Question: Thin Walled Cylindrical Vessels

[CVesGuy]

Question:

I work in a small sheet metal shop. We make cylindrical vessels (up to 4 ft in radius) that can hold up to a few psi.


The usual formula I use - which was given to me by the chief engineer - for wall thickness is PR/t=stress (Hoop stress) or t=PR/Allowable stress at temperature.

I then throw in about 40-60% extra thickness for a margin of error.

Do I have to add in axial stress (PR/2t) [which is perpendicular] and/or shear?

My understanding is that since hoop stress is the largest, it alone is sufficient.

This is not an ASME shop.

My understanding is that you do not add the hoop, axial, and shear stresses, but but go for the maximum, which in our case is hoop stress.

 

[Lumber Jim]

Question:

I work in a small sheet metal shop. We make cylindrical vessels (up to 4 ft in radius) that can hold up to a few psi.


The usual formula I use - which was given to me by the chief engineer - for wall thickness is PR/t=stress (Hoop stress) or t=PR/Allowable stress at temperature.

I then throw in about 40-60% extra thickness for a margin of error.

Do I have to add in axial stress (PR/2t) [which is perpendicular] and/or shear?

My understanding is that since hoop stress is the largest, it alone is sufficient.

This is not an ASME shop.

My understanding is that you do not add the hoop, axial, and shear stresses, but but go for the maximum, which in our case is hoop stress.

I think you and your chief engineer should find the appropriate industrial standard for the application and use it after you get comfortable and confident with it or find someone like me that provides those services.

I wouldn't expect more detailed guidance here than that based on what you have described but maybe I'm wrong...

Just because you are not an ASME shop doesn't mean you can't use ASME.

"Designed in accordance with..."
Or
"Certified to..."

Insert the appropriate design code or standards after that.

I've seen plenty of low pressure thin walled vessels blow the roof off service shops and design ALWAYS comes into question.

CYA

 

[Tiger15]

I agree with what @Lumber Jim said above. I'd encourage you to work with your chief engineer to determine what calcs need to be done and the appropriate thickness, and also to determine if you want to follow any codes (or just use them as a guide) even though you're not an ASME shop.

I don't have enough information to give you a full answer to your question, but I'll share some information with you. Just some stuff to think about.

• Normally stresses should be combined (not directly added; the formula is more complex than that) into what's called a "maximum and minimum principal stress" which is a stress, at some angle, that is higher than the stress in the basic directions. And also combined into what's called a "maximum shear stress" which again is the stress at a different angle where shear stress is the highest. Max prin and max shear are compared to two different allowable stresses (i.e. the shear allowable is different than the tensile allowable).
  • If you have axial and hoop stress ONLY, in a thin-walled pressure vessel, that is a special case called a bi-axial stress state where the max principal stress should equal the hoop stress. But it may still be important to check max shear.
• What allowable stress are you using?
  • It should be noted that the allowable stress at the welds and their "heat-affected zone" is much lower than the allowable stress of the unwelded sheet metal. This is where codes help you a lot because those weld calcs can get quite complex.
  • Also the allowables change at different temperatures. What temperature do your pressure vessels operate at?
• Are you considering any other loads besides internal pressure? Self weight (i.e. dead loads)? Liquid weight? Walking loads? Handing loads? Environmental load (wind, thermal, etc)? I'm sure there's more that I could come up with if I put more time/thought into it.
• Are your pressure vessels allowed to go into vacuum? Do you publish any vacuum capability? The calcs for a thin-walled pressure vessel buckling under vacuum can also be pretty complex.
• Is there any consideration for fatigue? Fatigue is a mode of failure where something breaks at a stress below the allowable stress after a number of cycles. You can think about it like this: even though you stay below the allowable stress, if you stress and unstress the pressure vessel enough times eventually it will wear out and fail. Again a different calculation to consider.
• Do your pressure vessels experience corrosion? Corrosion will makes your pressure vessels loose material (i.e. get thinner) over time. Should that be considered in the design? ... in the life estimates?

This isn't an exhaustive list of everything to consider, just what immediately came to mind. I'm sure there's some stuff I'm forgetting.

And I'm not saying that have to address all of these things. Since it's not an ASME shop, you are free to decide what you consider in your design. Maybe your chief engineer can use engineering judgement and say that most of this stuff doesn't matter in your specific application. That's fine, but realize that you all may be proceeding at risk if you were to ignore something important. If a customer experiences an expensive failure or worse, someone gets hurt, then your company and your engineer may be liable.

Sorry there are no easy answers. I hope this is helpful.

 

[Platypus Engineer]

I think there is an ASME standard for design of thin walled vessels. I am assuming vessel holds some type of gas since you say its low pressure. Stresses to consider also depend on if it is mounted horizontally or vertically.

Another organization to check for standards is API (American Petroleum Institute) www.api.org.

 

[Lumber Jim]

I think there is an ASME standard for design of thin walled vessels. I am assuming vessel holds some type of gas since you say its low pressure. Stresses to consider also depend on if it is mounted horizontally or vertically.

Another organization to check for standards is API (American Petroleum Institute) www.api.org.

There is. DOT also references ASME but exempts certain sections for low pressure.

Thumbs up to API too. API 650 is an example.

There are many references out there so you don't have to feel like you are on your own...

Many past editions are free if you search for them...