Looking to build more stable table

[EricSchimel]

Hey all! Right now on my 4x8 PRS I've got a pretty standard table setup:

Bottom layer is plywood, middle layer is sealed MDF with plenum routed into it and the top layer is LDF (Trupan). All three layers are glued together.

This setup has worked for several years got all of the plywood I've been cutting. Recently I've picked up some regular production work that's all in plastic. One job in particular requires very accurate pocket depths.

What I've been finding is that the Trupan moves daily, I'm not all that surprised by this.

I'm exploring ideas for a new table. I've priced out aluminum extrusion which I'm sure would work, but it's a LOT of money.

I've also looked at making my bottom layer and my plenum out of Extira. From everything I've read it seems to be really stable. My thinking right now is that I do that, and instead of glueing down my LDF I leave it loose and only put it on when needed. (I could keep it from sliding with some pins)

For the rest of my plastic cutting I have jigs that I could place (and vacuum down to) the plenum directly.

So the question(s) are: Has anyone else fooled around with building a really stable table? What materials have you tried?

[coryatjohn]

The cheapest and most stable material would be plate steel. Steel is literally an order of magnitude cheaper than aluminum. It's also stronger, stiffer, and has a lower coefficient of expansion. The negative is that a 4x8 sheet of 1/4" plate would be so freakin heavy that you'd need a forklift to move it around.

[MaxFrenzy]

Hey Eric,

So, the plywood base and LDF are going to be more susceptible to changes in humidity over the sealed plenum, which you've noticed. However, a lot of the success for these accurate depths are going to be dependent on making sure your vacuum is holding your plastic securely/evenly throughout the area you want to pocket. With plastics not being porous, I've always had good success holding them down with the bleeder & blower style vac tables. My question is if you feel like you have good hold down, and you zero to the top of your material and run a pocket, what sort of deviation are you getting when you measure depth of one area to another? Have you tried testing your z-zero on several spots of the piece before cutting? One thing to consider is that the thickness of the material may not be consistent throughout. Spindle tram is another, but I'm assuming you've covered that. I've had production runs like this before where the depth was important. Ideally, you'd be able to skip re-zeroing in between cuts since the work piece thickness SHOULD be identical, but in many cases isn't. In one instance, I found that re-zeroing to the top in between pieces gave me better consistency. It's something you could try. Overall though, in the best situations, I could keep around 5 to 10 thou difference between two pockets that were a foot or two apart on certain pieces but other times, it could be 40 or 50.

[EricSchimel]

So I've been far down the road on this...

Firstly here's a shot of the jig:

34189

Basically it's a jig that's made out of HDPE. I've run gasket around where the part goes and I've got a high pressure pump that sucks the rough cut plastic down to the jig. That part works awesome.

To make the jig I had to perfectly flatten my table. I even ran a dial indicator all over the table to make sure it was as dead flat as I could get it.

Then I probed the sheet of HDPE and it wasn't totally flat, close, but not perfect. So then I shaved just a bit off the top of the HDPE. I then cut in my gasket channels and installed the pins.

I have a probing setup that I run to put the jig down on the table and locate it in XYZ (And rotation). I also dial indicate it to make sure it's flat.

This all works, however every day/half a day it all goes to hell and I need to resurface the table.

Essentially I have this whole jig sitting on top of what's effectively a sponge.

If I can put the jig directly on the plenum I think I'll be much better off. the thing I want to be sure of is whatever I made the base layer/plenum out of is as stable as possible.

Hopefully that puts a little more clarity on it...

[coryatjohn]

I did a job that required cutting thousands of plastic parts (acrylic) to very close tolerances. One thing I noticed is that the wear on the end mill was the biggest contributor of error in my machine. Something to consider.

[EricSchimel]

Tool length is measured at the start of each job, sometimes twice daily so that's not an issue here. Side wear of tools has already been accounted for and solved on this job.

The issue is the the table moves throughout the day, sometimes it's higher in on spot, lower in the other, etc.

A few hours later you can probe it again and there's a chance things will be different.

I'm really looking to stabilize the table so I'm not constantly chasing down Z tolerance issues.

[dlcw]

Eric, another thing to think about is setting Z-zero to your table bed instead of the material. This way, if/when, the material varies in thickness, it won't be an issue as all depths are cut based on the table, not the material. This is how eCabinets works. This way you take out the variable of material thickness. It seems to work well with mortise and tenon joints for my cabinets. Just my opinion....

[EricSchimel]

Yep, I already do that. I have an inductive zero plate bolted to the metal frame of the ShopBot. I also have a Z prox switch at the top. All zero measures are done off of that plate and referenced to the Z prox switch at the top for maximum accuracy. The cap on the Z prox switch is made out of hard plastic so that regardless of the geometry of the bit I get an accurate reading.

While nothing is perfect, the big fluctuating variable is the table.

[waynelocke]

I am getting ready to use 3/4" phenolic for my plenum. You might look into that. I think that it is pretty stable.

[coryatjohn]

Question: Is your vacuum pump under your table? If so, that could explain the changes you're seeing. Vacuum pumps throw off a lot of heat and if that heat is hitting the underside of your table, especially on one side or end, it's going to twist the entire machine.

[EricSchimel]

Heat isn't an issue. I already checked with a thermal camera.

Wayne have you priced out phenolic? I've been seeing prices that get close to actually making buying aluminum extrusion seem with it.

[coryatjohn]

Have you checked 80/20's eBay store? They offer odd sizes for about half price. The odd sizes can be over 8'. I've bought maybe a dozen of them over the years at that length. They're probably remnants from custom cuts.

