[Shonky]

Hi, I've been lurking around for a bit, but decided to share my build, and figured it was better to get input sooner rather than later.

So, as with a lot of these builds that are a bit different, I have something at my disposal which makes me able to do this in a way that isn't really feasible for the majority of people. I work at a medium/large manufacturing company that makes very large stuff out of steel. Anyway, what it boils down to is that I can pay cost price (which is cheap when you buy as much as they do) for the square footage of steel, and then get it laser cut for free.

I'm going to use A572 grade 50 .5" thick structural steel for the frame of my new toy. I'm going to build it with tabs so that it nicely slots together, and should hopefully be reasonably dimensional without too much shimming. It will cost me ~$8.80 per square foot.

I believe that it should stand on it's own without too much interference, but for extra support I will put some tack welds or epoxy it together at the end when I have everything where I want it. I might also use brackets and bolts. I haven't decided on that yet.

I ordered my ballscrews and linear shafts+ bearings off the ebay about a week ago, so they should arrive in 3 weeks or so. I bought them from a seller called momo19830131. I couldn't find a whole lot of info on him, but the one youtube video I did see was favorable. I'll let you know how they turn out when I get them or I put them to use I guess.

On the left-right axis I have put an anti racking system, as described here: http://www.cnczone.com/forums/diy-cn..._solid-13.html We shall see how it goes, but I probably should should have just spent another $50 on longer linear shafts.

On the Z axis I have put an extension up the top with some optional mounts for if I find I have too much slop there. If I am doing thinner, stiffer materials I will take the upper linear bearings off and attach them at those holes to give me greater stability. I will also get around to making a brace for the top there.

The "box" up the top there will also have sides to it when I am done.

At the moment, I intend on purchasing a 381 oz/ G540 kit from automation technology for the electronics of this build unless someone here has a better suggestion.

Total budget is around the $2000 mark including software. Here is my spending estimate so far:
Bought:
mechanicals $562
anti racking parts $30
Yet to buy:
electronics $500
frame $200
Hitachi M12VC spindle $120
precisebits collet set $90
--------------------------------
total so far =1502

I'm sure that number will grow by the time I'm done.

Haven't decided if I will spend on Mach3 just yet or try my hand with LinuxCNC first.

Total travel should be around 500x600x300mm minus whatever it takes to put reliable end stop switches in. In "less Z slop mode" Z travel will be reduced to 150mm.

I'll attach some screenshots of my current design. There's a lot of sky-hooks and things not finalized yet, but it's well on it's way to being done. I want to wait until all my parts arrive and I have measured them before I complete it and send it out to get cut. I'd rather wait a little while than end up paying for it twice. Here's a quick gif showing how the frame should go together, but I'm not sure how long this link will live, and uploading animated gifs doesn't seem to work on the forum. http://gifmaker.me/ViewGIFAnimation....put_fCgSY4.gif

Any suggestions are welcome, and I'll try and keep this build thread somewhat updated with my mistakes and adventures for the good of humanity and all that.

Alex.

[embraced]

I'm working on a similar design in which there are tabs that slot together in laser-cut steel, all fusion welded at the seams.

On your design, I think 1/2" thick steel is beyond overkill. A system is only as good as it's weakest component - in this case, I believe it's the linear rails. Those supported round shaft rails will not keep up with the frame if you intend on machining anything particularly difficult.

If you do not have a need to take extremely aggressive passes in materials, you may be better off reducing the thickness of the frame.


I wish you luck anyhow, and look forward to the result!

[Shonky]

Hi embraced, thanks for your comments. You are almost certainly right about the frame being very far from the weakest component.

I may switch to 3/8" thick, but to be honest, I'm mainly doing 1/2" because it should make the construction fit together a little easier, and the guys who cut at my work seem to have the settings down really well for the 1/2". It usually looks cleaner than the others. They also cut more of it, so it's easier to get done when I want, and the price difference isn't that great to make it worth going to a thinner gauge. If anything, I may put some weight reducing holes into the design.

FWIW, the .5" is $8.80 a square foot, and the 3/8" is about $6.60 from what I remember. Not sure if it's worth it or not for the $50 savings. I guess the main benefit really comes down to moving the thing around when it is done haha.

