Vacuum Tank Query (Aluminium, Carbon Fibre, etc, Composition)

[guesses3]

TL;DR Summary

A quick query about if I wanted to construct a vacuum tank made out of, say, aluminium or carbon fibre.

Hi,

I'm hoping this will be a fairly easy question for someone on here to answer. If I wanted to make a (small?) vacuum tank (basically a hollow cylinder capped with two similarly hollow hemispheres - like so: (===)), how thick would the walls of the tank need to be for it to withstand the internal pressure going quite low (eg. 1 Torr or 1 mBar) at sea level? I would be looking to make it out of a strong but light material like aluminium or carbon fibre. I just need an estimate - I simply have no idea (1mm, 2mm, 5mm, 20mm?)

If there were some fairly simple equation (for estimating the approximate thickness based on the internal pressure of the tank) for each material, that would be really useful

Cheers.

 

[Baluncore]

Welcome to PF.

A high vacuum on the surface of the Earth can never be greater than atmospheric pressure.
Say 14.5 psi.

It is not thickness that is important unless the wall is flat. The wall needs to be designed so it will not buckle out of shape and collapse. Internal or external rings may give it the rigidity it needs.

What is the diameter of the tank?
Bigger tanks have flatter walls so they are more likely to buckle.

The choice of material will be decided by contamination and leakage rate.
What is the vacuum connected to?

 

[guesses3]

Welcome to PF.

A high vacuum on the surface of the Earth can never be greater than atmospheric pressure.
Say 14.5 psi.

It is not thickness that is important unless the wall is flat. The wall needs to be designed so it will not buckle out of shape and collapse. Internal or external rings may give it the rigidity it needs.

What is the diameter of the tank?
Bigger tanks have flatter walls so they are more likely to buckle.

The choice of material will be decided by contamination and leakage rate.
What is the vacuum connected to?

Hi,

Thanks for getting back to me. Let's say the tank is 50cm in diameter. I'm not sure what you mean by what the vacuum is connected to - the tank would just be connected to a vacuum pump in the air.

Is there an equation that could give me some sort of indicative estimate of the thickness required to ensure integrity based on the tank diameter and the internal/external pressure differential (ie without the need for bracing)?

 

[Baluncore]

I'm not sure what you mean by what the vacuum is connected to - the tank would just be connected to a vacuum pump in the air.

If it was only connected to a vacuum pump you could throw the tank away, along with the pump. What are you not telling us?

Is there an equation that could give me some sort of indicative estimate of the thickness required to ensure integrity based on the tank diameter and the internal/external pressure differential (ie without the need for bracing)?

You are focused on the wrong aspect. A 1 mm wall of almost anything would be sufficient. The wall thickness will be determined by the way you mount the tank and how flexible it is. Also the method of manufacture and how you make connections and seal any seams.

 

[Baluncore]

https://www.physicsforums.com/threa...ness-of-a-vacuum-chamber.986815/#post-6322290

 

[guesses3]

If it was only connected to a vacuum pump you could throw the tank away, along with the pump. What are you not telling us?
You are focused on the wrong aspect. A 1 mm wall of almost anything would be sufficient. The wall thickness will be determined by the way you mount the tank and how flexible it is. Also the method of manufacture and how you make connections and seal any seams.

I don't understand - why does the tank need to be mounted to something? And why could I throw the tank and the pump away if they are not connected to anything else? What am I missing? At the moment I am just looking into the hypothetical ability to create a single-piece vacuum tank that would not implode. If you are telling me that a single-piece tank made of 1mm thick aluminium or carbon fibre wouldn't implode when evacuated at sea-level, then that's great!

 

[berkeman]

I don't understand - why does the tank need to be mounted to something?

As discussed in the other thread linked by @Baluncore things like access doors, windows, pump connection points, etc., can be the weakest parts of such a vessel, so they need to be part of the overall design.

What is your application? What is your background with high vacuum? What pump(s) are you planning to use? What is your strategy for leak detection and fixes?

