vacuum: how hard for how long?

[drcrash]

I'm trying to figure out how to determine how much vacuum you need for various setups.

I have some little vacuum formers (roughly 12" x 18") that I use with a two-stage vacuum system with a vacuum cleaner and various small pumps (including some very small ones). I *think* I've learned something interesting from those, and want to apply the ideas to somewhat larger machines.

One of the reasons I'm interested is that I'd like to design relatively small inexpensive machines, relative to the size of plastic they can form. Small pumps and small tanks are both good, both for cost and for compactness.

With my system (a variant of Doug Walsh's two-stage design) I can suck the plastic down (mostly) with a vacuum cleaner, then open a valve to little 7-gallon (~ 1 cubic foot) tank for the hard pull. The vacuum level goes up to about 24 inches of mercury almost instantly, then fades quickly over the next several seconds, usually dropping two or three inches per second, until it levels off around 8 a few seconds later, or falls so low that low-vacuum system kicks back in (a check valve opens) and holds it around 5 inches. (Depending on just how small and slow a pump I'm using, and whether it can keep up with leakage around the platen edge at all.)

I know this is pretty lame, and I intend to improve it in several ways, but it seems to do a surprisingly good job---much better than a vacuum cleaner alone, and I'm guessing almost as good as having a tank several times as big and vacuum that fades several times as slowly. I *think* I'm getting most of the benefit of a much better vacuum system.

My mental model of this is that to stretch the plastic into/around the details only takes strong vacuum for a brief period, and a weaker and weakening vacuum is enough to hold it in place and keep it from springing back, losing the detail. (At least for relatively thin plastic like 1/16" or 1/8".)

For a VERY brief period, maybe a fraction of a second, the plastic is mostly being stretched like rubber, and then over the next second or so, the plastic mostly flows, with polymer chains sliding over each other into a new configuration. The plastic continues to flow more slowly for several seconds after that, relieving strain and reducing the force required to hold it in shape without springing back. Then over the the following several seconds, the plastic cools enough to harden.

If I'm right about this, it seems like a brief period of strong vacuum followed by a fairly rapid decline is just fine---as long as

1. The vacuum is strong enough initially to pull the plastic into the details, and

2. The vacuum fades more slowly than the plastic flows, so that the force is always at least what's required to keep the plastic from springing back, and

3. After the plastic has pretty well flowed and you're waiting for it to cool below its thermoforming temperature, you maintain enough vacuum to keep it from warping. (Presumably that's a lot less than the force required to form it.)

Does this sound right?

I'm thinking that the amount of vacuum you need---how slowly you need the vacuum to fade---has to depend on the thickness of the plastic sheet and the flow rate of the plastic it's made of.

If the plastic is thicker, more force is required to stretch the rubbery sheet and hold it stretched. Assuming you start at most of an atmosphere, though (getting less detail to start with and accepting that), it will decrease similarly after a certain point, going down by a certain fraction per second as flow within relieves the strain of stretching.

There's a subtlety there---for thick plastic, you need to hold the initial strong force a little longer, because it takes a little longer for the rubbery sheet to stretch into the details.

But after you've gotten most of the detail you're going to get, and just need to hold the shape while flow relieves strain, the vacuum can decline just as rapidly. The slope of the force-required-to-hold-it curve is the same after that point where you stop getting more detail.

If that's right, you don't need a whole lot more vacuum for thick plastic than for thin plastic. Just a fraction more, to mostly delay a relatively fast fade in vacuum level.

Does anybody know if that's right? Has anybody seen a discussion of these issues, maybe with some rules of thumb about how fast it's okay for your vacuum level to decline?

Is this consistent with people's real-world experiences?

I'm also wondering about different plastics with different specific heats and different flow rates.

For example, if a plastic has a high specific heat, so it takes longer to heat and to cool, that's not necessarily a problem. If the flow rate is similar enough, it will form reasonably quickly, and the strain will be relieved reasonably quickly, so essentially the same amount of vacuum is required to form the plastic and hold it in shape. You just have to hold a certain minimal level of vacuum after that while you wait for it to cool, so that it doesn't warp. (Just falling back on vacuum-cleaner-level vacuum might be enough, without needing a bigger tank or vacuum pump.)

On the other hand, if the flow rate is low, you need the vacuum to fade more slowly, because the strain won't be relieved as quickly.

Any comments would be very welcome.

[johnd]

The higher the vaccum the better off you are. This affects cooling time, and overall detail. Most hobbiest form by drape method. They push a male or female mold into a soft sheet and turn on the vaccum. Air pressure does a pretty good job of pushing the material down into or over the mold and seals until the plastic material cools enough to leak and then you loose vaccum. When you loose vaccum you loose detail, and when you go to strip the part off the part typically locks on the mold. Commercial formers use some type of perimeter seal on their mold to eliminate loss of vaccum and to allow atmospheric pressure to push down on the "plastic" sheet. If you are running 29 inches of vaccum or what ever, you try to maintain this thru the cycle. This mean you try to only evacuate the initial mold area and maintain this until you are ready strip the part from the mold. The perimeter seal can be made by using a clamp box with a mechanical seal or some form of o-ring locking the material to the mold surface. Other materials such as the oilefin families usally combine a moat groove around the perimeter of the mold along with a clamp box sealing to the mold surface.
Been there and done it for a few years.