As for what I am hoping for with the machine; speed isn't a big issue. The only real advantage to me taking deeper cuts is to go through less end mills. I do hope to be able to mill aluminium, and maybe the occaisional small thin steel stuff, gears and the like. My biggest issue there is probably going to be my spindle.

In regards to tolerance, if I can get .004" (.1mm) I will be pretty happy.

The extra height on the Z axis is because I hope to put a 4th axis on it at some point.

Hope your project goes well too! You should start a build thread too so I can steal your ideas see what you come up with.

[kevincnc]

Interesting design, and I nominate you for best thread title.

[harryn]

The 1/2 in thick steel is perhaps about right for what you are trying to do, especially since there is no "depth" in some areas beyond the steel plate thickness. Al cutting forces are fairly high.

It might be interesting to add in some ledges to deal with cutting fluid, especially if you are really planning to do much Al work.

Using the selected motors + screws drive ratios, how much force are you expecting to have at the cutter ?

[Shonky]

Thanks Kevin!

The 1/2 in thick steel is perhaps about right for what you are trying to do, especially since there is no "depth" in some areas beyond the steel plate thickness. Al cutting forces are fairly high.

It might be interesting to add in some ledges to deal with cutting fluid, especially if you are really planning to do much Al work.

Using the selected motors + screws drive ratios, how much force are you expecting to have at the cutter ?

I was originally thinking of putting some channels onto the table, and perhaps drain them at one end. The other thing I was thinking is that I could make a box to bolt to the table with a container for fluid, but I am a bit worried about the moving axis and fluid slopping all over the place. I do like your suggestion of some shelves. I can also get steel bent fairly easily at work too, so I could probably make some nice light gauge steel shelves to direct coolant to a container underneath. Thanks for the suggestion!

As for the force, I am fudging it a bit because I can't find the step in torque of the motors anywhere. It should have a fairly safe holding force over 3000N(if my postit note calculations are correct) based on the motors holding torque of 381oz-in. But as for the actual force from a stopped motor, I wouldn't even know what I could guess at with that. I'm purely going off other people's advice to other people on this forum.

Do you have any advice for me there?

Thanks for the comments!

[harryn]

I am still a student to all of this, but especialy for Al, I don't have a good idea of the cutting forces.

I attempted to do some measurements using a 1/2 inch wood router bit on a bosch 1617 evs router and (crudely) measured around 50 - 70 lbs of force needed. This is possibly conservative, as most people talk about lower numbers.

The stepper motor ratings are based on "at rest, but full power". In order to reduce motor heating, the G540 reduces the current after X seconds at idle position, so your actual torque is less.

The torque in a stepper motor drops off as the RPM increases, so for many, at 200 rpm, it is about 1/2 torque. If possible, it is good to try to keep your stepper motor in the 2 - 300 rpm max.

What is the diameter and lead of your ball screws ? In other words, how much linear motion per rotation ?

[dmalicky]

Interesting concept, and nice to have that laser cutter available! $0.40/lb is remarkable for steel, and mass never hurts (except maybe one's back). The vertical gantry supports are plenty stiff, of course. If you want a more even distribution of stiffness to cut aluminum, I'd suggest some reinforcements...
- More beef to the spindle clamps (that's a major source of flex in my FEA models).
- Add 'legs' to the Z plate so it's a channel cross-section. The legs need to tie the spindle mounts to the bearings. Or use a channel section.
- With the large Z clearance, much more horizontal separation for the Y bearings. Cutter forces in the Y direction create a torque on the Y-car; the anti-racking device will help, but stranded cable is not all that stiff. A rule of thumb for the 4 Y bearings is to have at least as much horizontal separation as vertical. I'd get new Y rails and redesign the frame to accommodate, unless reduced Y travel is ok.
- Legs for the Y plate, too, since it's so tall.
- A box section for the main gantry cross member. Even 1/2" steel plate is surprisingly flexy in torsion (cutter forces in the X direction). Have a look at the xls in this first link:
cnc machine stiffness calculator - MYCNCUK - The UK diy cnc machinists community.
Then check out the "Gantry Top Twist" tab. Basically, nothing can compete with a box section. You can try different plate or I beam sizes in the "Gantry Top" tab (cells in purple) -- for an I section, it takes a plate thickness of around 1" to compete with a 'wimpy' ~3x3x1/8" tube. I'd either switch to a tube for that part, see if they can cut 1"+ steel, or add another plate to the back of your gantry to close-out the section.

