Min. Thickness of exchanger tube 1

[NovaStark]

Hi all,

I am trying to calculate the minimum thickness of an exchanger tube under internal pressure. For this I am using ASME Section VIII Div 1. I am using an already designed piece of equipment to ensure that I am using the correct formulas. However I am bit confused as the ASME code is giving me a slightly different formula than the design calculation from the equipment.

I have uploaded the design calculation (document.pdf)

I redid the calculation using mathcad but this gives me a different min thickness than the design one (query.pdf)

And finally I did the calculation using the asme formula (asme.pdf)


I am not sure which formula to use here or if I am using the wrong code.

 

[TJOrlowski]

I would expect the result in your asme.pdf to be the most valid result. For cylinders under internal pressure, ASME Section VIII, Div 1 requires the designer to consider both the longitudinal (circ joint), and the circumferential stress (long joint), and use the thicker of the two for required thickness.

Review UG-27(c)(1) and UG-27(c)(2), and see alternative formulae in Appendix 1-1. I'd expect your required thickness to be something less than 2mm, even if you use a value <1 for E.

-TJ Orlowski

 

[NovaStark]

I'll check the other formulas then. But I don't get where the design calculation (document pdf) got the 20*SE thing from. That is the confusing part to me.

 

[chicopee]

In your last PDF,it seems that you are mixing english and metric units in the formula. Make sure the units are consistent with the formulae inyour three PDF files.

 

[EdStainless]

And what is the variation in allowed stress?

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[NovaStark]

I don't follow what you mean by variation in stress. Isn't the component not supposed to exceed the allowable stress as given in the ASME section 2?


Also once typed in MathCad, it will convert it to one common set of units (SI in this case) and then convert giving the output in SI. So even if I type it in imperial units, it should give me the same output whether I have all in SI units or whatever units.

 

[IRstuff]

While your Mathcad is using consistent units, your "document.pdf" is not. It's mixing bar and N/mm². The 20 appears to correct for putting S numerically in N/mm². This is shown be solving for the denominator, resulting in 2044 bar for the denominator, in which the 0.8*P term is nearly negligible. Therefore 20*S*E = 1931 bar, but S is actually 965.2 bar. Therefore, the equation is requiring S to be numerically in N/mm², with the 20 doing the 10x conversion to bar.

Old equations have to extremely carefully examined to determine whether there were shortcuts taken for empirical reasons or for convenience. Why it is 40 is not clear, though.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[NovaStark]

While your Mathcad is using consistent units, your "document.pdf" is not. It's mixing bar and N/mm2. The 20 appears to correct for putting S numerically in N/mm2. This is shown be solving for the denominator, resulting in 2044 bar for the denominator, in which the 0.8*P term is nearly negligible. Therefore 20*S*E = 1931 bar, but S is actually 965.2 bar. Therefore, the equation is requiring S to be numerically in N/mm2, with the 20 doing the 10x conversion to bar.

Old equations have to extremely carefully examined to determine whether there were shortcuts taken for empirical reasons or for convenience. Why it is 40 is not clear, though.

Right well document.pdf is from the vendor's documentation where my confusion lies, but I understand now, all the values are meant to be put in as is not to be converted to SI and then input.