OK here we go..! Wish me luck but I must warn you this will be a slow build... An eye candy from the latest stage of the design:
OK the design is almost finished.. It seems it will be; Y(travel) = 815mm Y(cut) = 800 mm X(travel)= 720mm X(cut)=720 mm Z(travel)= 160 mm Those numbers might change slightly but looks like this is what it's going to be. And another render from where the current design is:
An idea about vertical and horizontal fitting vise to Openbuilds v-slots/v-slot machine beds. Model details and downloads from GrabCAD are here.
I have a very similar machine idea floating around in my head. A little larger X/Z travels, but same setup and modularity/upgradability. The great thing is, something as beefy as this can be built for around the same price as a pre-configured kit, just takes a little more work. A view with just the extrusion beams with no covers would be cool as well.
I don't understand how the X and Y rails are constructed. It looks like they are boxes made of sheet material. I must be missing something as that won't be very strong/rigid.
Good call. I thought there was extrusion hidden inside the sheet metal. Kinda confused now. Those gantry plates are also looking about a third to a quarter of the thickness they likely need to be, sans any other reinforcement.
Hi, thank you for your comments.. They are very valuable for me.. I was thinking most of the load/strength will be needed on two part types.. One is the back plates (the red parts on this draft) and gantries. I might go for steel for those two and add 90 degree extrusion to connect back plates to v-slots (machine bed) 6 mm alu gantries do a good job with C-Beams.. But then again ofc steel would be much more sturdy and if gantries are seteel back plates will have to be steel too.. The yellow parts are just alu sheets to make the box and they connect to back plates and everything else with corner connectors. They do not receive weight or pressure on them at all. I am more worried about the feet to be honest as I have a Y connector down below and I can only use feets on the outer edges of the bottom plates.
What thickness are you thinking for the back plates? If I were you, I'd build up a test beam, attach a lever (proxy for the Z beam) at the center and see how easy it is to deflect the lever. I'd be very concerned that it will flex too much.
I honestly don't know, that engineering is beyond me. If I were doing it, I'd find some one who can do that kind of analysis.
Not sure I'd be slapping "Machine Design" with my logo all over a bunch of pretty renders if I hadn't done some basic mechanical testing and analyses, though. Sure does look pretty, but it's gonna be making oval holes and snapping tooling all over the place.
Well, true I like designing and making them look pretty. But not going to start building once I am 100% sure it is not going to make oval holes and tool snapping all over the place. Please note that this is not a commercial product or what so ever. This is me playing around only as a hobby. Would appreciate some productive input as I believe thats one of the philosophy of "open" building. Sure I could easily take this to bunch of friends in the University but then it wouldn't be so much fun (at least not for me).. Thank you for your "input" anyways as now I realize you are right.. This could be misleading.. Will change the pictures when I have time.. "EDIT: Deleted the medium with "machinery design"
Good answer. The fix is relatively simple: stop assuming your linear rail has any structural integrity whatsoever. Now design a machine frame that's going to hold it solidly in place. It can't flex in any direction or on any axis, for which plain sheet metal is the worst possible performer for all of these tasks. Now if your "sheet metal" was actually 3/4" aluminum bar stock, machined flat on both sides, maybe that would work. Unnecessarily expensive, though!
Thank you. In my mind I was not relying the integrity of my linear rails.. I was more relying on "back plates" as those are the ones that holds everything together. However your comment makes it obvious that sheets are not going to do it. In this case I am going to use c-beams again and come up with a design that keeps everything in a box for clean operating. (I also like the idea of having a polycarbonate "window" to see the goodies move while operating.
Fwiw, my design in my head is C-Beam based. It's the best bang-for-buck I've seen out there, and if it's rigidly held down to a work surface, the weak point is likely the lips holding the M5 tee-nuts in, not anything else. It doesn't look as cool as a truly custom design, I know, but without access to a medium/large VMC- VF2 or larger- it's difficult to create a long, strong, reasonably flat (+/- 0.001"/0.025mm or less) surface. Welded rib/spar construction would work well and provide mounting flanges, but again it would need machining flat at the end.
True.. Maybe its better to go for a smaller machine.. As you probably know I design, build and donate the machines to tech high schools so that they can use them for learning (like i do) and using for their projects.
I've used C-Beam before and it works great. The machine was a success.. C-BEAM-GTC I would like to go for something else this time.. Maybe something like this? 2 20x60 v-slots between 2 10mm 7075 plates. 1 has a recess for housing the linear rail as it is critical for 2 reasons 1 to have them lay in parallel, two help the machine to absorb the vertical or horizontal force.
Those are quite complex plates. Recess grooves on one side and undercut T-slots on the other. Do you have the capacity to manufacture those +/-0.005mm? If you can't manufacture precision, you have to manufacture adjustability. The linear rails will need sweeping in with a test indicator; best practice is probably to assume they're going to come wiggly and figure out how to make them accurate. So, it looks cool, and I like the overall strength, I just don't know if it's manufacturable.
I think simple grooves are enough with tnuts. It would be interesting to see an FEA of the assembly to see how it handles torsion.
Easier too yeah.. I am investigating fusion360' s analysis capabilities. As you can define the type of the material and surfaces, moving parts, etc it comes up with results showing the weak spots. Not sure if it will help but worths a try.