[cmadki4]

Just another build log:
This will be my first from-scratch design and build of a CNC machine. I've had machines in the past that I modified or rebuilt. A plasma table that I made more rigid for routing and a router that I modified just to work properly. But those machines are long gone. Now the only machine I have on hand is a 3D printer and I've got the itch for being able to cut wood and aluminum again. No specific application for the machine, just wanting to design and build one.

Features:
- about 24" x 37" x 5" work envelope
- constructed from laser cut 1/4" and 3/8" plate
- controlled via Raspberry Pi 3
- programmable control panel buttons

Since this is a just-for-fun built I'm going to utilize as many of the parts I've had lying around for years as possible to keep costs down. Most of the main parts are either on their way, recently received or have been sitting in bins in my shop for years. I'll piece together misc. parts as the build goes on.

Main components:
- Hiwin 20mm linear guides and blocks (300/800/1150)
- RM1605 300/800/1160 ballscrews with DFU1605 nuts
- FF12 and FK12 ballscrew supports x4 sets
- KL-5056 drivers x4
- KL232100-35-4B 381oz-in 3.5A stepper motors x4
- Raspberry Pi 3
- Protoneer RPi-CNC v2.58 4 axis hat
- 48V 12.5A power supply (for steppers)
- 24V 5A power supply (for worklights and control panel)
- 5V 2.5A power supply (RPi3 and logic)
- Balluff prox switches x6
- Misc. e-stops and control switches/buttons
- 2.2kW ER20 Chinese air cooled spindle w/ cast aluminum mount
- FR-E520-3.7kW VFD
- iPAC2 USB HID (for control panel buttons)

Screen shots to follow...

[cmadki4]

Attachment 373784Attachment 373776Attachment 373778Attachment 373780Attachment 373782

[fretman_2]

Good looking machine! I'd make that webbing on the back of the gantry solid. Probably no need to save any weight there.

[cmadki4]

Good looking machine! I'd make that webbing on the back of the gantry solid. Probably no need to save any weight there.

Thanks!

Thought about that. That particular plate is 3/8" thick. Cutouts save some weight, but maybe not enough to justify.

EDIT: the cutouts constitute about 5 lbs of removed weight.

[LeeWay]

I don't think I would cut them either, however there are pro's and con's for both. Nothing actually weighing heavily toward in or out.
It does look like a good design.
Look forward to following your progress.

[mactec54]

Good looking machine! I'd make that webbing on the back of the gantry solid. Probably no need to save any weight there.

The way that he has this designed, is more than what is needed, having it as a solid plate would be just add more unnecessary weight with no purpose

[mactec54]

Thanks!

Thought about that. That particular plate is 3/8" thick. Cutouts save some weight, but maybe not enough to justify.

EDIT: the cutouts constitute about 5 lbs of removed weight.

Stick with what you have, 5 lbs. is a lot when you start moving it around, the design is more than what is needed to do the job, and much better than what most use, ( Aluminum Profile ) the webs are large enough, for support, and the top and bottom plates add the strength that is needed, I have made the same design in aluminum and it works very well

I would add Ribs on the gantry support uprights, and put the same rib brace on the bottom of the main frame as you have on the top

[cmadki4]

Thank you LeeWay and mactec54. Great to get input from the more experienced operators.

@mactec54, when you say 'add ribs on the gantry support uprights' do you mean add cutouts for weight reduction? And carry that zig-zag pattern through to the bottom and throughout?

[LeeWay]

5 pounds means not much on a steel frame machine using linear ways and ball screws like this one. Especially dual drive and a small size like this. That gantry will stop on a dime edgewise either way.

The major con I had was that the plate may not be as flat as you like after laser cutting those supports out. In a perfect world, laser cuts would not warp the metal. That is not always the case. Granted that a flat plate really isn't required there, then i think the pro's equal the cons, so either way will not be a bad choice.

