Bulkhead thickness query

[Icosahedron]

I need an engineering formula for deckplan bulkhead design:

How thick does a flat steel plate of a given size need to be to withstand vacuum?
How does this vary if the plate is curved?

Just trying to get a feel for bulkhead and airlock thicknesses.

Thanks in advance.

 

[sabredog]

I'm no engineer, but I think it wouldn't have to be all that thick. Unlike a submarine you are keeping the pressure in (and it's not that much pressure) as opposed to enormous amounts of pressure out.

Think of an airplane - even the ones like the SR-71 are just made of really thin sheets of pre-stressed aluminum or titanium.

I think the reason bulkheads in Traveller are considered to to be "stronger" has more to do with what's inside them - pipes, stringers, conduits, and all that as opposed to the thinner (I imagine cubicle thickness) interior walls.

Being curved helps when the pressure is coming from outside because the pressure impacts on the object from all directions equally, but not sure if the reverse is true.

 

[CosmicGamer]

Well, the smart alikee answer would be that something as thin as tinfoil can 'withstand vacuum'. I believe it all depends on the pressure difference.

I can tell you that under one atmosphere of external pressure
- a soda can certainly can't hold a vacuum
- I believe a glass beer bottle can hold a vacuum
- I know a food can will hold at least a partial vacuum
- one atmosphere is 14.7 pounds per square inch
- those portable air tanks for inflating tires can hold well over 40psi and I don't think they are very thick

An interesting fact: Some boxers can hit you with over 1500psi!

A quick search provided:
"one inch of steel plate will yield beyond its ability to recover its original shape at approximately 36k psi, and will fail at approximately 60k psi. "

So I think the thickness of a bulkhead may be fairly negligible compared to the life support plumbing, electrical wires and other items that may need to fit inside.

 

[aramis]

Keep in mind: the aluminum walls of the lunar landers were able to be punched through with a fist... or cut with school scissors... but held against 8+ PSI.

A soda can can hold in 2-3 ATM; it can't hold out the same, however, since the tensile strength is significantly different from the flexation strength.

 

[Icosahedron]

Thanks for the replies. I've seen the drinks cans in action and I did some quick calculations / thought experiments before I posted the question.

Thin material is ok in small areas, particularly if curved, but working on 15psi, that's about 10 metric tons per sq metre, so 20 tons pressing on an airlock hatch. That's like parking ten land rovers in your ship's corridor and then sitting your ship tail-down in 1G, so the vehicles rest on the airlock...
I reckon you're not gonna hold those babies back with tinfoil.

And a whole bulkhead, say 10 metres wide by 3 high, might feel 10x3x10 = three hundred metric tons of force. The strength and thickness would have to increase as the area increases.

Nevertheless, I think you're right that bulkhead thicknesses will be more about inertial stresses and structural rigidity than about vacuum integrity. I just wish I could find some comprehensible figures to prove it.

 

[atpollard]

With a high pressure inside and a lower pressure outside, the minimum thickness (for strength) would be minimal. The Space shuttle and aircraft both have very thin skins.

Look up pressure containers on wiki ... you will discover that the mass of the container scales directly with its internal volume, so a 100 dTon pressure vessel will weigh exactly 10 times as mich as a 10 dT pressure vessel and 100 times as mich as a 1 dT pressure vessel.

Imagine a 1 meter cube whose shell is made of stainless steel plate with a thickness of 0.001 meters (1 micrometer) and a density of 8000 kg per cu.meter. The area of the shell is 6 sq.meters, the volume of the shell will be 0.006 cubic meters (6 x 0.001) and the mass of the shell will be 48 kg (0.006 x 8000).

Now imagine a 10 meter cube (10x10x10=1000 cu meters) or about 71-74 dTons (based on 13.5 or 14 cu.m. per dT). Since the weight of the shell scales with its volume, the 10 meter cube will have both a volume and weight of 1000 times the 1 meter cube, so the shell of the 10m. cube will weigh 48,000 kg (1000 x 48). The surface area will increase with the square of its length, so the 10 m cube will have 100 times the surface area of the 1 m. cube (10x 10=100), or 600 square meters of steel plate. Dividing the mass of the shell (48,000 kg) by the area of the shell (600 sq. meters) yields a shell that weighs 80 kg per square meter (48000/600=80). Dividing the mass of the shell (80 kg/sq.m.) by the density of stainles steel (8000 kg/cu.m.) yields the plate thickness in meters (80/8000=0.01) or 0.01 meters (1 centimeter or 10 milimeters) for the 10 m. cube.

