1.5m by 1.5m cnc at rigidity
Hey everyone
I had this bug in my head for a while - i want to build a large cnc at home (1.5 by 1.5 meters) on the budget
Mainly to cut mdf panels but also sometimes aluminium and to try cut steel
I have seen on the internet some people actualy made some machines like that but there was no information one the accuracy of cut and rigidity
Sadly i cant find aluminium profiles in my area so i was thinking about steel square profile
Im alittle afraide the gantery can bend abit at such length
Do you think i can get 0.1mm or less accuracy (due to bend only) of cut, cutting mdf and 80x80x2 steel gantery?
Re: 1.5m by 1.5m cnc at rigidity
Hi MrR - Sorry will need to be about 200x200x6 prefer 8mm thick to do that. Peter
Re: 1.5m by 1.5m cnc at rigidity
Can you weld? If so, I might consider building a truss structure out of the smaller profiles you do have access to. Be sure to use whatever low-stress methods that are available to you, and afterwords stress-relieve it as best as you can.
Re: 1.5m by 1.5m cnc at rigidity
Hi MrRocket - This forum is full of discussion of people filling sections with stuff. They do make it heavier which seems to reduce noise (but it does not make it stiffer) but you must get the size of the section right in the first place. Everything needs to be much bigger then you think for the machine you are describing. Peter
Re: 1.5m by 1.5m cnc at rigidity
Steel box tubes are stiffer than aluminum extrusions, and cost less. If you build a machine with big enough pieces of box tube and pay attention to details, you might end up with a machine that will do an acceptable job of machining aluminum. Steel, not so much - take a look at mills with that part envelope designed for cutting steel to get an idea of the mass and rigidity required.
Sag in your gantry is proportional to the height and thickness of the tube selected; if you use a massive enough tube, sag won't be much of an issue, although inertia may be.
Re: 1.5m by 1.5m cnc at rigidity
And a fully triangulated frame is going to be stiffer than a pair of otherwise unconnected beams...
Re: 1.5m by 1.5m cnc at rigidity
You're getting responses that factor in the most extreme part of your request. If you just left it at mdf, you'd get a different response.
Re: 1.5m by 1.5m cnc at rigidity
OK, firstly; I'm not a mechanical engineer - everything I know is self-taught. With that in mind...
Quote:
what size of profile will be enough with this structure?
In this context, more is better. I would get the largest that you can afford; and I wouldn't go for less than 2.5mm wall thickness. The profile will need to be large enough to fully support your linear rails; so if your rails for example have a mounting width that is 20mm wide, make sure your profiles are at least 20mm, preferably 25 or greater.
Quote:
do you have any examples of people that built such diy machines?
Not that I can easily point to - I've seen them online over the years; but I'm afraid I didn't keep bookmarks to them.
Quote:
and wouldn't all these connection create a system more susceptible to vibrations?
Not exactly. There are several parameters involved here: deflection magnitude; vibration frequency resonance, and damping.
By cross-connecting the beams into a truss, they effectively behave as a single beam of larger size (yes, I'm horribly oversimplifying that - perhaps an actual mechanical engineer could jump in here?). When you do so, the deflection per unit of force applied goes down, the resonant frequency goes up, and I believe that the damping also goes up somewhat (but, adding something like an epoxy-aggregate filler will really boost the damping; at the expense of added weight on a moving part). Of all the ways to cross-connect them, my understanding is that a triangulated truss-type structure is the most efficient in terms of stiffness and strength per unit of material.
So, with all that in mind, what I think I'd do is get profiles of the largest size that's affordable (and I wouldn't go for a wall thickness of less than 2.5mm); and arrange them kinda like this (pardon the ASCII art, but the forum doesn't have a CAD component)
Side view; the X indicates a square box profile, the single lines ( | \ and -- ) indicate the triangulation cross connectors.
The linear rails would attach to the left side of that "drawing"; kinda like this:
.... And I just got interrupted. I'll pick this up again later. :)
Re: 1.5m by 1.5m cnc at rigidity
Hi MrR - It does not matter if you cut al or steel once a year or every day. The requirement to do this is far above MDF. For instance my router was not designed to cut metals. It will fly through timber, plastic and foam. But put a piece of aluminium under it and it will wobble and struggle. Plus the spindle speed is totally different for al and steel. You can cut aluminum if you use a small DOC say 0.2mm but if you want to cut a 16mm piece then you are there for a few days to get through it and you wear the tool out many times as its just using the tip.
Now bolted connections are good from a vibration point of view as they have friction so absorb motion. Welded joints transfer motion 100%. So if you really want to cut al you need to put everything on steroids which drives the cost and complexity up on your MDF machine for a few al cuts a year. In the end its up to you. If you design to cut timber and MDF you can use 25mm belts and save a lot of $$$ and be very accurate and happy. When you have an Al job to do find a machinist you will get a far better result. So keep thinking and designing, a long way to go yet. Peter
Re: 1.5m by 1.5m cnc at rigidity
After welding it; I’d grind any protruding welds flat; and see if I could get it stress-relieved as best as I could (from what I understand from asking about it online, heat treatment is best, but vibration of the right frequency works as well; and hammering on it with an air tool - needle descaler or an air hammer gun - is supposed to have some effect as well).