[Shonky]

What is the diameter and lead of your ball screws ? In other words, how much linear motion per rotation ?

The lead on the screws is 5mm so it shouldn't be too bad. With 200steps per rev, that should give me a resolution of .025mm if my math is right. If that 70lb figure is correct, even if my driving torque is 1/10 of my of my holding torque then I should have about enough. At least to do wood anyway. Thanks for the idea of the force required

Interesting concept, and nice to have that laser cutter available! $0.40/lb is remarkable for steel, and mass never hurts (except maybe one's back). The vertical gantry supports are plenty stiff, of course. If you want a more even distribution of stiffness to cut aluminum, I'd suggest some reinforcements...
- More beef to the spindle clamps (that's a major source of flex in my FEA models).
- Add 'legs' to the Z plate so it's a channel cross-section. The legs need to tie the spindle mounts to the bearings. Or use a channel section.
- With the large Z clearance, much more horizontal separation for the Y bearings. Cutter forces in the Y direction create a torque on the Y-car; the anti-racking device will help, but stranded cable is not all that stiff. A rule of thumb for the 4 Y bearings is to have at least as much horizontal separation as vertical. I'd get new Y rails and redesign the frame to accommodate, unless reduced Y travel is ok.
- Legs for the Y plate, too, since it's so tall.
- A box section for the main gantry cross member. Even 1/2" steel plate is surprisingly flexy in torsion (cutter forces in the X direction). Have a look at the xls in this first link:
cnc machine stiffness calculator - MYCNCUK - The UK diy cnc machinists community.
Then check out the "Gantry Top Twist" tab. Basically, nothing can compete with a box section. You can try different plate or I beam sizes in the "Gantry Top" tab (cells in purple) -- for an I section, it takes a plate thickness of around 1" to compete with a 'wimpy' ~3x3x1/8" tube. I'd either switch to a tube for that part, see if they can cut 1"+ steel, or add another plate to the back of your gantry to close-out the section.

Hmm food for thought there. Thanks for the suggestions! Incidentally, this is partially why I took so long to get back to you guys. I've made some significant changes.

-The spindle clamps and Z plate are unfinished for now, so I'll work on those more when I get to that end of things.

-Based on your recommendations, the top cross bar is now 3"x8"x1/4". Hopefully that should make it plenty sturdy.

-With the horizontal clearance for the Y bearings, I have put wings onto the plate they mount to, which will give me the option of spacing them out if/when it becomes a problem. That way I can keep them close for softer materials where I need the travel, and spread them out if I need stiffness. I put pockets into the redesigned uprights for the lower wings to go into when not being used, and the top will clear the side of the frame anyway. At the suggestion of my dad, I have also designed the machine to be modular, in that the uprights can be swapped out for shorter if I want more stiffness, or longer if I wanted to machine the top of a tall object, like an engine block or something. Might come in handy if I actually get any accuracy out of this thing.

-As the plates are now supported by a frame, I have reduced the width of the steel to 3/8" for some added lightness as Colin Chapman would say. It's pushing around 200kg/440lb at the moment. I think it's going to have plenty of mass for my needs. I'm already intending on building a custom stand to support it.

Thankfully the new modular system will make it easier to make it at work in parts, and then do final assembly at home. Hopefully some aluminium foil shims will be enough to get the thing straight.

Here are some new screenshots:


In other news, my slides and ballscrew assemblies arrived. They are as expected +-2mm in length, so I will need to adjust my design to suit. I kind of expected that. The bearing to ballscrew ends are a tight fit, so I might have to stick the screws in the thermal testing unit at work at -40c to get them to fit nicely. The linear bearings aren't great, but will do for now. The ballscrew nuts and thread are actually surprisingly good.

I also ordered and received the electronics. I got the the G540 and power supply of ebay and saved myself $50, but still got the motors from automation technologies. I've got them "running" with the mach 3 demo for now, and they apear to go back to the same place after being jogged and then "return to zero" being pressed. I figure that's about all I can really do for now on that end.

I'm waiting on some couplers from china still, but other than that, I just need to finish my design, build the thing, and get a router and I'm set. Boy, it sure sounds easy when I say it like that haha.