[tcom-frazzled]

Folks,

I have a large CNC router that has a 1/2 thick steel plate table. It was originally used in a limited capacity, with fixed size material carrier pallets that would sit in machined registration flats on the table. Location registration was achieved using retractable pneumatic registration pins and the pallet was then pneumatically clamped in place.

The original application for this machine is no longer needed, so the machine is being repurposed for general use. As a part of this modification, I would like to finish the entire work envelope of the table surface so that it can be drilled and tapped in a manner similar to a large tooling plate, with a dead flat surface. The surface is currently just the factory rolled finish on the plate. Finishing the table in this manner will give me the option to use clamping and fixturing solutions that currently aren't available.

That said, I was considering milling the surface and then using a cup wheel in the existing spindle to finish grind the surface flat using the machine itself. I realize it'll be a painfully slow process, as I'll only be able to take .0001 at a time. I'm assuming the process will likely take several days to complete, but since the machine isn't doing anything else right now, I'm not too worried about that.

Before I go off and try this, I wanted to sanity check the idea with folks here to see if this is the best way to go about it.

[txcncman]

Only curious as to why you need the precision of a ground surface. I would most likely fly cut or face mill it and be done with it.

[tcom-frazzled]

one of the fixture types I plan to use is a double-sided vacuum fixture that would need to make a seal against the table surface. I agree with fly cutting it to start and if the finish is adequate for that seal, I can live with that unless something comes along that requires a higher tolerance finish.

My primary concern with grinding is that a cup wheel would wear enough during the operation that the last pass would be out of plane from the first pass. I'm surfacing an area approximately 40" x 90".

[tcom-frazzled]

...and I tend to be a bit of a perfectionist on some things like this.

[Geof]

I wouldn't touch it!!

If, as you say, it is simple plate it has all manner of internal sresses from the original rolling process that made it. As soon as you skim a bit off the surface you start to relieve these stresses and the plate is likely to warp in all manner of unpredictable ways.

In your place I would get a piece of 3/4" aluminum tooling plate (MIC6). Tap the existing plate around the perimeter spaced about every 10inches and tap a 10" grid pattern; 5/16" NC would probably be a good enough size. Bolt the MIC6 down tightly with bolts recessed 1/8" below the surface and then check the surface for flatness.

No matter what you do it is highly unlikely you will ever get a surface flatter than +/-0.0005" per foot and very likely your requirements don't even need it this good. If it is not good enough for your requirements take a very light facing cut over the whole surface with a fly cutter.

[tcom-frazzled]

Perhaps worth mentioning that it is welded to a base structure consisting of 5/16 wall C and I channel framework (4x8 typical dimensions). The largest span between support structures is approximately 4ft x 3 ft.

I'll take a closer look at the plate to see if there is any indication it is cast rather than rolled.

Not sure if that changes the internal stress argument or not, but thought it worth mentioning.

[Geof]

If the top was cast certainly it would not have the same internal stresses but if structural members are welded to it you can be confident it is not cast. Also if structural members are welded to it you can be confident that there are additional stresses present from the welding. And if it is only 1/2" thick and in places it spans 4ft by 3ft it will not be rigid enough just to use it by itself as a mounting surface. It would tend to vibrate like a big drum.

[tcom-frazzled]

Keep in mind that it is a cnc router, not a machining center. The primary use is cutting wood and plastics with occasional light duty work in aluminum plate.

[Geof]

Point taken but I still think half inch is thin. You can build up a surprising amount of deflection if you hit a resonant vibration even with a light cut.

Another thing about trying to face it that I did not mention is that rolled plate tends to have a tough oxide skin. You need to be able to take a heavy enough cut to get under this otherwide your facing tool will just rub itself to death; even if it is carbide.

An alternate possibility to facing or topping it as I suggested could be to fabricate a rim around it and pour on a layer of self levelling epoxy. The sticking point to this idea (excuse the pun) is that it could be non-sticking. You might have to sand the surface to get a good key for the epoxy, or mildly etch it which introduces the hazard of acid fumes floating around.

[tcom-frazzled]

Excellent points, and certainly ones that warrant additional consideration before I do anything that could end badly.

That said, I think I'm going to gather some additional info. I'm going to start by mapping the surface to see how much "correction" would be necessary. I can throw a touch probe in the machine and write a macro to sample the existing surface. I'll dump the data into a spreadsheet and crunch the numbers to see what I'm working with. I guess the approach I take from there is going to be highly dependent upon the deviation between the minimum and maximum table surface heights.

May be a hairbrain thought, but I could make and install threaded inserts extending above the table to allow me to pour a 1 inch epoxy/granite layer that could include cast-in vacuum table grid that can be used with gaskets (I have a 10HP Becker vacuum pump waiting to be added to this project). I would still be left having to grind the final surface, but it would add some additional clamping options, help dampen vibration, and cost a hell of a lot less than a 4ft x 9 ft piece of 3/4 MIC6 cast tooling plate. Not without it's own challenges, but I can create a small form and test the idea pretty easily.

[Geof]

Not really hairbrain. Ambitious, yes, but if you made it work it would be a very nice table. And I don't think you would need to do any grinding or surfacing because you could easily do a topcoat of self levelling epoxy, it would certainly stick to the epoxy granite substrate.

[Analias]

@tcom-frazzled any updates on your progress? I'm interested in what you decided or any new information.

[tcom-frazzled]

I decided to clean up the existing surface and then install a modular E/G top made up of plates that will be bolted to the table. I'm still kicking around a few designs that would provide a combination surface that would provide a combination of t-slot, vacuum, and tapped tooling plate style options.

As luck would have it, a production job came in that now has the machine tied up with a jig, so I'm not sure when I'll have the downtime to get more done on this.