The basic mechanics are pretty much wrapped up and the machine functions pretty well.
I've got roughly a 40x20x3 cutting capability. The 20 inches being a limit due to short linear rails. Replacing those rails will bring it closer to 40x32x3.
I'll be adding cross bracing as soon as my shipment of 8020 shows up. I figured it was time to focus on getting the machine as close to level as possible. Also, I'm in the midst of moving the control board underneath the machine and tidying up the wiring.
As shown in the pic below I chose to not cut the extruded whenever possible so that I can ultimately use the 4' pieces on another machine... someday .....
This pic sort of shows what i was doing in order to gauge the relationship between the rail and the top surface of the piece of 8020 that it's fastened to, which is where I'll be mounting the surface for cutting.
Prior to placing that piece of aluminum plate on top of the rail, I was sweeping the gantry by hand from left to right with the gauge making contact with the center of the 8020. As expected the results were less than spectacular. From 0 to about 44" the max deviation was about 0.08", sloping up from left to right. The error seemed linear as I swept from the left to right, which seems to indicate that I just need to drop the right side of the rail. I'm not sure what kind of numbers i should shoot for ... 0.01", 0.005"? I'm sure there are some variations on the 8020 surface itself... I know I can do a lot better than 0.08" though! The other side produced similar results. Unfortunately, I loosened the left side socket screw holding the rail against the 8020 and jacked it up all the way and locked it in place, and then let the right side drop by loosening and letting gravity do its thing, but I'm still stuck at 0.08. I may need figure out another way to mount the rails or perhaps widen the width of the channel a bit so I have more adjustment available.
So here is the dilemma. I'm not sure if i should bother with this exercise and get it as close to dead on as possible, when I am going to be fastening another surface to the top of the 8020. Whether that surface is going to be a piece of 3/4" mdf or a 1/4" or 1/2" piece of aluminum plate is yet to be determined. I'm not sure what the tolerances are for mdf over a 4x8' sheet, or on the thickness of aluminum plate from my local supplier. So I could get this thing down into the high thousandths and then blow it all to heck with a warped surface piece.
Options:
1) get it as close to dead on now before tossing the surface plate on
2) get it close to dead on and the tweak further when surface plate is anchored on...
3) just mill the surface flat after you toss the plate on, dummy!
A close up of the Gantry uprights, so far they seem to be pretty rigid, ill add in an angled cross brace and do some redesign if needed. The rails are more than up to the job.
I added some leveling feet earlier tonight, I figured the feet might provide a wee bit of dampening. I guess time will tell. I never realized my floor was so far off!
Just a quick UPDATE: After an hour or so I was able to get very decent results on one side. I can sweep several passes up and down the 4' section of extrude and it's now +/- 0.002". I don't think I'll waste anymore time on that side.
I basically had to remove the rail and bolt it in using every other hole (instead of 3 socket head total at the front/mid/rear) This rail is around 1x1", yet it still was deflecting under the weight of the gantry. I was surprised.