If you make a hole anywhere in the uprights, make it so that you can eyeball the exact location of the tool in the spindle from the side.
There are other more precise methods, but often when cutting out the rest of a sheet or something, eyeballing is close enough. Line of sight is good to have.
I would not worry about weight reduction there though. In fact, you should add supports to the plate to help stiffen them.

[cmadki4]

I would not worry about weight reduction there though. In fact, you should add supports to the plate to help stiffen them.

Something like this in red?

[cmadki4]

Quick look at the brain I just soldered pins on. I have Mach3, but wanted to try using the Pi, so I'll be using bCNC. The Protoneer RPI-CNC hat will feed GND/ENABLE/PULSE/DIRECTION signals to the KL5056 drivers. It will also receive inputs from limit switches and a few of the control panel buttons (E stop, CYCLE START, FEED HOLD, ABORT) as well as control the VFD. Neat little package.
Attachment 373806Attachment 373808

[LeeWay]

Something like this in red?

That is the right idea, but insufficient. Most use gussets of some shape and at least a pair of them on each side.
Steel plate gives or has movement and the gantry and Z axis are leveraging forces on the uprights. It is structurally sound in the long axis, but needs help in the short axis to achieve a little more rigidity. Those too will add weight, so now you have another reason to lighten that back plate.

[wizard]

Or go with boxed sections. The focus on the gantry is important as that is often the weak link in a design. In the same regards I'm not seeing where stiffness in the gantry beam itself is coming from. Maybe I'm missing something but it looks like a fabricated C style beam with not much to the two flanges. Your stand is only 24" plus what ever the saddle is so it might be stiff enough but I have reservations and can't really tell what is there from the renderings.

A fabricated frame has one issue in general that can be a problem, that is distortion that welding causes. I'm assuming you are welding the frame up here. In any event expect to have to do some post welding processing to flatten or align parts. These guys: TabandSlot.com | Laser cutting tab and slot for welding have a short design video that explains the theory on how they design their products to reduce weld distortion by the people that ultimately assemble them. They actually have a few tutorials. If you can work these concepts into your design and materials it might help a bit with distortion control, minimizing after welding processing.

Maybe you have already done that, I just have this fear that the side plates might end up to wavy to mount linear rails on. That can be resolved in a number of ways if the surface is too wavy but something to be aware of.

Something like this in red?

[wizard]

Quick look at the brain I just soldered pins on. I have Mach3, but wanted to try using the Pi, so I'll be using bCNC. The Protoneer RPI-CNC hat will feed GND/ENABLE/PULSE/DIRECTION signals to the KL5056 drivers. It will also receive inputs from limit switches and a few of the control panel buttons (E stop, CYCLE START, FEED HOLD, ABORT) as well as control the VFD. Neat little package.
Attachment 373806Attachment 373808

Nice!

Just about every day I hear about a new CNC solution. bCNC is in fact a new one to me, hopefully you can report on how well it works in a few months. By the way the Beagleboard guys have just announced a new pocket BeagleBoard. It is shocking how much compute one can stuff into a tiny little PC board these days.

[cmadki4]

Here's a look at the gantry construction. In blue the 1/4" thick on top and bottom and 3/8" in the middle. Beam length is 811.35mm.
Attachment 373816Attachment 373818

@wizard: Nice find. I kind of implemented some of the Tabandslot.com ideas, but could clearly do with some refinement. Watched a few of their videos. Their design/assembly methodology definitely inspires some confidence I can keep these pieces relatively flat/straight.

And back to the gantry supports...?
Attachment 373822

I keep fidgeting with a 3/8" thick scrap. I just can't imagine that much deflection while cutting wood or aluminum (6061 and 5053). But I'm more than happy to add rigidity if it doesn't make any major compromises elsewhere.

The gantry supports don't necessarily need to be offset. They could be more rectangular. Then I could box them in.