As far as the actual thickness of the shell, it will increase with the cubed root of the volume. Increasing the volume by 1000 times will increase the thickness of the pressure container by 10 times (1000^0.3333=10). Doubling the volume will increase the shell thickness by x1.25 (2^0.3333=1.25).

Structurally, the shell can be quite thin. The problem in general becomes one of radiation and micro-meteorite impact protection requiring a much greater 'minimum thickness' to protect the occupants. For example, a 747 or the Space Shuttle would probably not explode from the pressure of travelling in deep space, but solar radiation would kill the crew and pebbles would punch lots of holes in the hull.

 

[HG_B]

Also, remember that you aren't dealing with steal. You are dealing with materials that, hypothetically, can withstand kinetic energy many thousands of times higher than anything currently known. Based on that, it can be VERY thin.

 

[BraselC5048]

I've spent some time caculating the weight and thickness of structure for starships. I decided that a pressure bulkhead would be able to hold 1.5 atmospheres of pressure. Assuming everything is made of out titatium (grade 5 alloy),the walls are flat, and all the stresses are below the fatauge limit, results in pressure bulkheads being 1.5 cm thick. But, they need to be supported on each side every 5 feet by box beams about 6.5 inches tall and 4 inches wide, with one of the ship's major structural members every 10 feet. Since in my universe removable 1/16" thick pannels go outside the box beams, to cover up the piping and whatnot in the wall, that would be a total wall thickness of about 14".
On the other hand, I had a hard time coming up with information of flat walls, since just about all pressure vessels in the real world are round, and apparently one caculation says a round wall of 10 foot radius would be a tenth of the thickness, without the beams. Atpollard, could you tell me where you got the weight and thickness for that cube? I could really use some more information.

 

[sabredog]

Except for possibly a close structure (or box in other versions of Trav) most designs will have mainly round outer hulls. The interior walls, bulkheads, and everything else is just built inside the pressure hull - which isn't really so much like a pressure hull any more than on a high-flying airplane.

High Guard has always said that "armor" for a ship (in CT) isn't just thicker outer skin - it also consists of interior bracing and greater compartmentalization. Plus, as another poster pointed out: crystaliron, superdense, and bonded superdense are the materials of the future so those could be as little as (assuming no armor added) a cm.

Airlocks could always be beefed up a but "just in case", but I think mass would be at a premium so unless it's armored a ship would be as light as possible.

 

[atpollard]

I've spent some time caculating the weight and thickness of structure for starships. I decided that a pressure bulkhead would be able to hold 1.5 atmospheres of pressure. Assuming everything is made of out titatium (grade 5 alloy),the walls are flat, and all the stresses are below the fatauge limit, results in pressure bulkheads being 1.5 cm thick. But, they need to be supported on each side every 5 feet by box beams about 6.5 inches tall and 4 inches wide, with one of the ship's major structural members every 10 feet. Since in my universe removable 1/16" thick pannels go outside the box beams, to cover up the piping and whatnot in the wall, that would be a total wall thickness of about 14".
On the other hand, I had a hard time coming up with information of flat walls, since just about all pressure vessels in the real world are round, and apparently one caculation says a round wall of 10 foot radius would be a tenth of the thickness, without the beams. Atpollard, could you tell me where you got the weight and thickness for that cube? I could really use some more information.

I just selected an arbitrary (small) initial thickness for the 1 meter cube. I was focusing on how the size of the ship will affect the weight and thickness of the shell.

The actual thickness varies with the tensile strength of the material, the pressure and the shape of the 'box'. I suggest that the steel casing for the SRB motors on the space shuttle would be a good place to start as a 'typical' thickness.

 

[Supplement Four]

I believe there are acutal thicknesses listed in the game. I may be thinking of just the hull, though, not a bulkhead.

Many of the design systems get pretty detailed about that--especially MT, TNE, and T4.

I don't think it had acutal thickness, but some of the combat info in Snapshot may be helpful. There's some CT rules there for getting through a bulkhead.

 

[PFVA63]

Hi

One thing that you want to be careful of in any type of back of the envelope type anayses like these is to not focus in too much on one factor at the expense of others.

For instance in the design of ocean going ships, while hydrostatic pressure against a hull is one issue, you must also consider other loads like global bending loads, thermal loads, and the like. Also, concerns about erosion/corrossion etc also have to be considered. Thus, even if calcs suggest that you need only a relatively thin amount of material to overcome the pressure differential, you may actually need for the plating to be thicker for these other reasons.

