Vacuum chamber walls in tension?

High strength vacuum chambers could be built with the proper configuration of shell, but your example of a vertical box with cylindrical walls does not provide the whole story. While it is true that the cylindrical walls could sustain tension, the problem is that they dump large reactions at each vertical corner of the box. This would require strong beams spanning from top to bottom which in turn would require compression members top and bottom to carry the beam reactions.

One thing you have in your favour is that the internal pressure is uniform throughout the structure. You don't have to contend with unbalanced loading conditions (other than perhaps wind and snow outside the structure).

Probably the best configuration would be a spherical dome. Buckling is still a problem, but you could create the dome out of folded plates or use stiffening ribs in radial and circumferential directions.

As I understand well you need to build a vacuum chamber of 100mm at its smallest distance.

Will this be one free room or is it allowable to have columns in the room?

Design it as a tent. Calculate the structure as if you need to support 1m of water over the whole structure.

You will end up with chain line structures. It is a bit different than hyperbolic.

In excel it is called the COSH function.

There are a couple of applications I'm interested in.

There have been a few proposals floating around about making launches from the ground for small satellites by bringing the rocket to full orbital velocity on the ground then directing it up to space once it reached sufficient orbital velocity.

Because of the air pressure on the ground it would be advantageous to do this acceleration in a vacuum chamber. However, you might need the chamber to be 10 to 100 meter across for this application. Vacuum chambers this wide and only a few meters long are normally massive constructs. For this application, the chamber would have to be kilometers long. It would be advantageous to be able to make this without having to use for example a million tons of steel.

For the second application, it has been known for some time now you can make much cheaper telescopes if you make the surface be a liquid and rotate the container so that the surface forms a paraboloid. The cost can be up to 100 times cheaper for large scopes. A six meter scope has been made this way.

There are plans for a solid mirror scope 100 meters across. Such a scope might take ten years and cost a billion dollars to make. If this were done with a liquid mirror telescope it might take a year or two and cost $10,000,000 dollars.

However, there are problems with the liquid mirror scopes the larger you make them. A key problem is the large container and mirror rotating itself produces such strong winds that it destroys the surface flatness of the liquid mirror. Therefore what I want to do is create a vacuum chamber large enough to contain the mirror and the entire building for such a large mirror. This would take a chamber at least a 100 meters across and perhaps a few hundred meters high. For the chamber this high also you would want also to use a minimum amount of material.

You see for both of these applications you don't want support members crossing in the middle of the chamber.

Bob Clark

Application 1: how will you enable the sattelite to leave the concealed vacum chamber and move up to the space? The sattelite will hit the air at a level which is way worse then the return of a spacecraft, it also has to travel through the air for several hunderds of Km before it is in space (unless you release vertical, which make you add extra fuel to move to an orbital path)

Escape speed is approx 11 Km/sec (a bit more) Good luck with your heat shield and shutter valve.

Application 2: a transparent roof, 100m diameter with a complete vacuum beneath: the stresses in the glass/polymer layer will make the passing light unusable.

I can see where you're thinking goes, but in the first application, how do you propose to transition the rocket from the vacuum chamber to atmosphere? I see a significant problem there, and also with fireproofing the walls if you want a membrane over a frame for the structure. For the second application, I've read about liquid mirrors, and recall them being mercury (Hg). This evaporates, and I suspect lowering the pressure on it will only enhance that effect. How do you propose to control that? Also, the transparent membrane cover (if it even exists) will certainly complicate veiwing. Computers are used to compensate for atmospheric distortion (twinkle effect) by comparing to an artificial (laser) star. The computer power algorithm may be able to account for the membrane, I can't say...

Yep, as the others have said your vacuum building would need to extend all the way into space.

For the parabolic mirror you could spin molten glass in a vacuum and let it set. See

http://cr4.globalspec.com/thread/16602

Hi to everybody,

this information about what to do and how to use did clarify a lot.

I think the problem has to be divided into two different ones:

First should be pretty long and not too big in diameter to launch inexpensive satellites.

Second should be big in diameter and in height for an astronomical imaging telescope, this I would assume to be impossible as stated by others above that the optical properties of the dome will never be good nor the dome be cheap.

There is an existing example: the South African howitzer can shoot at a distance of45km (the NATO reaches 23km!) .

The air in the gun barrel in ordinary guns is compressed by the projectile and this is limiting the range.

So if we assume a position on an elevated place 5000m high most of the air is below.

