about as rigid in the sim as a similarly sized manual mill designed to make massive cuts with big mills should be, and trying to beef it further gives diminishing results, so I'm stopping at that.
>blocks your path
>>2956337Milling motors are basically like routers, speed and on-off is not connected to the cnc at all and is just manual on the thing itself, right? VFD sounds better in this case.
>>2959481No if you actually searched the milling motor it has a cnc connection. God i hate this fucking site>hurdur dumbshit because i don't bother researching anything because i know everything
>>2959491>didn't see a question markWhat a cunt you are, sir.
>>2959449That thing is a toy.
>>2959494U r a nigger
>>2959515A nigger with a sherline which is about the same size but can actually cut and uses full size end mills.This is what the column on that toy looks like.
>>2959526This is the 'table'.
>>2959527This is the motor.It's a stretch to call this thing a mill.
>>2959439>It's about as rigid in the sim as a similarly sized manual mill designed to make massive cuts with big mills>i dont have a clue but muh computer program shit out some garbage ill preach as gospel why are green jr engrs so fucking arrogantly retarded?
>>2959552Can you show me on this doll where did the sim touch you?
>>2959526you are still a nigger
>>2959651Yeah we all know a tube is the most rigid shape, but LOL not that tube. It's fine for the tiny RC car motor they put on it but put a real motor and watch that thing shake itself apart.
>>2959439I thought about providing a sketch but I'll let you have as food for thought why that might be a bad idea if no one has shared here an arragement I can point you to.That being said, do you have access to topology-optimizing software? That would also be useful to figure out where and how to add reinforcements.
>>2960891>do you have access to topology-optimizing software? That would also be useful to figure out where and how to add reinforcements.I could probably find something to use but I won't bother for a few reasons:- I don't have the manufacturing capabilities to build anything that isn't primitive enough to just optimize manually by trial and error. It'd probably take more work to beat the program into submission so it won't propose me shit I can't make.- for that kind of optimization to be even worth the effort you need to have a very good idea about expected loads and type of work the machine is being designed for, and the ones derived from a simple testing procedure from and old standard for a different type of machine do not anywhere nearly cut it. Without that it's garbage in - garbage out.- I'm already halfway into making actual hardware and I'm not throwing it away just because there's something 5% better.
>>2961758Post pictures.I've spent the last month migrating a DIY parallel port based controller into a professional looking housing. It's nuts how a power supply, some wire, screws and a box have cost me like $200.
>>2961761>>2959439
>>2950898shit fea if you got only 7% dif in flection with and without the braces>t. senior structural engineer, ansys spec
>>2961808Naw it makes perfect senseThe forces involved (the X component specifically) are twisting the frame around the axis pretty much parallel to the braces so they don't do shit. They do help resisting the Y and Z forces but the frame is strong enough in that direction already and the majority of deflection happens elsewhere, so again they don't do much.
>>2959526>Big empty space you can fill up with molten lead, concrete, or epoxy sandit'll be fine with some mods
>>2962533That amount of lead or epoxy will cost you the same as saving a couple of hundred dollars more for a chinese mill made out of steel.
beer can cnc status check?