This isn't a very big machine:

[cmadki4]

@LeeWay: I reread your previous comment. What about something more along the lines of a gusset on the inside faces of the gantry uprights?
Attachment 373826

[mactec54]

That is the right idea, but insufficient. Most use gussets of some shape and at least a pair of them on each side.
Steel plate gives or has movement and the gantry and Z axis are leveraging forces on the uprights. It is structurally sound in the long axis, but needs help in the short axis to achieve a little more rigidity. Those too will add weight, so now you have another reason to lighten that back plate.

My post # 8 already said the same as what you are trying to say, when a piece of bar is add in this way it is called a Rib, it would help if you understood the correct terminology, of fabrication, here is a little information on Ribs

Adding ribs to your design will increase the ‘bending stiffness’ of a plate. The ‘bending stiffness’ refers to the applied bending movement of a plate. The ribs allow the part a degree of flexibility that effectively increases its strength, as movement is absorbed rather than inflicting shock.

It is common practice to set a rib’s thickness at 60 percent the thickness of the wall it is supporting. So if a wall is 5mm thick, the rib supporting it will be roughly 3mm.

So in your case if your side plates are 3/8" then the Ribs would be 1/4"

What's the difference between a rib and a gusset?

This was a question in an engineering exam, ( Snip ) his answer was "gusset" the correct answer is either a "rib" or a "web". Does anyone know why this is a "rib" and not a "gusset" he asked, answer Because it does not represent a Gusset, it looks like a Gusset as most know it , but the correct answer was a Rib or a Web

[NIC 77]

I've got the itch for being able to cut wood and aluminum again.

Is this a design log or a build log? Please tell me it's a design log! I want you to succeed.

If you have the ability to get parts laser cut you can do so much better. I think where you got messed up is in thinking that by making it as light as possible you will get better performance. This isn't true. Your machine won't be able to handle better accelerations or faster cuts because it isn't rigid enough. More weight would give you better performance because you would have the stiffness to handle better acceleration and deeper cuts, allowing you to take advantage of all of the motor torque available. Allowing you to minimize bit deflection, reduce chatter, and attain decent material removal rates. You are not making an aircraft wing, or a helium filled airship!

Forget about cutting aluminum even reasonably well with this.

Your gantry is basically an I beam. I beams are bad in torsion. Not only that but yours is made from aluminum, which is not nearly as stiff as steel, and you have lightening holes cut out in it! Who cares about 5 lbs? You're not thinking about it right. You should add an extra 100 or 200 lbs to what you have.

This project will fail hard as is.

To get good results you'll need a different design strategy. A laser cut aluminum gantry with internal ribs, front and back plates, and filled with epoxy granite, that is what you want. Don't be discouraged by my comments. Use them as a springboard to come back with something good, that will wow everyone with performance as much as your good cad work and laser cutting ability already has it appears, in this thread.

Look at this thread:

http://www.cnczone.com/forums/uncate...aluminium.html

He has a gantry made from aluminum plate and filled with epoxy granite, and it works awesome. Best to find something that works and use that concept. Have you seen any designs like yours that work really well? It's a rhetorical question, no need to answer.

Toss this design out and start again from scratch. Use designs with proven performance to base your concept on. You will be glad you did. Nice place to start, nice work so far in coming up with a concept, so please don't be offended, or give up, but like I said, I want you to succeed.

[mactec54]

cmadki4

Now just bend the ribs down and make longer where needed, to run down through the center of the Linear Bearing mounts and you will have a good rigid side plates

[NIC 77]

Stick with what you have, 5 lbs. is a lot when you start moving it around, the design is more than what is needed to do the job, and much better than what most use, ( Aluminum Profile ) the webs are large enough, for support, and the top and bottom plates add the strength that is needed, I have made the same design in aluminum and it works very well

That's probably the worst advice I've seen on this forum. 5lbs is nothing. Aluminum Profile (I assume you mean T slot) is a better choice Than a 3/8" aluminum I beam with lightening holes.

Really, you've made this design and it cuts aluminum well? What a load of hogwash!