Need Help!- Steel Rod

[Drools]

In a different thread I was informed that 5/8 steel rod was not strong enough for a 3' x axis or a 2' Y axis. Today I upgraded my order to 1" steel rod for both axis will 1" steel rod be of for this?

[ahren]

Drools,

It depends on your application. What sorts of cutting loads will you have, and what sorts of deflection and error can you tolerate? Basically, the deflection you will get is a function of the load put on the rod, and the rod geometry. Check out some of the deflection calculators on engineer's edge to make these calculations.

You're on the right track by increasing the diameter, especially since stiffness is a function of diameter cubed. So your 1" rod is 4 times stiffer than the 5/8".

Personally, I'm not a huge fan of unsupported rails, which is why I live open pillow blocks with supported rails, or skate bearing systems like the one I sell. But for a short span, with a big enough rod, you can get acceptable results.

Ahren
www.cncrouterparts.com

[Drools]

Thanks for the reply Ahren, I was starting to think I'm asking my questions in the wrong forum. I want to do PCB milling and some wood. I was hoping my X axis would be 3 feet and the Y 2 feet. I really hope the short distance of the Z axis will ok with the 5/8 rod.

[BobF]

Drools,
Look at some of the Joes2006 builds. You will see how he used double pipes on the long axis supported by a torsion box to a increase the stiffness.

[ahren]

For PCB routing, your cutting forces will be low, so you can probably get away with the rod. The deflection you get is a function of the rod diameter cubed, but it's also a function (in the other direction) of the rail length cubed, so if your z axis is short, you'll be in good shape with the 5/8. If you're set on using unsupported rods, I think this is about the best you can do without getting ridiculous. What are you planning on running on the rods? Round linear bearings, or something else?

Ahren
www.cncrouterparts.com

[Drools]

I was planning on using aluminum angle and skate bearings.

[Drools]

I have noticed the use of what is called I think "gas line pipe" (hollow pipe) this is not the stuff I purchased. I went with 1" solid steel rods on both the X and Y axis. Will the solid steel rods really flex quite a bit under load?

[ahren]

Your rods are slightly stiffer than the pipe, but not much. The material towards the farthest outside diameter provides the greatest resistance to bending loads. Use the calculators below to answer your questions. The first calculates the cross sectional moment of inertia of your tube (the default is conveniently 1"). The second allows you to plug in this moment of inertia, along with different loads and lengths, and see the deflection. For your setup, I would figure 50 lb loads at a minimum:

http://www.engineersedge.com/calcula...re_case_11.htm

http://www.engineersedge.com/beam_be...flection_2.htm

Using these values, and the default stiffness in the calculator (which happens to be correct for steel) for a 36" span, I get a deflection of 0.033", which is about 1/32". You might be able to live with it depending on what you want to cut, but I'd personally be kind of bummed about it, especially since this would probably cause an angle that would be multiplied over the length of your z axis. Additionally, if you wanted to make more aggressive cuts, you could get above 50 lbs of cutting load, and this deflection will get worse. This is why I prefer supported bearing rails.

Ahren
www.cncrouterparts.com

[Drools]

Thanks again Ahren, I will look through Joes section for some examples of supporting the 1" rod I have. I can't take it back as it is cut to lenght so I have to figure out a way to use what I have.

[BobF]

(0--0)
If you use 2 on each side, you build something like my crude ASCII drawing above. Imagine the drawing rotated 90 degrees. The 0s are rods. The -- represent MDF or wood standing on edge between the 2 rods (or pipes). the ( ) represent your skate bearing carriages, one on top one on the bottom with a means to adjust the clamping pressure.