Design Study: Rail Inline With Ball Screw

[Bravin Neff]

Honest answer, if 9mm is good in that design you can use less, triangulate and make it lighter whilst still being stronger and more rigid.
Or triangulate and make it far stronger and more rigid for little added weight.

I might not be understanding you correctly, but I think you have in mind either gussets between the horizontal and vertical plates, or what I have been imagining, a small vertical detail that connects the horizontal plate with the Gantry, which would fully box in the ballscrew mount and ballscrew. I have been leaning away from that, because the design as is leaves that surface (the horizontal one that the ballscrew mount attaches to) fully exposed on a vertical machining center, making it easy to backside machine. In other words, I am trying to balance the design with ease of Machining these details.

If you can imagine this entire Gantry as one welded fabbed unit, then putting it on a vertical Machining Center, upside down, makes the rail Carriage surfaces easy to machine flat, and then you can back side machine the opposite side, making the ball screw Mount Services flat as well.

[Robin Hewitt]

Is this your first design? What is it supposed to cut? How big is it?

[magicniner]

Basic Engineering Design revolves around Triangulation, without which you will never have optimum rigidity for the weight of your design.

[Bravin Neff]

Is this your first design? What is it supposed to cut? How big is it?

Yes. Wood and aluminum. Roughly 950 x 630 x 135.

[Robin Hewitt]

If you are trying to put the screw close over the linear blocks, why don't you sink the nut in to the plate and drop the screw down? How are you going to keep the cuttings out of the screws? You have made it from heavy section to cut aluminium but that will not help you get the speed to cut wood efficiently. If your gantry weighs 50 lbs then you need 50 lbf plus the cutting force to accelerate it at one G.

OTOH I think you need to build something wrong and then fix it with the Mk2. Understanding CNC design does rather come with experience. Getting it wrong is not actually a bad idea.

[Bravin Neff]

If you are trying to put the screw close over the linear blocks, why don't you sink the nut in to the plate and drop the screw down?

I don't follow you here. Do you mean machine a pocket into the plate in order to get the ball nut lower (i.e, closer to the linear carriages)?

How are you going to keep the cuttings out of the screws?

Guards, covers, bellows. They aren't shown.

You have made it from heavy section to cut aluminium but that will not help you get the speed to cut wood efficiently. If your gantry weighs 50 lbs then you need 50 lbf plus the cutting force to accelerate it at one G.

I don't follow you here. You seem to say it won't be fast enough by saying it won't accelerate enough - is that what you mean? Regardless, I assume there is always a servo strong enough. Some of the machines I sell (for my day job) weigh over 30,000 kg and can accel at 1G. Of course those motors are huge... I'm personally looking at the Clearpath servos. Nema 23 and 34 frame sizes.

OTOH I think you need to build something wrong and then fix it with the Mk2. Understanding CNC design does rather come with experience. Getting it wrong is not actually a bad idea.

I think that's sage advice. What part are you seeing that is wrong?

[cropwell]

Why not put the screw inside the frame

[Neale]

I would consider moving the ballscrews just outboard of the ends of the gantry. This allows a more rigid fixing of gantry to carriage feet. If the ballscrew mounts are on outrigger brackets welded to the outside of the frame tubes, mounting pads for the ballscrew bearings could be machined at the same setting as the frame tubes. There is no benefit in having the ballscrews over the slides as a gantry driven at both ends is balanced anyway. It's not like the gantry ballscrew which is more difficult to fit between the linear rails.

The idea of being able to machine the top of the long rails is so far from reality for most of us that I doubt if many have designed for that! I'm in the "if I can't machine it, build in adjustment" camp...

[Bravin Neff]

I would consider moving the ballscrews just outboard of the ends of the gantry. This allows a more rigid fixing of gantry to carriage feet. If the ballscrew mounts are on outrigger brackets welded to the outside of the frame tubes, mounting pads for the ballscrew bearings could be machined at the same setting as the frame tubes.

Great point and idea. I'll try to mock something up. I am definitely motivated by making the machining of the machine as straightforward as possible and only requiring a vertical mill. When I see machines, similar to what I just mocked up except the ballscrews are outside of the frame, mounted to the outside frame vertical wall, this is either (1) asking for a large boring mill or horizontal machining center, with (2) multiple setups, just to machine the ballscrew mounting pads. That starts to add up fast. Or the pads are painstakingly fitted by hand with shims and numerous iterations of racking the gantry back and forth to get alignment, and then I question how flat and parallel the mounting surfaces can be.

The idea of being able to machine the top of the long rails is so far from reality for most of us that I doubt if many have designed for that! I'm in the "if I can't machine it, build in adjustment" camp...

I am surprised to read this. Granted, my machine is a little on the small side compared to many I see here, but it will fit on an 80" x 40" machine. The going shop rate is $60/hour near me, so I am counting on all the machining work coming in just a few hundred $. Given how much money people invest into designing and building their own CNC routers, I would thinking have the rail surfaces and mounting pads machined would be easy to justify.

[routercnc]

My current machine uses something similar. I was trying to get the gantry sides as stiff as possible, and make a good connection to the rails. Fitting the ballscrew in there as well was just convenient.