Humble Beginnings - For Alu

Hello :)

I figured I best make a thread, it's good a place as any to start asking the myriad of questions and getting lots of criticism of my random ideas. This is going to be a bit of a living document (I'm going to change my mind when I feel like it ). You can click the pictures for bigger images.


Goal:

• To build a rig that can CNC aluminium designs

Must:

• Cut a 500mm x 500mm size (Ok, so I picked these sizes at random :P )
• Cut aluminium
• Cut with accuracy
• Cut with repeatability
• Take a beating

---

Entry 1: (26 Feb 2013)
I have been working my way through the build threads and the huge amount of info that's around, trying to learn the basics. While I've been doing this I decided to kick off some CAD just to get back in the swing of it as it's been a few years since I last used it in anger. So I drew up some parts from data sheets as practice:

http://i.imgur.com/GImdHcxm.png

NEMA 34, 15mm Rails


And then I started doodling in CAD, I started with 30x30 mild steel box section with 3mm side wall, soon binned that deciding it wouldn't be strong enough.

I've got to here so far, it's using 50x50 mild steel box section with 5mm side wall.

http://i.imgur.com/sdVuwhxm.png

http://i.imgur.com/i2VUWMFm.png

http://i.imgur.com/n7x6SVBm.png

• Bear in mind this is more of a concept than a design at the moment, but feel free to comment on it :)
• The idea is to have the screw in the channel between the rails.
• At the moment it looks like everything is welded, only because I haven't designed in any bolted joints yet, they will be there though.
• I am not a welder, nor do i frequently work with steel, so if it looks like I have drawn an impossible weld or cut, let me know.
• I made the bed 20mm Alu and there is another rail under it supporting.

Cheers,

Husky the noob

---

Entry 2: (27th Feb 2013)

Ok, had a play with some more CAD :)

Deleted the side rails & related
Got rid of the aluminium bed
Added more bars for a new bed
changed the rails for slightly longer ones
Added bolted section to allow fine tuning of angles


Some pics:

http://i.imgur.com/qIOlgECm.png


http://i.imgur.com/kEpuAj5m.png


http://i.imgur.com/vsCrAZam.png

---

Entry 3: (3rd March 2013)

Shortened the rails.
Changed the bed from being welded to bolted, not quite finished yet.
Changed the angles from box section to plate.
Added bolted gantry cross beams
Added bolts to attach gantry to rails. Had to enlarge box section on top of sliders to 60mm, 5mm wall.

http://i.imgur.com/uAKRsrUm.png


http://i.imgur.com/WMJagxjm.png


I need to get round to sorting the materials. I have the box section all set to mild steel, but all the fixings and rails are just set to a generic high strength steel.

Current weight of what's on the screen ~55kg

---

Entry 4: (3rd March 2013)

Ok, drawings changed.

Added bolts etc
Changed gantry runners
Changed gantry sides

http://i.imgur.com/LcPeihIm.png


http://i.imgur.com/JLiUEtYm.png

Like your drawings. For serious metal cutting I still think a fixed gantry would be better, and a much stiffer (double) gantry cross beam. G.

I'll shoot first. .Lol

Don't need Nema 34's 23's will be more than enough.

Don't need 2x rails per side they won't add much to the machine other than expense. Linear rails can handle massive loads so one rail with 2 bearings will be fine.

The Single piece Box on Gantry is NO where near strong enough for cutting Aluminium and will need massively beefing up.

The 20mm Plate for Bed is Good but will still need a Decent frame building under it and 1 cross piece won't be enough.
Don't under estimate the forces for cutting Aluminium.! . . To achieve decent quality of finish the machine needs to be very very ridged.

What you have drawn so far won't cut Aluminium overly well and full fill your criteria.!! . . . . . More beef required, Esp on the gantry and don't be afraid of the weight for cutting hard materials it helps.

Good thing I haven't drawn the gantry yet :joyous: I'm still scratching my head on that one, the bar was kind of a place holder, don't panic haha.

Ok, I was worried as I have never used the runners before, I shall go forth and delete the outside ones! Lets me get rid of those angles too, that would have been a pain to line up right in the build.

Yeah I'm going back down to the 23's. I drew those before I really read into it and got some advice. They still look rather pretty though, so I put them up anyway.