This is not a problem of the launching tube but a problem of turbulence and drag in the atmosphere.

You will have to find a lightweight membrane that can be blasted away some microseconds before the accelerated "satellite" is passing. This would not be a big problem as fabrics made from cellulose-nitrate coated with a high explosive will do this.

May bet he "satellite" can be hardened a bit so the membrane can be cut by the nose of it.

Let us calculate the length of the vacuum tube :

assume a constant acceleration of a=10g or 100m/s² , a necessary velocity of V=14km/s (too much I assume, to be modified if we know more about the drag), then the tube length has to be L=v²/2a or 1000km. (250km if 7km/s is sufficient)

Time to accelerate would be only 140 (70) seconds!

So find a mountain where you can bury your gun barrel, go with acceleration to 1000g get a length of 100 (25) km. This is still impractible.

So it would be a good idea to have a two stage design? First as a cannon second as a rocket?

If we take a more practical 1km length of the gun barrel the available acceleration time is t=√(2L/a) = 4.5s.

Compared this with the total necessary acceleration time of 140 (70) s , the result is: this is not useful.

So only a very high acceleration device may be suitable.

For medium size satellites or pure transportation to ISS or other satellites it may be interesting.

Remaining problem: how to isolate the accelerating projectile from the tube?, or better
to make contact with the wall as in a real cannon? How to accelerate if there is no contact between tube and projectile?

The once propagated railgun seems to no longer alive. A somewhat different approach is still existing but there is lack of information because of military funding.

What is the community thinking?

RHABE

"What is the community thinking?"

Well, the part of the community that I can speak for (me) thinks that was a good answer, and have rated it as such. I read (past tense) a lot of literature, both scientific and scienceficticious, about the proposed railgun some years ago, but as you pointed out, it seems to have gone moribund. I do not know why, it seemed to be an ideal solution for launching inanimate payloads. I suspect it was a political decision unless there was some technical difficulty that I missed hearing about, or that wasn't publicized well. If it was/is a military project that's gone black, we won't know more until they shoot something with it, and maybe not even then. Curses!

http://www.utexas.edu/research/cem/Electric%20Gun%20Integrated%20Launch.html

Hi,

the above link seems to be active work on an improved type of electromagnetic gun.

Driven by current pulses from a decelerating big rotor in a big generator.

Current pulses feeding coils that circle around the gun-barrel thus generating magnetic fileds that interact either with aluminum cylinders (outer wall of the moving mass) or interact with longitudinal permanent magnets (many of them with small diameter to suppress eddy currents).

Current pulses (high voltage and very high current) are switched by ultrafast plasma switches (? my estimate).

The current pulse has to be shorter for the next coil according to the gain in velocity.

The peak current should be higher for any next coil to get the same velocity increase per coil. As the generator will run down with acceleration of the moving mass, I think it necessary to have any coil equipped with a pulse transformer that is compressing the extracted energy into a shortened on-time with amplified current amplitude.

In total an approach that should be regarded as promising.

Any rocket transportation does cost more than 100times the theoretical energy because the rocket has to accelerate its fuel that is necessary to carry with it and the structure to support it.

RHABE

.

"Any rocket transportation does cost more than 100times the theoretical energy because the rocket has to accelerate its fuel that is necessary to carry with it and the structure to support it."

Right on - which is why rail guns (of any configuration) have always made good sense to me. The fuel stays on the ground where it belongs... And for strictly cargo shipments, the acceleration can be anything the packaging can be designed to withstand.

This question is somewhat off the OP topic, but seems relevant to the direction the discussion has turned: is the orbital launch rail gun any more or less feasible than the space elevator given current state of technology?

Here's one report on a group investigating reaching such high speeds on the ground:

Huge 'launch ring' to fling satellites into orbit.

http://technology.newscientist.com/article/dn10180

They are only intending to launch small satellites using this method, 10 kg and less. For such small satellites you could design the shell holding it in the form of a thin missile that could survive the very high Mach speeds while still in the lower atmosphere.

However, while doing a web search I found a report on creating inflatable vacuum chambers, where the walls are filled with pressurized gas for strength. Such chambers could even be buoyant if the walls were filled with a lighter than air gas such as helium.

This then could be used to extend a vacuum travel path from the ground all the way to high altitude for orbital rocket launch.

It also would allow the vacuum shroud for the telescopes to reach all the way to the stratosphere, below which most of the atmospheric distortion takes place.