 

[Chestermiller]

I think what you really need to be concerned about is not a strength issue, but a compressive buckling issue. This happens to cylindrical vessels subjected to a higher pressure outside than inside, like, for example, submarines. The buckling takes the form of modal panels with 3 or 4 sides, depending on the ratio of the length to the diameter. You can see this for yourself if you take an empty 2 liter soda bottle and suck some of the air out. It very rapidly goes to a 4 model buckle shape. For high pressure differences, you need to worry about the structure severely buckling. There are analyses of this phenomenon in the literature, and, of course, you can analyze it yourself as a buckling problem.

 

[guesses3]

As discussed in the other thread linked by @Baluncore things like access doors, windows, pump connection points, etc., can be the weakest parts of such a vessel, so they need to be part of the overall design.

What is your application? What is your background with high vacuum? What pump(s) are you planning to use? What is your strategy for leak detection and fixes?

I don't have an application as such yet. But as I said, at the moment I am just considering a single-piece vacuum tank with a pump attached to it - no access doors, windows, etc, just the pump connection point. I have no background in high vacuum. Leak detection does not seem to me a great concern as the pump will be attached and I'm not planning on taking the pressure down to ridiculously low levels.

 

[guesses3]

I think what you really need to be concerned about is not a strength issue, but a compressive buckling issue. This happens to cylindrical vessels subjected to a higher pressure outside than inside, like, for example, submarines. The buckling takes the form of modal panels with 3 or 4 sides, depending on the ratio of the length to the diameter. You can see this for yourself if you take an empty 2 liter soda bottle and suck some of the air out. It very rapidly goes to a 4 model buckle shape. For high pressure differences, you need to worry about the structure severely buckling. There are analyses of this phenomenon in the literature, and, of course, you can analyze it yourself as a buckling problem.

Yes, buckling is my main concern. And I guess that will be an issue when using a cylindrical shape. Presumably an oblate spheroid design instead would be more robust?

 

[berkeman]

I don't have an application as such yet. But as I said, at the moment I am just considering a single-piece vacuum tank with a pump attached to it - no access doors, windows, etc, just the pump connection point.

Well, those things are important to include even in your initial investigation.

As an analogy: in my EE work, we often deal with RF shielded chambers (and even full-size shielded rooms) for various kinds of testing of RF devices. The weak/RF-leaking parts of such enclosures are usually the access doors and cable feed-throughs that are necessary for such enclosures.

If you can't open up your vacuum chamber, you can't put anything inside it or take it out later. Not a very useful chamber, IMO...

 

[guesses3]

If you can't open up your vacuum chamber, you can't put anything inside it or take it out later. Not a very useful chamber, IMO...

Or at least you can't put anything in or take it out while it's evacuated?

As I said, basically at this stage all I want to be able to do is evacuate the tank without it collapsing. If 1mm thick material is fine - provided the design is appropriate (shape + bracing) - then I'm very happy with that

 

[berkeman]

Or at least you can't put anything in or take it out while it's evacuated?

As I said, basically at this stage all I want to be able to do is evacuate the tank without it collapsing. If 1mm thick material is fine - provided the design is appropriate (shape + bracing) - then I'm very happy with that

Please just buy one for your initial experiments. Here is a common one that should be safe and meet your initial needs:

https://www.zoro.com/sp-scienceware...KRe0p1huEpAireAcKayfnlVc-L7TWHDxoCE3sQAvD_BwE

And you didn't answer my question about what pump(s) you will be using, or what your application is...

 

[Baluncore]

Yes, buckling is my main concern. And I guess that will be an issue when using a cylindrical shape. Presumably an oblate spheroid design instead would be more robust?

A short cylindrical section with spherical ends would be better than an oblate spheroid.

Your hypothetical vacuum chamber, floating in the atmosphere without a connection to the solid world will need a thicker wall than one that has solid mountings. That is because the mountings can be arranged to prevent buckling, and so allow a thinner wall. Wall thickness can be the last consideration once the supports, shape and connections are identified.

 

[Baluncore]

And why could I throw the tank and the pump away if they are not connected to anything else? What am I missing?