On the subject of the bed... Yeah, I'm not sure why this happened with only a 20kg load, but...

http://i.imgur.com/DHxwdN3m.png

I pressed a wrong button somewhere I think :very_drunk: But still, I'll put more bars in as you say!


I'll do a fixed gantry version next, I just can't get past the fact it'll be double the size of the cutting bed, minimum.


Thanks guys!

I,m sure you pressed a wrong button. Not quite double the frame (not bed) length, but much easier to build a really strong machine IMHO. G.

Too many Desperado's me thinks!

I admit to being biased!!! G

Quote:

---

Originally Posted by GEOFFREY

I admit to being biased!!! G

---

I agree 100% Geoff fixed gantry is Much Much better option for cutting hard materials if you have the space.

OK, talked to a mate at work regarding my results shown above. Basically he said to look at the results like a scaled graph, it's scaled up so you can see it more easily but the actual numbers are the important thing. So, it could be that it's only deflecting a tiny bit, but scaled up it looks horrific. Also, I used a distributed load rather than a point load, I kind of knew this at the time but couldn't find the correct option, just wanted to see some results :P

Looking forward to chopping it apart tonight for the next version.

You've set it to display stress, which is useful when determining if the material will fail, but that's clearly not the case here as the maximum stress is small. Also, you need to be careful at corners or near bolts as the stress outputted by the simulation at these boundaries can be a lot higher than reality, due to some of the assumptions in the calculation. We're interested in the deflection so set the display to displacement and find the maximum points. The results you get will depend greatly on the boundary conditions, so how you attach the bed to the frame in the simulation is important. The mesh quality on the simulation is very poor too, but there's not much you can do about that. If you're going to try and use FEA, then at least start with applying forces to the gantry as the gantry is going to be far less stiff than the machine bed even without additional supports. For the current simulation you needn't include the gantry and X-rails, so you might as well remove them as a significant number of the nodes in the mesh are consumed in these parts.

The extra two rails wont make much difference because the vast majority of the force will be apparent on the other rail, with two blocks, since the structure in between is not particularly stiff. Just use one rail on each side and space the blocks out as far as you can, since the stiffness is proportional to the block spacing which means the deflection is inversely proportional to the bearing spacing squared, so increasing the spacing a little can make a big difference. That's assuming the structure they're mounted to is very stiff, which currently it isn't.

Just a quick stupid question, what software do you use to analyze stress? is that part of the AutoCad package?

Regards,
dsc.

I'm using Autodesk Inventor Professional 2012 :)

Ok, had a play with some more CAD :)

Deleted the side rails & related
Got rid of the aluminium bed
Added more bars for a new bed
changed the rails for slightly longer ones
Added bolted section to allow fine tuning of angles


Some pics:

http://i.imgur.com/qIOlgECm.png


http://i.imgur.com/kEpuAj5m.png


http://i.imgur.com/vsCrAZam.png

I'm guessing this is the Spindle I want to go for :)

WATER-COOLED SPINDLE MOTOR 2.2KW AND INVERTER VFD DRIVE FOUR BEARING | eBay

Yup that is the one.

If you are mounting a metal frame on top of two bits of what looks like wood... Don't.

Wood swells and expands depending on moisture content, the arrangement you have will warp the frame and throw your rails out of alignment. You also have your linear rails going over a joint. Don't do that either, unless you are going to take your frame to a big engineering shop and have the whole thing surface ground you will have zero chance of fitting the rails properly with that arrangement.

Ok, wood is out :thumsup:

Ah ok, good point about the rails! I'll shorten them back down.

Shortened the rails.
Changed the bed from being welded to bolted, not quite finished yet.
Changed the angles from box section to plate.
Added bolted gantry cross beams
Added bolts to attach gantry to rails. Had to enlarge box section on top of sliders to 60mm, 5mm wall.

http://i.imgur.com/uAKRsrUm.png


http://i.imgur.com/WMJagxjm.png


I need to get round to sorting the materials. I have the box section all set to mild steel, but all the fixings and rails are just set to a generic high strength steel.

