Vacuum Chamber

[rlara]

Hi to everyone:

I am trying to make a test vacuum chamber and wanted to know how to go about it.
Inside vacuum will go down to 6.67e-3 Pascal = 9.67 e-7 psi = 5 e-5 torr.
This vacuum chamber is small, 3 x 8 x 3 inches, machined out of aluminum 6061-t6.
My concern right now is to determine wall thickness in order to support the vacuum
Inside. To determine this I was thinking that the load on the outside walls of the chamber
would be the atmospheric pressure 1 ATM or 14.969 psi, and inside would be considered
as zero pressure, so I applied a pressure of 14.969 psi to the walls. My question is if this
is correct or is there other factors I should take in to account.


Thanks

 

[skills]

That's the way I'd do it. I have had to build several recently for a barometric sensor I designed.

 

[rlara]

thanks for the response Skills,

Well thats what i thought to, but im wondering why they make them out of thick steel. Whats the main purpose of the material thickness, is to make the chamber strong or to prevent diffusion throught the material?

 

[arto]

also a couple of other things to maybe consider:

- buckling/elastic stability - maybe not here, but on bigger vacuum vessels, it's the main criteria

-distortion/deflection & it's effect on sealing/gasketing surfaces

 

[MadMango]

You're applying 14.969psi to the walls, but you should also incorporate some level of safety factor (2x or 3x) to the wall thickness. You never know when your compressor or pump is going to go wacky, and you don't want an implosion. "The attempt and not the deed confounds us."

 

[hacksaw]

I've yet to see a vacuum chamber made of thick steel unless it was used in auto-clave work where the hot strength or mechanical stiffness, and material costs were an issue. A lot of small chambers are made from chemical resistant glass or stainless. There easier to clean.

Al works fine at the pressures you are looking at. If welded assembly is used, make sure you get good low-porosity weldments.

 

[skills]

Well, I would simply say this. The panel that is 3*8 will have about 360 pounds of force on it trying to push it in. Will that force cause enough deflection in the center of panel to cause a failure. I don't know. Young's modulus is about 11e6 psi and this is stress/strain or F/A/dL/L. At some point of area with a 14.7 psi pressure a container would 'beer can' but I don't believe you're vessel is nearly that size yet.

 

[skills]

I would add a rule of thumb to my above observation. As was pointed out above, the wall thickness must increase as the panel areas increase. I assumed that your wall thinkness is about 1/4" and that would be safe enough. The main concern I have here is that if you used 1/8" alum then your seals at the seams might be moving or worked to the point of failure. There is no danger of implosion with a 3X8X3 inch vessel holding a 14.7" vacuum at this size. 1/4" Plastic would probably be just fine; but, you would certainly see wall sinking occur about 1/4" and that might look weird.

 

[djw2k3]

MadMango,
I agree that a safety factor should be included but only in terms of material strength and deformation limits. Its impossible to draw more than -14.7PSI (or atmospheric pressure) in a vacuum, no matter what the vacuum pump does.

 

[Biggadike]

To my mind, a vaccum chamber is just a pressure vessel in reverse. The good thing is - you know the maximum pressure (i.e atmospheric).

Normal problems with pressure vessels as far as I know are:

1/ Material inhomogeny (welds, inclusions bubbles)which raise stress.
2/ Brittle failure due to strength of material required.
3/ Stress raising geometry at junctions to the vessel and other features additional to the nominal geometry.

The down side to vaccum chambers is they are difficult to do a 140% hysdrostatic proof test with. These have the advatage of pre-yielding the material in weak spots making it stronger.

 

[Biggadike]

My friend designs vacuum chambers for a living. Here are some of his comments:

********************************************************

Over engineer it massively essentially. The extra cost of huge weld fillets and thicker material is rarely of any consequence compared to the thing collapsing because it wasn't strong enough. Remember it's probably a one off, so not too much point in spending tons of time streamlining the design. And don't be afraid to put big strengthening ribs all over it to stop it bowing if it's a big bugger. This is important not just for strength, but avoiding repeated stressing (and hence cracking) of welds because of repeated pumpdown and venting operations. This unfortunate situatio happened to a friend of mine recently, because he did not strengthen his chamber enough, and therepeated flexing cracked a load of welds. Bummer.

As for shape, cylinders are good, obviously, but I have never had any problems with wierd shapes and square ones, providing you adhere to the above. If He is making a square type chamber with a door on it, then he would be well advised to make the door bowed in shape. I suyppose that's fairly obvious really.

I never use less than 3/8" thick stainless for what its worth.

Inclusions and other items you mention in welds are important not just for stress, but for the fact that they can create a "virtual leak" ie a little pocket of gas that bleeds out while you are trying to pump it down, and which you can't find. For this reason, all vacuum welds should be INSIDE the chamber, and any external welding for strength should be kept to stitch welds, and MUST NOT be full length. If you do get a leak on the internal weld, at least then you can trace it. This is true. I feel I am an authority since I designed a chamber just a couple of weeks ago, and will be off to see what it looks like this week. Mine is 1" thick stainless steel, but that's because of the application. I have had plenty of aluminium vacuum chambers also. Really it depends what pressure he wants to go down to. If he wants to go really really low then aluminium is not the best choice.