Typically in ships, and I think aircrat and stuff like the space shuttle, the vessel typically has stiffened shell plating, similar to the post from BraselC5048 above.

Here are some small sketches of the space shuttle, a jet aircraft, a ship, and a submarine for reference.

http://history.nasa.gov/SP-4225/imagery/diagrams/shuttle/3-7_m.jpg

Regards

PF

 

[Andrew M]

One thing that you want to be careful of in any type of back of the envelope type anayses like these is to not focus in too much on one factor at the expense of others.

He asked one simple question...

..but sure run with it, we're all here to learn +/- waste time

 

[PFVA63]

Hi

Hi,

Sorry that this has taken so long to respond, but it took me awhile to find a book that I was looking for.

In the book "Synthesis of Subsonic Airplane Design" by Egbert Torenbeek there are some equations for estimating the weight of the shell of an aircraft base in part on the differential pressure that the aircraft shell will withstand.

Specifically the book suggests that

Wsk = constant x delta P x Df x Sg x fref/f

In US units

Wsk is the skin weight to resist design cabin diffential in lb
constant = 0.007
deltaP is the cabin pressure differential in psi
Df is the diameter of the main fuselage lobe in ft?
Sg is the gross shell area (with all openings faired over and no blisters, etc) in sq ft
fref is reference mean hoop stress level = 12,000 psi
f is anticipated mean hoop stress level in psi
(if f is unknown the book suggests assuming fref/f = 1)

So if you only want a fuselage to be able to have a standardpressure inside, with a vacuum outside, then I'd guess that

Wsk/Sg = 0.1029 * Df

If we assume a standard density for aluminum of 0.098 lb/cubic inch (or 14.1 lb per inch-sq ft) then you have a minimum thickness of about 0.073 inch (1.85mm) for a 10ft diameter hull or a 0.729 inch thick shell (18.5mm) at 100ft diameter.

This does not include the weight of stiffeners, frames, or other such stuff, and doesn't really consider the effects of stresses from thermal loads (such as part of the ship in sunlight and part in shadows or heating due to re-entry), impact loads, or re-entry drag (especially where part of the vessel may be subjected to higher resistance than other parts leading to additional stresses).

There are also some additional equations for the weight of floorng or bulkheads at the ends of carge spaces, etc (if you are interested)

Anyway, just wanted to pass along this additional info.

Regards

PF

 

[Icosahedron]

Thanks PF - and everyone else - that formula shows how the different factors relate to each other and should be useful.

I think the example of under one inch of aluminium on a 100ft shell demonstrates that with a hull of crystaliron or superdense material, the actual material thickness will be negligible compared with the struts necessary to support its own weight/inertia (as everyone else asserted). Basically, if it can hold itself up, it will keep a vacuum at bay.

For my purposes, then, bulkhead walls are simply a thin skin that encloses the structural support members. The thickness of the panelling on the bulkheads does not have to be greater than that on cabin walls, the inner and outer skins will simply have a greater seperation thanks to the larger supports between them.

However, this also means that bulkheads should be no more weapons resistant that cabins - unless the skin is designed as a structural component and is made of a stronger material... All a matter of choice rather than necessity, perhaps.

Interesting discussion.

 

[Matt123]

FWIW, I was randomly looking at collected links & came across this essay.
http://www.io.com/~thrash/hulls.html

Might be useful, might not.

 

[HG_B]

FWIW, I was randomly looking at collected links & came across this essay.
http://www.io.com/~thrash/hulls.html

Might be useful, might not.

Very interesting write-up. It doesn't follow canon though.

 

[jawillroy]

Very interesting write-up. It doesn't follow canon though.

On the other hand, if you're like me and are less interested in OTU canon than the implications of the original rule-set, it's a lovely rationale for LBB2 ship construction!

 

[HG_B]

On the other hand, if you're like me and are less interested in OTU canon than the implications of the original rule-set, it's a lovely rationale for LBB2 ship construction!

As far as I know the limitations imposed in the doc listed don't follow B2 design rules re limits on M-Drives, etc. But, it is a good starting point for an alternate rules system, that's for sure. Lots of good data.

 

[jawillroy]

As far as I know the limitations imposed in the doc listed don't follow B2 design rules re limits on M-Drives, etc. But, it is a good starting point for an alternate rules system, that's for sure. Lots of good data.

Oh, I'm sure it's not dead-on, but I haven't my LBB with me to check since I oughtta be working... but I do like the basic rationale for keeping the ships small, on average.

(with allowances for superhightech gee-golly-whiz ships of enormosity, but I don't really need design rules for those so much...)