[EricSchimel]

Unfortunately they closed that store earlier this year. Even at a discount you still need to get the hardware which roughly doubles the price. You're in the 2-3k range for that, plus I'd need to do some fancy cutting to get my vacuum plumbing in there.

I do love the idea of a metal table, but I can't help but thinking that the benefits don't justify the cost.. which why I'm exploring extira...

[bobmoore]

I use MDO as a base which is much more stable than any other plywood that I have worked with. I put corian on top of that with the vacuum grid (4 zones) carved into it. The spoilboard tops off the system but it is doweled not glued. It stays quite stable for weeks at a time but if I am carving tight tolerances I dress the table fresh before carving. One thing to watch when dressing the spoilboard is the amount of vacuum you pull on wide open spoilboard as apposed to carving a sheet that has all of or most of the spoilboard covered. I cover as much spoilboard as practical as I dress it to keep the vac as high as I can.

[Chuck Keysor]

Eric, please don't laugh, but I'd like to suggest you investigate using a type of concrete called "GFRC".

I have long been interested in Glass Fiber Reinforced Concrete, "GFRC", (but failed to do anything with it :(). It is used a lot in architecture to make exterior cladding for buildings, concrete counter tops, and various decorative items. It can be cast thin, and even will bend some. Plus you can embed hardware into it for securing your plenum layer. (They embed hardware when casting architectural panels, to facilitate attaching these to the building.) I have seen videos where they make panels that are just 3/4 of an inch thick. And some people mix in vermiculite to make the panels even lighter.

There are tons of YouTube videos on the material. Here is a typical video, this one by Smooth-on: https://www.youtube.com/watch?v=wA3GygYmz0g.

In addition to being stable, I believe that relative to your other options, that this is low cost. Of course there is the learning curve too...... You could probably find a concrete counter top fabricator in your area and see what he'd charge to make what you want. It should be right up their ally, as this would be just a big blank slab, requiring no sink or faucet holes. (Don't let them talk you into mixing in seashells or LEDs into your slab ;)!)

Thanks for your interesting posts, Chuck

[evan]

In 2005 /2006 I put in a 1.25" thick Paperstone bed. (I was cutting a lot of countertops at the time) If you are going to cut a vacuum grid into it (probably any phenolic top) you should keep in mind that it'll probably "warp" a bit because of all the stress relieved in cutting into the top over such a large surface (mine is 60" x 120 "). I had to flip it cut some grooves in the back to get it to relax back enough so when I flipped it and bolted it down there wasn't a lot of strain on the table frame. If I were to do it again I'd probably put it together out of four smaller pieces and biscuit / epoxy them into a solid table after having cut the grid into each of them. I think a lot of the problem I had was because it was one monolithic piece.
Anyway the long and short of it is the table is very stable now. I find it's the spoil-board that causes the changes in height, as you pull through the board you introduce whatever humidity is in the air and that causes the fluctuations in thickness. It can very significantly across the board after only a couple hours. Of course using a vacuum-jig like you showed eliminates that issue, unless you are using the table vacuum to hold the jig down. ;-)

[coryatjohn]

>> I find it's the spoil-board that causes the changes in height, as you pull through the board you introduce whatever humidity is in the air and that causes the fluctuations in thickness.

That makes sense! Never thought of that before.

[jerry_stanek]

Have you thought about granite for the base

[EricSchimel]

Have you thought about granite for the base

Surely that would be stable, but there'd be no reasonable way to cut a plenum into it.

[Chuck Keysor]

Not only would GFRC be far cheaper than granite, and lighter depending on the ingredients, you could route a form in which to cast the plenum layer. For that matter, one could integrate the base and the plenum layer as one cast piece.

[evan]

Not only would GFRC be far cheaper than granite, and lighter depending on the ingredients, you could route a form in which to cast the plenum layer. For that matter, one could integrate the base and the plenum layer as one cast piece.
I like this idea, although if you should happen to accidently send the bit through the spoil board (which I've never done... NOT LOL), there'd be hell to pay.

[jerry_stanek]

Surely that would be stable, but there'd be no reasonable way to cut a plenum into it.

If you used a wet saw you could cut the plenum and a hole saw for the vac hook ups. or just have a counter top maker cut them for you.

[Burkhardt]

One thing I have not seen mentioned is the stiffness of the thin table, given the 8x4' dimensions and the different expension coefficients (by temperature and humidity) vs. the steel frame. I think there is no way to make a thin slab that would not move a little with environmental changes. From all the earlier proposals I think most promising would be the granite (but hard to get in 4x8 and impossibly heavy), a thick and ground MIC-6 cast aluminum plate (really expensive) or the epoxy glass concrete (would probably need some surface grinding to be flat and be super heavy as well). At the end you may effectively be looking at building a new machine.

One low cost compromise may be to look for a sandwich torsion table design. That means lightweight rigid plates (e.g. good quality plywood) spaced apart (let's say 4 inches) by a grid of stringers. This hollow symmetrical sandwich could serve as vacuum plenum as well when sealed well. It would probably not be totally flat as constructed but the spoilboard surfacing could compensate for that. For that matter, if you glue a spoilboard on the top I would glue the same board to the bottom of the sandwich.

Now, I have no idea if that could be fitted to your machine and you may lose some z-travel. Just talking theoretically....

FWIW, my 3x4' machine has an 80/20 extrusion table. Works pretty well but it is just a bunch of extruded aluminum sticks. To make flat table that does not flex, I bolted them to a rigid sandwich table as mentioned above.

[curtiss]

Digital Vacuum gauge

Not sure who is selling this, but for those who have to be “digital everything”

https://springfield.craigslist.org/tls/d/springfield-digital-vacuum-gauge/7228719866.html