Power Route - anti-racking system upgrade build

[petek]

Before i found some people were replacing their anti-racking cable system with a dual y ball screw system, i started making this heavy duty antiracking system. Its designed to reduce out of squareness deflection at 5 newtons (~5kg) by a factor of 6. So the 0.003”-ish should become 0.0005”-ish, in theory anyway, to be measured at the extreme RH or LH of the deck, taken in the y-axis.

Anywho, heres a pic of it mostly assembled.

i also upgraded the beam there too (its the bottom y - link (between gantry plates under the deck). That should reduce deflection under the same load from 0.001” also in y, to somewhere around 0.0002”, in theory anyway.  All to be measured later. 

[oldtech]

Interesting. I have been thinking about dual motors for the y axis and other upgrades, but the more I looked into it the more complicated it became. I concluded that approach is a hard way to build a new machine.

So now I'm keeping the PR, doing only maintenance, and have started designing and building my own machine that will be more rigid with servos and linear rails.

[petek]

I recognize the limitations of a gantry design vs traditional CNC Mills. Discussions on practical machinist forums have discussed this at length. Surface finish will likely always be weaker with gantry types supposedly. I think this is vibration and vibration absorption issue related.. and benefits from heavy masses of cast-iron.. but that’s a trade off for a much larger bed footprint (gantry), and at a fraction of the cost to traditional professional mills. For me that means accepting that, for 3D convex and concave surfaces that are aesthetically important, I will need to sand them. This unacceptable for a factory, but fine for my niche-product aspirations, esp for now. I’m happy to work off this power route platform for my part size requirements, production volume (tiny), and investment cost standpoints. If my products don’t sell, making/buying a better CNC won’t help.

It would be great to see what you build, if you don’t mind - please post! Would u just get an old knee mill and CNC-ify-it?

[oldtech]

I'm building a gantry system. Based on my understanding of rigidity - bending resistance and vibration absorption - I started looking for materials. Metal is strong, but poor at vibration, unless you add a lot of mass - that's why the old mills and lathes were made out of massive chunks of cast iron. And in general materials that absorb vibrations have less bending resistance so there is not one ideal material that will satisfy rigidity. A few people are building composite machines using epoxy granite over a metal frame and I thought about approach, but it's expensive and epoxy is both difficult to work with and expensive.

So I looked at other materials that absorbent vibration, relatively cheap, and easier to work with and decided on Versa-lam engineered wood beams. Wood is better at absorbing vibrations than metals and while not as strong as metals on a per unit basis it can be just as strong with careful sizing and engineering.

If my calculations and my memory are correct two 60 x 60 aluminum excursions in a gantry spanning 32 inch with a 1Kn load (~225 lbs) has a deflection on the order of 0.2 mm. While two 3.5 x 7 inch engineered beams, that I am using, has a defection on the order of 0.002 mm (0.00008 inches) for 1Kn.

My first step is to assemble the frame including the gantry using the 25mm linear rails. The gantry will consist of two solid aluminum bars ~1-2 inch by 5 inches for vertical support with two 3.5 by 7 inch beams. I'm also planning on an aluminum base plate mounted on top of the beams for the linear rail. I'll finish by putting an epoxy coat on the beams.

I still have to determine how to fasten everything together. Obvious choices include steel cross dowels or inserts glued in with epoxy - that will take some experimentation. And then make decisions about the rest of the machine.

BTW: the cost for a 8 foot 3.5 x 7 inch Versa-lam beam is ~$85 at Home Depot, while a series 30 60 x 60 aluminum excursion (used in the PR) is about the same (80/20.net lists it for $1.0 per inch). Yet, you could safely support the weight of a car on the Versa-lam beam (rated at 14,000 lbs).

[petek]

Would you need to address moisture effects on the wood?

[oldtech]

Yes. According to Boise Cascade Versa-lam is dimensionally stable*. But, I'm not counting on that alone because I am using a balanced design. In other words if the gantry should widen by 1 mm the beams supporting the linear rail on the y axis will also widen by 1 mm because they are attached to two other base beams. In addition, I am planning on coating the beams with epoxy to minimize dimensional change.

I do realize that this is an experimental design. That's one of the reasons that I just plan on building the gantry and y-axis as a first step and to test the dimensionally stable claim along with the homogeneity of 'engineered' Versa-Lam.

A fall back design would be to just attach the beams to a solid aluminum frame like they do for epoxy granite machines.

*So far I am impressed with the straightness of the Versa-lam top and bottom edges. I put a 6 foot straight edge on top and did not see any gaps. Of course, I still need to take more measurements. The sides, however, are rough and unfinished. They are not something to admire without finishing and are intended to be inside of structures. That's another reason for the epoxy.

[jms]

Just a heads up, check your local lumber yard instead of big box stores, My lumber yards can be 50% cheaper. You think big box is cheaper but they are not.