Current weight of what's on the screen ~55kg

Will there be more frame work to that base frame.? If not then then you'll need to be sure what ever it sits on is 100% flat and true other wise that frame will easily twist and put the rails on different planes.
Profiled linear rails don't tolerate twist (Or bend) very well and will bind to some degree. Doesn't take much to lock them up solid.!!

One other thing I see thru experience of mostly cutting Aluminium is that chips get fired every where and with this design the profiled rails are in direct firing line. They will act has stops and chips will collect building up in the profile of the rail. While the bearings will mostly push these out the way some of them will get thru and into the bearings, which over time will considerably reduce the bearings life. Over time It will also have an affect on finish quality because it will reduce the bearings smooth running causing notchyness that passes thru to the cutter.!

You could get away from this very easily with a slight design change to how gantry mounts to bearings.? Instead of using the short piece of Box section has riser then make that piece full length of the machine and mount directly onto the frame. Put the linear rails on this then mount the gantry to the bearings directly using a plate to replace the short box section.
The plate could be 20mm Ali or 10mm steel either would work ok. Personally I'd use Ali plate has welding thick to thin steel creates heat distortion and make the Triangle brackets out of Box section along with larger Triangle plates for strength. (see pic)

This one change will give several advantages other than moving the rails out of direct firing line.?
#1 Increase travel 100mm because rails can now go right to end of machine.
#2 Make aligning rails parallel and planner to each other easier by simply shimming or adjusting the Long piece of box section.
#3 Make gantry mounting and bolt access easier.
#4 Will be stronger and looks neater.
#5 Easy future height clearance upgrade if required by simply inserting another piece of box section or spacer between bed frame and linear rail box section.

Hope that helps.

Quote:

---

Originally Posted by JAZZCNC

Will there be more frame work to that base frame.?

---

Yeah, Originally I planned to lay it atop some massive ass bits of wood. But it seems that this is a bad idea... So I need to draw up some legs and cross braces etc for it now.

Quote:

---

Originally Posted by JAZZCNC

You could get away from this very easily with a slight design change to how gantry mounts to bearings.? Instead of using the short piece of Box section has riser then make that piece full length of the machine and mount directly onto the frame. Put the linear rails on this then mount the gantry to the bearings directly using a plate to replace the short box section.
The plate could be 20mm Ali or 10mm steel either would work ok. Personally I'd use Ali plate has welding thick to thin steel creates heat distortion and make the Triangle brackets out of Box section along with larger Triangle plates for strength. (see pic)

---

I had to read it a few times to get what you meant but figured it out in the end haha.



Thanks for the info, I'll get drawing! :)

Ok, drawings changed.

Added bolts etc
Changed gantry runners
Changed gantry sides

http://i.imgur.com/LcPeihIm.png


http://i.imgur.com/JLiUEtYm.png

Well hello there people (yes oh dear it's actually been more than a year wow). So I got caught up in renovating my house (still doing) and lots of life things. I've been wanting to get back to this for so long now and I can't resist staying away any longer. I've just got to figure out where I was!

So I loaded up my old CAD files and gave this thread a re-read to try and get back in the swing of it. I prodded CAD for a bit and had a nosey round my design.

What I can see is a heck of a lot of bolts... What do you guys think about them? My thoughts are that they are a prime point for the frame to slip out of alignment, perhaps I should really just put them all down as welds? I think my initial idea was that bolting them would give me the ability to fine tune the angles and alignments when built to get them spot on, but a fresh look is making me think twice about it.

Secondly, chopping up and tapping all those aluminium blocks is going to be such a big task i may as well take it somewhere to get it precisely welded anyway.

Here's a picture (because we all love pictures) of the design as it stands, I made it all rusty to represent the year of neglect :numbness:

http://i.imgur.com/Ptxmomp.png

I think it's not so much the metalwork that needs lining up but rather the support rails upon that metalwork. This is where the epoxy method etc. comes in to play, obviously the metalwork needs to be rigid so welding might be a good idea for some of the parts anyway.
Maybe you should have a look at some build threads that have been posted in the last 12 months, you might get some other ideas.

weld without a doubt ...

those bolts if overtightened will crush the box section .... and even if they are torqued up right, with vibration the frame will flex causing the bolts to come loose (admittedly you can lock tight em in)

id weld it up slowly, then epoxy the face