Engineering minimises the cost of meeting a requirement. If you will not use the vacuum, apart from testing the pump, then there is no reason to have a vacuum chamber. That suggests you do not need a pump or the power to run it. How will air flow into the chamber if there is no inlet connection apart from the pumped outlet ?

You have generalised your question to the point where there can be no helpful answer. You now want an answer to the question of minimum wall thickness when there is no decision yet on shape, mounts or plumbing fittings. The wall thickness will also depend on the wall material, and that will dependent on compatibility with the source of fluid that flows into the tank from the unmentioned source.

Why do you insist on designing it backwards?

 

[Chestermiller]

Yes, buckling is my main concern. And I guess that will be an issue when using a cylindrical shape. Presumably an oblate spheroid design instead would be more robust?

If you really want to do this right, you probably need to do some serious finite element structural modeling (probably a shell model), including buckling stability perturbation analysis as you evolve to a prototype design. If you don't have the experience to do this yourself, I recommend getting help from a consultant. Anyway, that's what I would do.

 

[guesses3]

Thanks guys for being so generous in sharing all your knowledge. I appreciate that I have been rather vague, but that's because I'm just trying to ascertain whether or not what I would like to achieve is technically feasible before attempting a proof of concept, and from what you've said it looks to me like it is. Sorry for not being more forthcoming, but if anything does come of this (not any time soon!) I will be sure to give you guys credit for the help you've given me

Just one final thing: Baluncore said "A short cylindrical section with spherical ends would be better than an oblate spheroid", but then the vacuum dessicator that berkeman recommended appears to me to be basically oblate spheroid in shape...

(The reaonsing seems to me self-evident given that a sphere is inherently more robust structurally than a cylinder.)

Anyway, thanks again - your help has been very much appreciated!

EDIT: And yes, I will definitely be leaving the actual implementation to experts

 

[Baluncore]

(The reaonsing seems to me self-evident given that a sphere is inherently more robust structurally than a cylinder.)

For a fixed radius, the poles of an oblate spheroid are flatter than the equator of a sphere, or the surface of a cylinder, so it will need thicker walls.

An oblate spheroid is flatter and has less volume than a sphere, so it is a doubly poor modification to a sphere.

Better to insert a short cylindrical section into a sphere to get the most increase in volume for the least increase in surface area with a fixed wall thickness.

 

[bobob]

There isn't a simple equation that will answer your question. A better question is do you want to build a vacuum chamber or is this an exercise of some sort? If you want to build a vacuum chamber, how light does it have to be? How big does it have to be? If you aren't seriously constrained, just make it cylindrical, weld a cap at one end and a weld a flange on the other end so you can use o-ring seal for the mating flange and go for a little overkill. Without knowing your requirements, giving you more informations is difficult although I wouldn't use carbon fiber. Carbon fiber will outgas. I'm also not sure that using aluminum (at least 6061) over stainless will make your chamber lighter, since stainless is a lot stronger than aluminum so you could make the walls thinner with stainless. With standard 6061and a reasonable definition of small, consider a thickness between 1/8" and 1/4".You can also consider using an aluminum alloy like 7075T6 which is comparable to some steels in strength, but is more difficult to work with and can't be welded (if I recall correctly).

In any case, if this is a question about building a real vacuum chamber, then a little overkill is highly desirable so that you have a safety margin.

 

[Nik_2213]

Vacuum dessicators, IMHO, are an uneasy compromise between handling ease and robustness. Traditionally, they were made of glass-- thick, annealed, mating faces ground to plane, the tap port doubly or triply thickened against handling accidents. As glass will flex very slightly, the body had to be extra-thick for stiffness, with over-sized lips to tolerate handling chips and increase any leakage path lengths. A smidgeon of vac-grease, typically Apiezon (TM) or its kin, was the final touch...

Always, always, there was the worry that any tiny, un-noticed scuff or scratch might 'Go Large' and implode, scattering fragments like a broken 'Bottle TV' tube. Be NOT There !

