Good choices for torsion rods?

I’m playing around with some ideas for a larger-format CO2 laser, looking for about 1250mm by 700mm cutting area (in pass-through, I’d like to be able to put something 4’ wide in it).

To drive Y motion, I am playing with two ideas. The first thing I came up with was a C-shaped belt loop with the gantry attached to the sides of the C, and the motor at one corner. That would be about 5500mm of belt, at which point I start to wonder if it would stretch dynamically enough to affect quality. (Also, 5500mm of 9mm Gates belt would cost over $80, so not a cheap mistake if I ended up throwing it away.) I haven’t seen this belt layout in a large format machine; I think it was used a few times by people experimenting with reprap, but I don’t think it’s ended up in successful designs.

The other option is to run a torsion rod down the front or back with a loop of belt at either side and a motor driving the torsion rod. This is a relatively common setup in 3D printers (part of the ultimaker movement, also used for stacked movements in printers like the Ender 5) and I see it also in a bunch of lasers. Picking a random example from the OpenBuilds forum with a similar size, the Vulcan uses a torsion rod setup. However, it has a dual-shaft motor in the middle, whereas I have a specific single-shaft servo motor in mind.

Thoughts on what kind of torsion rod to get that’s about 1300mm long that will be both light and stiff? I could see “whipping” being a problem. I guess I could spread a few radial bearings in blocks at irregular intervals (no two on low-order nodes of the same harmonic series) but is there a state of the art that I should pay attention to? I’d like to avoid anything heavy to reduce ringing, if I take this route.

Beamlazer also uses a torsion rod design. It has a BOM where it just lists 2 x 1 meter 12mm rods, without a link to the source in the BOM, so I don’t know what kind of rod for sure.

The open source Lasersaur also uses a torsion rod redign:

I saw a reference to Lasersaur here recently, but took it less seriously when I saw the CW-3000 specified for chilling a 100W laser tube in the BOM.

Following the Misumi link, it looks like they are using 12mm induction-hardened bearing steel precision rod for Lasersaur.

I really have two concerns for solid rod: inertia and whipping. There’s an inverse relationship; to reduce inertia I want a thinner rod, and to reduce whipping I want a thicker rod. I just looked at carbon fiber wrapped rod and it’s available relatively inexpensively. There, I could use a fairly wide section with 1mm wall, which ought to be lower inertia, and if I guess right on width would probably be stiff enough.

Maybe I’m over-thinking this. :slight_smile:

We operate a Lasersaur (with a CW-5200) in our FabLab for five years now and never had a problem with the 12mm rod. I can’t feel any inertia or whipping from that rod. The 12mm diameter is very small related to the pullys and weight of the gantry. The effect of long belts if for sure more noticeable.