Stability Analysis of an Inflatable Vacuum Chamber.

http://arxiv.org/abs/physics/0610222v4

Bob Clark

Try to buld one (small scale) or to calculate or both if you don't trust the results.

Or think about a slingshot or a trebuchet?

RHABE

An air supported structure would resolve one of the structural obstacles, namely the design of a frame capable of large spans, carrying atmospheric pressure (100 kP or about 2100 psf). With an air supported structure, you would not need a frame at all. Instead of a vacuum, you would pressurize the interior space to a little more than atmospheric pressure.

Air would not be a suitable gas to use for the telescope mirror as it causes waves on the surface of the mercury. But, if filled with helium, would the waves be a problem?

Both suggestions: dome structure or hyperboloid structure are pretty near an optimum design.

To simplify the thing it would be advisable to have one large cylindrical section and two spherical end sections.

Have a visit to Moscow. There is a university dedicated to airplane research and development, the Moscow Aviation Institute.

They have a really big vacuum chamber designed to weld titanium aircraft, I saw it once but I forgot the dimensions, I estimate 10 x 8 x 4 m.

I am sure you can have a look and may be use it.

Workers wear scaphanders inside. As this chamber is not big enough to weld a complete big aircraft the individual sections are screwed or bolted. This concept seems to be better as cost for the chamber and cost for any repait can be hold low.

RHABE

Isn't this approximately the same problem as a suspension bridge. Use catenary shaped cables in tension to support (cables to support) the surface of the structure.

This is probably stupid, but, not sure what constraints you're working to: you might consider building on top of a high mountain, though you'd have to take higher wind speeds into consideration. Mount Whitney in California would give you a start of 0.532 atmospheres (mountains in Colorado and Washington come close); in Alaska you could do better, and, Kilimanjaro would be better still (and the weather is not too inclement).

"I had thought that the chambers could be made lightweight because the formulas for the thickness of a vacuum chamber would be almost the same as the formulas for the thickness of a pressure chamber (containing high pressure within.) The only difference I thought would be you would use the strength in compression of the material in the formula rather than the strength in tension"

This statement is not correct. Please find herewith the ASME code, Section VIII, Div. 1 formulas to calculate:

Thickness of Shells and Tubes Under External Pressure (UG-28).

Where,

A = Factor determined from Fig. G of Subpart 3 of Section II, Part D.

B = Factor determined from the applicable material chart, psi.

D_o = OD of cylindrical shell course or tube, in.

E = Modulus of elasticity of material at design temp., psi

L = Total length, in. (see Fig. UG-28.1).

P = External design pressure, psi.

P_a = Calculated value of max. allowable external working pressure for the assumed value of t, psi.

R_o = Outside radius of spherical shell, in.

t = Min. required thickness of cylindrical shell or tube or spherical shell, in.

t_s = Nominal thickness of cylindrical shell or tube, in.

(c) Cylindrical Shells and Tubes :

(1) Cylinders having D_o / t ³ 10)

Step 1. Assume a value for t and determine the value of L/ D_o and D_o /t.

Step 2. Enter Fig. G at the value of L/ D_o .

· For values of L/ D_o > 50, enter the chart at a value of L/ D_o = 50,

· For values of L/ D_o < 0.05, enter the chart at a value of L/ D_o = 0.05.

Step 3. Using the value of L/ D_o , move horizontally to the line for value of D_o /t. From this point of intersection, move vertically downward to determine factor A.

Step 4. Using A, enter the applicable material chart, move vertically to an intersection with the material/temp. line for the design temp. (see UG-20).

· If A falls to the right of the end of the curve, assume an intersection with the

horizontal projection of the upper end of the curve.

· For A falling to the left of the end of the curve, see Step 7.

Step 5. From intersection obtained in Step 4, move horizontally to the right and read the value of B.

Step 6. Calculate the max. allowable external working pressure, Pa = 4B/3(Do/t).

Step 7. For values of A falling to the left of curve, Pa = 2AE/3(Do/t).

Step 8. Compare the calculated P_a with P. Increase t until P_a ³ P.

(2) Cylinders having D_o / t < 10 :

Step 1. Using the same procedure as given in UG-28(c)(1), obtain the value of B.

· For values of D_o /t < 4, the value of A can be calculated using the following formula : A = 1.1 / (D_o / t)²

· For values of A > 0.10, use a value of 0.10.

