www.TK560.com

[drcrash]

I did an experiment with a well-sealed platen but not using any kind of edge sealing---not forming plastic around the edge of the platen, or having a gasket, or anything.

I just put a cold sheet of plastic (1/16" HIPS) flat on the platen, and sucked it down with my two-stage vacuum setup.

The platen is my 12 x 16 inch many-hole platen, which is really 9.5 x 13.5 inches, giving it a 46-inch perimeter. The outermost holes are about an inch from the edge, and the surface is just epoxy-coated MDF. (Pretty smooth, but not perfect.) That gives a roughly 1-inch wide band of flat epoxy surface about 46 inches long for the plastic to suck down to and make a "seal."

First I did it with just a 7-gallon (about 1 cubic foot) tank, and the pump off, so I could figure out the flow rate by measuring how fast the tank filled up.

When I opened the main valve, the vacuum level immediately dropped by 3 inches of mercury, from 28 to 25, then dropped steadily by half an inch per second, so that after 10 seconds it was 20 inches and after 20 it was 15.

(Between a perfect vacuum and a half of an atmosphere, flow is not much dependent on the vacuum level; it depends almost entirely on the size of the leak, and half an atmosphere of pressure difference only changes the flow rate by a percent or two---less than my measurement precision.)

That tells me that the flow rate of the leakage under high vacuum is about 1 cubic foot per minute---each inch of mercury filling a 1-cubic-foot tank is about 1/30 of a cubic foot, and so 1/2 in Hg. per second is 1/60 of a cubic foot per second or 1 cubic foot per minute.

That's an interesting baseline. No vacuum former should perform any WORSE than that, if its edge seal is working at all, or even if it's not working, but is not managing to make things worse. Flat cold plastic against flat cold plastic. If you're losing more than about 1 CFM for a small platen, or 2 for a 2 x 2 foot platen---which has less than twice the perimeter length---you have a big leak somewhere.

So if you have a 30-gallon (4 cubic foot tank) and a 2 x 2 foot platen, your vacuum level should only drop about half that fast, as long as you suck the plastic down flat to the platen around the edges and keep it well sucked down there. (After the initial pulldown, that is.) You have about four times as much volume to fill and only twice the perimeter, so once you've sucked most of the air out of the platen and mold(s), your vacuum level should only drop about 1/4 of an inch of mercury per second, down to around 12 or 15 inches. (After that it gets weirder to calculate, because pressure differences matter more.)

For example, even if your vacuum drops from (say) 27 to 20 when you pull the plastic down and pull most of the air out of the molds and platen, it should take another 20 seconds for it to drop to 15 inches. So after 20 seconds, it should still be pulling 2.5-3x as hard as most good vacuum cleaners. If it's not, you have a big leak. If your platen is not sealed on the bottom and sides, that's probably it---you are sucking more air straight through three square feet of unsealed MDF or wood than you are actually using for vacuum forming.

I did the same experiment with the pump on, so that it would continue pulling air out as air leaked in. (You probably know how that works, but the basic deal is that the pump is teed to the tank on one side and the platen on the other in my two-stage setup, so that the pump can pull from the platen, or from the tank, or both, with ball valves to control which.)

My pump does about 3 free air CFM (it should do 4.5 but I have an undersized check valve & hose barbs, oops).

At a high vacuum level, like 28 inches of mercury, that's really not much at all because a cubic foot of air doesn't have much air in it---about 1/15 as much air as normal atmospheric-pressure air, so about 1/5 of a CFM.

---

That's something to keep in mind about pumping reasonably high vacuum against leakage. The air that's leaking in is several times as dense as the air that you're pumping out, so whatever the CFM of your leakage, you need several times that to maintain a vacuum level. How many CFM you need is proportional to the leakage rate times the ratio in the pressures.

Conversely, the smaller the difference in the pressures, the lower a ratio you need to maintain that difference, so a slower pump can keep up at lower vacuum levels. Maintaining 3/4 of an atmosphere of vacuum takes about twice as much pumping as maintaining 1/2 of an atmosphere, because the air you're pumping out is half as dense, and maintaining 7/8 of an atmosphere takes four times as much pumping, because the air is one fourth as dense. (It's not quite that simple, but that's the right general ballpark.)

---

With the pump on, when I opened the valve, the vacuum level immediately dropped from 28 to 25 as the air was sucked out of the platen and plumbing, then began dropping at about 1/2 in Hg per second (as before), but the drop rate quickly slowed so that after 20 seconds, it was still 20 in. Hg. Then it kept dropping, more and more slowly, so that after 30 seconds it was 18.5 inches, and after 1 minute, it was 15.5. It stopped dropping after about a minute and a quarter, and held steady at about 15 inches of mercury.

That's an interesting baseline, too. For a small platen like mine, a 3 CFM pump should be able to maintain half an atmosphere of vacuum *indefinitely* even without an edge seal.

For a 2 x 2 foot platen, 6 CFM should do that---if you have a 6-or-more CFM pump, your vacuum should never drop below 15 in. Hg; if it does, you have serious leakage somewhere.

(Of course, if you get a good edge seal, it should be able to hold a substantially higher level of vacuum indefinitely--20 or 27 inches, say, or even more, depending on just how good your edge seal is and your pump's specs.)

By the way, when actually forming plastic over my 12 x 16 platen, and forming the hot plastic around the epoxy-covered-MDF edge, I do get a much better seal and maintain a much higher vacuum---the vacuum level falls to around 25 but rises again to around 27 and stays there---only an inch or two below the pump's maximum vacuum level. (So far, anyway. That's for 1/32 and 1/16 HIPS.)
_________________
Paul (a.k.a. Dr. Crash)

Tired of buying cheap plastic crap? Now you can make your own! www.VacuumFormerPlans.com

[drcrash]

I did a crude experiment to measure how much air I could suck through a sheet of MDF with high vacuum.

It's a lot---about 1 cubic foot per minute per square foot of surface, give or take a third of a CFM either way. So maybe three CFM for the bottom of a TJ 2 x 2 platen.

I was using 1/2" MDF, but I wouldn't expect 3/4" to be much better. (The outer surface is what matters, and the core is very porous.) Your mileage is likely to vary from brand to brand or even batch to batch, though. Porosity of MDF varies.

For reference, for a platen the size of my small one, that's about 1.5-2x as much leakage as I get around the edges of the plastic, even with no edge seal at all. (Just cold flat plastic sucked flat against cold flat plastic.) For a large platen, it would be worse, because the area goes up faster than the perimeter length when you scale up.

If you have an unsealed MDF platen and aren't able to hold a good vacuum for a long time, that's probably why---you're losing most of your vacuum straight through the MDF. (If you're getting a good edge seal, the large majority is likely going through the MDF bottom.) Seal your platen before considering a bigger tank or a faster pump.
_________________
Paul (a.k.a. Dr. Crash)

Tired of buying cheap plastic crap? Now you can make your own! www.VacuumFormerPlans.com

[knightshade]

Very interesting topic.

I'l have to compare timings... but I've noticed that with my pump running, Over time I drop to 7 in/Hg before things level out.

I'll have to validate, but I seem to recall that my pump is a 10CFM pump.

That said, I've got a 2' platten made out of MDF, and I've only sealed the sides.

Guess that I need to re-evaluate the platten.

[jegner]

Thanks Dr. Crash for the info. Good to know. Sealing the next time I have my machine down for repairs.

Jim

[sethb6025]

That is very interesting Dr Crash, thanks for putting that theory to the test.