I can give loads of other advice, but it tends to be specific depending on useage, size and all those other lovely things.

The other thing is, when vacuum chambers collapse they don't go with a bang, they just kind of crumple. It is exceedingly funny when it is someone elses design. It has never happened to mine.

*********************************************************

If you have any questions, I'll put them to him.

 

[rlara]

Thanks to all the response

Well I am applying a safety factor, the concern was that when I made a simple stress analysis, my stresses were very small, around 1500 psi, reason why I thought there was something else I should take into account.

The vacuum level will be down to 6.67e-3 Pascal = 9.67 e-7 psi = 5 e-5 torr. Which as I understand it a
“high vacuum level” . But as one of you guys said the chamber is small, so pressure effects are not as big as in larger chambers. We have decided to make it out of .5 inch aluminum 6061-T6. BiggaDike said that aluminum could not be a good choice, at certain pressure levels, is the pressure above of concern when using aluminum?

rlara

 

[hacksaw]

Most metals are fine for your pressure range at room temperature.

Square chambers kinda/sorta are a problem because of the stress concentrations.

If you make it a cylinder with flat plate ends(o-rings or welded) you're better off. The cylinder is always in compression. only the end pieces need to support a tensile load (due to bending).

 

[Biggadike]

Ive forwarded your question to my friend..

I would like to know why you have decided to go for aluminium as the material of choice?

 

[Biggadike]

Words of wisdom from the vacuum sage..

Aluminium is fine at those pressures. It's actually more process related. Aluminium outgasses (self explanatory I think) more than stainless steel, so for critically clean processes, which generally require a much higher level of vacuum (much lower pressure) stainless is better. This is a good choice of aluminium, and is appropriate for these vacuum levels, and is the same grade I use for vacuum fixturing and chambers etc. It's what everyone else in the vacuum industry uses too, so that should be even more reassuring. While I'm waffling about it, aluminium is fine down to 1 X 10 -7 torr.

What are you using the chamber for?

 

[rlara]

The chamber will be used to test a microwave component. Most of the things we do are antenna components for satellites. The purpose is to test this component at space like environments. Not only will there be vacuum but also a temperature cycle, which makes this a lot more complicated. And to make it worst the component must operating during the environmental test so we will need some windows for this.

Concerning the thermal cycle (-20C to +50C) we were thinking of using a thermal plate and putting the chamber on top of the plate. But after talking to the guy who is selling us the pump, he recommended putting the plate inside the chamber saying it would make insulating allot easier. Problem with this is the plate is a big (18” x 7”) and it needs flowing water to cool down.

We use aluminum because we have it in stock and we can easily machine it. The inside of the chamber will be machined out; (i.e. we will machine a pocket on a block of aluminum, with large radius on the corners to prevent stress concentrations and a top.)

We thought this would be a lot easier, but by the looks of it this is not as simple. So we are contacting an expert for a quote, if it’s too much we will have to rig something out instead

Thanks

rlara

 

[Biggadike]

rlara,

More response from he who is wise...Most interesting.

There are lots of standard windows, liquid feedthrus, power
feedthrus, motion feedthrus etc etc on the market to help accomplish these things. I can advise on specifics if he wants. As for heating, remember how inefficient radiant heating can be, which is all you get in a vacuum. Personally, I would have thought Inconel heating elements mounted inside the chamber would give you a better chance at
getting the temperature up. Also there are a number of quartz type heaters on the market that work quite well in vacuum, though it

Make of that what you will.
sounds like the chamber is quite tiny.

Sounds like a bit of a fun project actually.

 

[sancat]

Just as a curiosity, how will the heat move from your cooling plate to your tested device in the vacuum?
I think that most of the heat a device in space receives or loses is transported by infrarred radiation, so you would need an infrarred lamp to heat it, and lots of time at a VERY COLD temperature on the aluminum case to collect the infrarred radiation your device will emit while cooling itself. I would also paint the chamber pitch black to ensure that all energy is absorbed, and none is bounced back
sancat

 

[vacuumsage]

I am in fact the one who was advising Biggadike.

It sounds like you are matching the size of the test chamber closely to the device. In this case, if the device is mounted intimately to the base or wall of the chamber, then it would make sense to use that wall of the chamber to provide heating / cooling. It sounds like this is what you originally intended. My comments regarding heaters are more suited to larger chambers.

One possibility is to machine channels in one wall, and use hot water to provide the +50 temp. I have done something similar to this successfully, but it really needs a good contact between your device and the thermal surface. The same thing could work for the -20 temp, but obviously you will need an appropriate coolant.

Hope this all helps.

 

[rlara]

Vacuumsage:

You are right, the component is small so we are trying to match the chamber to the component. We just got this peltier thermo electric plate that we want to use to produce the thermal profile. The product specification requires a certain profile (i.e. ramp up and down at specified rates) due to this we need to have control over the thermal sequence.

Have you had any experiance using these types of thermo electric plates?