Step 2. Using the value of B, calculate a value P_a1 using the following formula : P_a1 = [2.167/(D_o/t) - 0.0833] B

Step 3. Calculate a value of P_a2 using the following formula :

P_a2 = [2S/(D_o / t)][1 - 1/(D_o/t)]

where, S = Lesser of 2 times the max. allowable stress in tension at design metal temp., from the applicable table referenced in UG-23, or 0.9 times yield strength of the material at design temp.

Values of yield strength are obtained from the applicable external pressure

chart as follows :

(a) For a given temp. curve, determine the B value that corresponds to the

right hand side termination point of the curve.

(b) The yield strength is twice the B value obtained in (a) above.

Step 4. The smaller of P_a1 or P_a2 shall be used for the max. allowable external working pressure P_a . Compare P_a with P, and change t until P_a ³ P.

(d) Spherical Shells, seamless and butt welded :

Step 1. Assume a value for t and calculate A = 0.125 / (R_o / t).

Step 2. Using A, enter the applicable material chart. Move vertically to an intersection with the curve for the design temp. (see UG-20).

· If A falls to the right of the end of the curve, assume an intersection with the horizontal projection of the upper end of the curve.

· For A falling to the left of the end of the curve, see Step 5.

Step 3. From intersection obtained in Step 2, move horizontally to the right and read the value of B.

Step 4. Using the value of B, calculate P_a = B / (R_o / t).

Step 5. For values of A falling to the left of curve, the value of P_a can be calculated as : P_a = 0.0625E/(R_o/t)²

Step 6. Compare the calculated P_a with P. Increase t until P_a ³ P.

(f) Vessels intended for service under external working pressure of 15 psi or less, which are to be stamped with the Code symbol denoting compliance with the rules for external pressure, shall be designed for a max. allowable external working pressure of 15 psi or 25% more than the max. possible external pressure, whichever is smaller.

Hello RGClark,

In a perfect world, how large a box would you need (L x W x H)? Would interior columns be permitted? If this cannot be achieved, would a smaller size be an option?

Do you really expect to achieve a perfect vacuum inside the box, or can we assume a lesser value for design purposes?

Without some idea of these design parameters, it is not possible to come up with anything very helpful.

My 'expertise' in this field is at best minimal. However, what you describe sounds like it would be amenable to a reverse monocoque chassis as is used in racing automobiles.

Another thought, there have been some absolutely amazing topographical structures constructed that seem to defy classical geometry. I do not recall the name of the designer (he took Buckminster Fuller's concepts to a new level), but there may be some value to investigating this. It could render a framework over which a suitable gas-tight membrane could be suspended.

In a previous life I designed UHV, high temperature crystal growth reactors, the largest of which was 70" tall, 36" wide (178 cm x 91.5 cm roughly). It had a water jacket pressureized to about 50 PSI (gauge) (3.5 bar), and an ultimate vacuum of 7e-9 torr (9.3e-12 bar), but was also designed to have a positive pressure as high as 25 bar. The inner wall was roughly half a cm thick (0.25"), the water jacket passage was another 0.25", and the outer wall was 0.125" thick. the chamber was designed as being segmented, with thick flanges and a baseplate, each roughly 0.75" thick and O ring sealed. The top was domed, and there were protrusions all over the place for observation windows, feedthroughs, vacuum ports etc. Material was 316L stainless steel and the reactor had an operational temp of over 2500 degrees C. Given all that, I can tell you we never had any problems whatsoever. The design was modeled with COSMOS by me, and checked by our manufaturer (actually an industrial boiler maker, far cheaper then getting something custom built by a vacuum company, and we were even able to get the interior electropolished) using Autodesk inventor. We found that we could get away with even thinner walls if we used ribbing between the inner and outer walls, but for the sake of simplicity we decided not to go that route. Based on this "real world" example, you might be able to derive some general ideas for your own chamber.

Avery Montembeault

A tension argument would indeed be appropriate if acknowledgement were given to areas of greater tension due the type shape or construction. If construction be steel and shape be other than tubular or round special attention needed to ascertain specific stress points. Areas considered stress points need gussets, range of radius corner to spread stress or thickening of containment materials. Tensile vs. ductile vs fatigue hmm, round globe shaped vessel use your scale rule then begin gradual thickening starting 2/3's from top to create bottom thickness 1-1/2 times wall thickness. Chromium/ 7% molybdenum steel with 15%-21% tungsten or some such to extremely light weight high strength material. 521 series stainless steel, tough stuff and retains toughness at full brinell hardness.

Round is a shape requiring greater force to implode than explode, yes?

Your cylindrical walls with square ends is surprisingly close to that used to create force in early steam engines (: