Recommended Method to hold Aluminum Extrusion in Gantry Style Mill

[Steven Smith]

The ShopBot thread alluded to this, and I've heard a few folks talk about a "rail mounting" system for their gantry mill.

The specific question is, does anyone have a good picture or description of the system they use to hold "rails" which I'll define as square/rectangular aluminum tubing of lengths between say 2 to 6 feet, in a manner to allow sufficient rigidity for FRC use. If so, would you be wiling to post pictures?

My Experience

We have a 8'x4' CNC Router Parts Pro gantry mill. For the most part, we have figured out how to keep pretty darn accurate (~0.002-0.003" over ~12") parts on it. Any vibration of the parts though will generally lead to rapid heat generation and failure. Our bed is dual MDF sacrificial boards, and for the most part we just use hold-down screws to keep sheet metal/plastic held down well, and resurface a couple times a year.

The first year we had it, I tried milling pockets into MDF and using finger boards to hold down the aluminum. Many things have changed since then (VFD/spindle/Compressed air/coolant), but that didn't work so well. It also wasn't very accurate (or I was a poor machinist!) for being able to flip a part, locate it again, and cut another hole, such that the two holes were perfectly aligned. The last two years, we've just done the operations on a manual mill with "ugly" lightening/access holes. Quality was pretty good, but it ended up taking quite a bit of time, as we don't run manual machines every day.

Current Idea for Feedback

I'm playing with the idea of using a ~12" steel C channel, something on the order of 15-20# per foot, and mounting some 4" HF drillpress vises every 1' or so along the piece. I could then drill/tap the steel and align all the fixed jaws to each other.

Potential advantages I see... the C gives the thing a little more rigidity, allows any coolant runoff to be contained a little better, and the whole thing should be pretty hefty (~120-150#), so just a few screws should keep it from walking around. Also, the total cost would be under a few hundred dollars.

Disadvantages/Concerns include... I'm not sure how much movement i have in the Y axis as I move along the X (need to check), due to twist in gantry rails, etc, and what that means in terms of viability. I know the cheaper vises will generate lift on that edge of the part (need to quantify). Also need to verify how much room I have in the Z axis in the shop today, and what that means in terms of tooling I can use. Also, not sure how much twist/warp to expect on structural steel. It could be no where near the accuracy desired. I can potentially shim all the vises to correct this, but... might end up chasing my tail a bit.

Any rate, any thoughts anyone has on the subject, or preferably pictures of a setup you have that you like would be appreciated.

[mman1506]

I've been thinking of using Taig milling vices to hold box tube on my router. They are better made and lower profile than cheapo drill press vices.

[JesseK]

Attached is what we are experimenting with this offseason. It was precision-made using a manual mill. It can hold extrusions 10" - 42", and can rotate 360 degrees freely. We have worked out most of the chatter in it, but there is still a little bit. I can get more info on it Thursday, if you have questions on how it was made.

[asid61]

Originally Posted by JesseK

Attached is what we are experimenting with this offseason. It was precision-made using a manual mill. It can hold extrusions 10" - 42", and can rotate 360 degrees freely. We have worked out most of the chatter in it, but there is still a little bit. I can get more info on it Thursday, if you have questions on how it was made.

A very interesting device you have there! How is the tube clamped to the endcaps? Can it do sizes of tubing other than 2x1x0.1"?

[JesseK]

The endcaps are press-fit for normal 1/8" wall, or snug-slip-fit for VEXPro extrusions. Turns out that normal 1/8" wall extrusion just needs some filing (for a taper) to get it started on the endcaps, then a rubber mallet does the rest. The shoulder of the endcap that goes into the extrusion is somewhere between 1/8" and 1/4" - I don't remember.

We could make additional endcaps for other-sized tubes if we CNC them often (e.g. 2x2 for telescoping what-nots). For now though, we'll just clamp the tubes down to the table like we did last year.

The tricky part about the construction of the endcaps is that both sides must be perfectly symmetric about the hole they rotate. Otherwise the two sides of the rail will not precisely align if doing mirrored CNC cuts after the 180-degree rotation (which is the point of this particular design). An easy way to check is to see if, after a 180-degree rotation on the jig, the spindle has to be re-zero'ed.