I’ve started printing pieces for the MPCNC, and have my conduit on hand. I was going to start cutting the conduit today, but while planning out my cuts, I noticed that I can actually build a slightly larger machine than I was originally planning. My original plan was a 28" x 38" cutting area (the foamboard I want to cut is about 2’ by 3’) with a 4" z direction. However, while planning my cuts on the conduit, I actually have enough to do 38" x 38" x 6".
Quick and dirty analysis shows that I’m approximately halving the accuracy I’ll see by doing this, but with the 20lb load I was assuming that still only gives a deflection of about 3/32". The z-axis I’m not too worried about, I was going to keep the material I’m cutting as high in the machine as possible anyway.
Am I missing anything major with this line of thinking, or does it sound like I’m good to go?
Just looking at the difference in calculated deflection between the two lengths I looked at. There will be other sources of error too, but just from the rails I went from around 3/64" of deflection to 3/32" of deflection. As a side note, I didn’t bother figuring in any mid span supports for now. I just wanted a quick feel of just how much worse the deflection could be, so I’m fudging to the extreme worst end.
The 20lbs is a number I just pulled out of thin air because I knew it was stupidly high. Its about 5 or 6 times higher than what I expect to see. Spindle will be a pound and a half at most, 3D printed parts won’t be much. Actual deflection in the rails themselves will probably be smaller than what I could reasonably measure.
The way I’m looking at it is that if this is the result I’m getting at absurd loads, I’m not seeing much reason to not go with the larger size.
Stupid me cut the first piece an inch short anyway because apparently I have no idea how to read a tape measure. So 37" x 37" it is I guess. I’m not going to try and keep a 38" long and a 37" long axis straight.
I think it should be quite easy to add diagonal braces to the corner conduits to significantly increase the rigidity if you wanted a bit more height. Here’s one example: https://www.thingiverse.com/thing:1730371
Sort of. I would think those braces would solve part of the problem but not all. The Z axis needs to go back down every inch your legs go up. Then if you are using a spindle/cutter you will have a LOT of torque challenging the “Plumb” or vertical alignment of your Z axis. So if you use long legs and these braces you would need to raise your work piece and spoil board in order to reduce the torque forces applied to the tip of your bit.
+1 to what Aaryn said. The deflection of the legs is much smaller than the deflection of the Z axis. Therefore increasing the stiffness of the legs won’t help much.
There are cases where additional Z legs but not z carriage length is required, such as thicker materials or a vacuum box, but generally you are correct, keeping Z height to a minimum is a good idea.
In the construction industry they use something called CFT, or concrete filled steel tubes. This makes steel tubing much stiffer without the cost of the alternative which is thicker walled pipe. One downside is that it is very messy to fill a 1 inch tube with concrete — don’t ask It also adds to the moving mass, so you might need to upsize the X and Y steppers to get the same acceleration out of the machine.
Another alternative would be to use stainless steel pipe, which is much harder than the steel used for conduit, but very expensive.
The modulus of elasticity for concrete is actually much, much lower than steel, so for a given deflection concrete filling would barely add any force. Or simply put it wouldn’t add stiffness. Intuitively concrete seems stiff because we usually encounter it in big blocks but young’s modulus is only about 17 GPa compared to steel at about 200 GPa.
For thin wall tubing maybe it helps prevent crushing, or perhaps buckling under axial compression, but for our loads it doesnt help.
It also adds to the moving mass, so you might need to upsize the X and Y steppers to get the same acceleration out of the machine.