Then, considering your circumstances, I think I would probably use a very low viscosity self leveling epoxy. Build up a suitable set of “dams” (out of something that the epoxy won’t stick to) around the metal frames, clean the metal thoroughly, lay it flat (use the best level you can get your hands on), pour the epoxy, cover it so bugs don’t land on it while it’s curing, and hope for the best.
Um... where are you, geographically? Or more to the point, you can order stuff from Amazon, eBay, Banggood, AliExpress, etc, without paying insane shipping prices, right?
Re: 1.5m by 1.5m cnc at rigidity
Quote:
Now bolted connections are good from a vibration point of view as they have friction so absorb motion.
Um, if the force of the vibration is less than what it takes to overcome the clamping force exerted by the bolt, does the friction still apply?
Re: 1.5m by 1.5m cnc at rigidity
Hi Britt - Yes the load path is via the asperits and it is a known fact that bolted frames are damper then welded ones. Damp structures have tortuous loadpaths that's why EG is so good, millions billions and penta billions of internal load paths. Its how metals are damp as well, its the friction within the grain boundaries that dissipate the energy into heat from the motion.
Vibration stress relief although done a lot does not do a good job. Usually done with objects that are too big to put in an oven. Peter
Re: 1.5m by 1.5m cnc at rigidity
Interesting... I wouldn't ah' thunk it (software guy here... not a trained M.E.).
I suppose a question for MrRocket is... can he get that put in an oven where he is?
Hmm... I suppose one could put rivet nuts into the long beam profile sections, and bolt triangulation cross members to them on the back side (um... perhaps make the cross members out of C channel?). Some of the cross pieces that connect the upper left beam and the lower right beam would probably still need some plates welded to the ends so that there was a surface to bolt to... but that should be a lot easier to deal with than a 1.5 meter long welded structure. With care & appropriate temperature measurement equipment; I suppose you could even do the heat-treatment on those in a fireplace (adjusting their position in the fire so that they don't get too hot or cool too rapidly).
Re: 1.5m by 1.5m cnc at rigidity
Hi Britt - To do an effective stress relief the structure needs to be brought to 650C entirely then allowed to cool. If you did it in a fireplace and one area got hot while the other area was cold you would have a differential temperature therefore a differential thermal growth. On cooling this would further stress the object vs what you are trying to do. Plus when cherry red steel is quite pliable (watch a smithy work some red hot steel or have a go yourself with an oxy set quite interesting) so if badly supported the structure will sag under its own weight and you'd be an unhappy camper!! That's why buildings fall down in a fire the steel gets hot and losses all strength and it collapses under it own weight.
Best practice if welded is to take the frame to a heat treater and stress relieve, then finish machine. If not possible then you have to make do.
Mt Rocket and others - If you use channel you can design a frame that is entirely bolt up with no welding. If you weld make it just for the bolting flanges or webs. A channel frame would be very suitable for the task. I'd make the rail support structure as best as possible, take it to a machinist and have it surfaced. Bring it back and with a good machine level, fit up the rail supports level and square. With a bit of luck when you surface the rail supports they won't move much. Allow a good margin of metal for a second cut if needed. Normalised steel usually doesn't move much, ask the machinist for advice on metal that's suitable..... Glue the rail piece onto the support structure vs welding it on, in fact glue the whole thing together there are great glues out there now. Look up Plexus and Scotchweld...Peter
https://itwperformancepolymers.com/products/plexus
https://www.youtube.com/watch?v=4RCHd-PLhTM
You can weld your bonding flanges on, set up the table with clamps as you glue the bits together. There are heaps of automotive panel adhesives now which are easy to use and give great results. If you know what you are doing a simple toughened epoxy will be good as well. I'm amazed at how good construction adhesives are these days as well. Who uses nails anymore?? Cheers Peter
Re: 1.5m by 1.5m cnc at rigidity
Quote:
badly supported the structure will sag under its own weight
My dad is a civil engineer; and I’ve visited some of the projects he’s worked on. Several of them have been re-builds after fires; it’s quite eye-opening to look up at steel I-beams that are 10, perhaps 12 inches wide, and see them drooping down like cooked pasta.
Quote:
If you weld make it just for the bolting flanges or webs.
Sorry - that’s what I meant about doing in a fireplace - just the small connectors with welded-on bolt flanges. I suppose it’s not really necessary; but would be an optional step.
I’m operating under the presumption that he doesn’t have access to a machine shop (he can’t get aluminum profiles in his area?!?); but I still don’t know roughly where he is, so...
If he was someplace with more industrial resources, my thinking would be different... probably think about getting the profiles ground after milling, if there’s room in the budget for that.
That 3M adhesive is definitely interesting... although I think I’d still clean the area with acetone or isopropyl alcohol before using it. Hmm....
Re: 1.5m by 1.5m cnc at rigidity
Hi Britt - I suppose adding oil is a low priority option in reality :) But I agree I'd feel strange not cleaning a joint before gluing it. But MrR its entirely possible now to glue a frame together vs welding.
Here's the welder you want. https://youtu.be/rsTBPh2vKL4
Peter