The tough plastic versions are a lot safer in that respect, but much less tolerant of local heating...

DIY or not, consider adding 'safety mesh', a stronger, lab-grade sleeving than that used to 'sock' Christmas trees...

Slightly OT: When we dried or ashed lab samples, we generally cooled their crucibles in ordinary, non-vac glass dessicators, whose base held a dish of P2O5 dessicant or activated silica gel crystals. Before lid taps were routine, this usually meant leaving the lid slightly ajar. Snag was warmed air expands, and lids placed just-wrong could easily shift slightly, become totally clamped to base when that air cooled.

Samples could be re-tested with fresh material, but extricating crucibles without breaking them or the non-vac dessicator seemed impossible. Much bad language generally ensued, followed by a shout for me.

For a small-ish dessicator, I'd fasten an elastic band around the mating surfaces. Bigger dessicators had to be slanted. Pop into *ambient* vac-oven, pump down very, very gradually. When pressure difference sufficed, the dessicator's lid would hop. Either it landed sufficiently skew, or band would slip into gap and prevent re-sealing. Often, process left sample intact, too ! ;-) ;-) ;-)

 

[arydberg]

Vacuum tanks are used in various cars. here is a link. They are strong enough.

https://shop.advanceautoparts.com/c4/vacuum-reservoir/13673

 

[Baluncore]

They are strong enough.

They have a small volume, and are tightly curved.

 

[enigma]

Why on Earth would you want to design and build it from scratch without the expertise to spec it? Go online and source a component to buy. I guarantee that COTS will cost less than designing and testing a new part, even if you need to have a custom interface designed by the vendor.

 

[guesses3]

Why on Earth would you want to design and build it from scratch without the expertise to spec it? Go online and source a component to buy. I guarantee that COTS will cost less than designing and testing a new part, even if you need to have a custom interface designed by the vendor.

I am not wanting to actually design and build it: my questions were simply to establish feasibility. That is the fundamental pre-requisite of any project. Now that I know that it is at least feasible, I can progress from there in an appropriate manner.

 

[guesses3]

There isn't a simple equation that will answer your question. A better question is do you want to build a vacuum chamber or is this an exercise of some sort? If you want to build a vacuum chamber, how light does it have to be? How big does it have to be? If you aren't seriously constrained, just make it cylindrical, weld a cap at one end and a weld a flange on the other end so you can use o-ring seal for the mating flange and go for a little overkill. Without knowing your requirements, giving you more informations is difficult although I wouldn't use carbon fiber. Carbon fiber will outgas. I'm also not sure that using aluminum (at least 6061) over stainless will make your chamber lighter, since stainless is a lot stronger than aluminum so you could make the walls thinner with stainless. With standard 6061and a reasonable definition of small, consider a thickness between 1/8" and 1/4".You can also consider using an aluminum alloy like 7075T6 which is comparable to some steels in strength, but is more difficult to work with and can't be welded (if I recall correctly).

In any case, if this is a question about building a real vacuum chamber, then a little overkill is highly desirable so that you have a safety margin.

So carbon fibre is a gas-permeable material (ie not "vacuum-tight")? But aluminium/stainless steel are not gas-permeable?

 

[Baluncore]

So carbon fibre is a gas-permeable material (ie not "vacuum-tight")? But aluminium/stainless steel are not gas-permeable?

The choice of material will be decided by contamination and leakage rate.
What is the vacuum connected to?

You have failed to identify the application of the vacuum tank, so the leakage rate and potential for contamination cannot be assessed.
Why is the use of the vacuum confidential?

 

[guesses3]

You have failed to identify the application of the vacuum tank, so the leakage rate and potential for contamination cannot be assessed.
Why is the use of the vacuum confidential?

By definition, I can't provide a meaningful reply to that question

 

[Baluncore]

By definition, I can't provide a meaningful reply to that question

By definition of what ?

 

[berkeman]

Thread closed for Moderation...

 

[berkeman]

By definition, I can't provide a meaningful reply to that question

Best of luck. Please be safe in your project. Thread closed.