Rolled Plate analysis. 1

[nuche1973]

If you take a cylinder, capped at both ends, made from .25in 304SS, and cut it in half. Does ASME Section VIII criteria for external pressure still apply? The half cylinder would be fully welded to a flat plate. How does one analyze the stresses in the rolled plate? The plate is rolled to a 24in radius. The whole assembly is welded together then welded to a larger flat plate. The length of the cylinder is 36 inches.

I used Roark's, bent beam analysis, but since my R/d is much greater than 8, I am having trouble believing the results. I am trying to determine the amount of external pressure the half cylinder can withstand. I am also interested in the reactionary loads at the welds.

Any suggestions would be greatly appreciated. Thanks.

There are days when I wake up feeling like the dumbest man on the planet, then there are days when I confirm it.

 

[metengr]

nuche1973 \;
I am absolutely confused by your post above. Is the original cylinder a pressure vessel? Was it designed for internal pressure or external pressure? Are you sectioning along the meridional or circumferential direction?

 

[chicopee]

A sketch of what you want to do is worth a thousand words.

 

[MikeHalloran]

It's too short to not be supported by the ends, so they're involved too, but the first thing I'd try is evacuating it and waiting (at a distance) for the bang as it collapses.





Mike Halloran
Pembroke Pines, FL, USA

 

[MikeHalloran]

A full cylinder of perfect geometry and long enough for end effects to not interfere should collapse with an external pressure around 8.4 psi.

I have no idea how to deal with a short tube, or a short half tube.



Mike Halloran
Pembroke Pines, FL, USA

 

[nuche1973]

I generated a drawing. I believe that this will make it clear on what I am working on. thanks.

There are days when I wake up feeling like the dumbest man on the planet, then there are days when I confirm it.

 

[rb1957]

this looks like a domed end to a pressure vessel ... hoop stress = pR/t, "axial" stress (from one closed end to the other) maybe not the usual pR/2t. the hoop stress will put some load inot the fasteners along the sides, and bend the cr@p out of the attach flange.

at the ends, the pressure will be reacted by shear into the endplates (rather than hoops stress in the plate).

 

[racookpe1978]

The sketch is for rolled 12 GA, but the 1/4 thick plate I buy is 11 gage. Will that be a problem?

The 5/8 holes are to bolt down the flat 2" inch strips down each side?

The rolled plate isn't a half-cylinder since its drawn considerably "flatter" than that, but the load will be greatest at the sharp bend between the flat 2 inch strips and the rolled upper cover as the cover is resisting a force "blowing off" the cover. The wall itself will be strong enough for this light a pressure, but the cover might leak around the bolt holes.


I'd analyze the cover as if it were flat. That's conservative, but not too far off.

 

[MikeHalloran]

A full cylinder of perfect geometry, 24" R and long enough for end effects to not interfere, made of 12 gage stainless, .1046" thick, should collapse with an external pressure around 0.6 psi, not the 8.4 psi calculated for a 1/4" thickness.

The way the drawing calls out the 5/8" holes suggests that the material is perforated, but gives no details that would allow analysis.



Mike Halloran
Pembroke Pines, FL, USA

 

[rb1957]

generally perforated plate sucks at reacting pressure ...

 

[nuche1973]

Ok, not following the 6 P's on this one and it shows: Let me start over.

First, the original post is valid. I have what is essentially a half cylinder as stated and the original data is correct.

Second, the previous drawing, is wrong. Close, but useless for what I am trying to convey. Sorry about that. We see so many ambiguous questions and posts, there's no need to add any more.

Third, the attached drawing is what I am dealing with.

Finally, as stated before I analized this using Roark's, bent beam criteria. Since, it seemed applicable. However, after reviewing the results and thickness. I am wondering if I can use ASME Section VIII or basic pressure vessel theory. Which would help me generate a sound methodology.

I hope this clarifies what I am asking. Thanks.

There are days when I wake up feeling like the dumbest man on the planet, then there are days when I confirm it.

 

[MikeHalloran]

3.089" ??

Mike Halloran
Pembroke Pines, FL, USA

 

[nuche1973]

Mike:
Good catch. Sorry about that.

There are days when I wake up feeling like the dumbest man on the planet, then there are days when I confirm it.

 

[JStephen]

I would assume that if an entire cylinder is adequate to resist an external pressure, that a half-cyclinder fixed along the two straight edges would be stronger, rather than weaker. Actually proving that might be difficult. My thinking is that small buckles would occur equally in either case, but larger buckles would be prevented in the half-cylinder.

When analyzing a cylinder with internal pressure, the assumption is made that loading is uniform around the perimeter, which means there is no bending in the cylinder, just hoop tension. If you start putting attachments on the pressure side of the cylinder wall, that is no longer true, and you need to start looking at what kind of bending you are inducing.

 

[rb1957]

this is a small blister on a larger vaccuum tank ?

 

[rb1957]

no length shown on pic either ...

in any case, given Steel, 1/4" thk, R = 24", p = 13.7psi ...
pR/t = 1.3 ksi, conservative allowable stress (assuming 50" long) = 2.2 ksi (and that's conservative by a factor of something like 5) ... reference NACA TN 3783 (available online)

 

[nuche1973]

rb1957:
thanks for the reference. It will come in handy.

There are days when I wake up feeling like the dumbest man on the planet, then there are days when I confirm it.