Why Don't The Required Thickness Don't Match?
2

Question

Why Don't The Required Thickness Don't Match?

Christine74 (Mechanical)

(OP)

5 Aug 04 12:43

For vessels with a specified corrosion allowance, when I run a calculation for the minimum required thickness of a cylinder based on the I.D. formula per UG-27(c)(1), I get one result.

However, when I run a calculation for the minimum required thickness of a cylinder based on the O.D. formula per Appendix 1-1(a)(1), I get a slightly different result, even though all of the geometry in both cases is exactly the same.

Which result is correct?

What am I doing wrong?

Thanks.

Christine74 (Mechanical) · Answer 1 · 2004-08-05 12:46:40

Sorry, that title should read "Why Don't The Required Thicknesses Match?"

Is there any way to edit our posts here?

deanc (Specifier/Regulator) · Answer 2 · 2004-08-05 13:16:50

Do the same as when you do a "layout". Use the centerline or neutral for thickness of the plate/part and adjust. ODs and IDs do not work well in real life. You will never go wrong on a centerline.

JoeTank (Structural) · Answer 3 · 2004-08-05 14:34:15

Christine74,
You may not be doing anything wrong. Both equations give essentiall same answer, but they are not identical. Remember also that ASME VIII requires that all calcs be made in the fully corroded condition. Therefore, you must deduct a full corrosion allowance when performing the Code calcs. In your case the inside radius INCREASES by the amount of the corrosion allowance. The outside radius remains unchanged. I am assuming that the corrosion allowance is for internal corrosion purposes.

Steve Braune
Tank Industry Consultants
www.tankindustry.com

codeeng (Petroleum) · Answer 4 · 2004-08-09 15:49:58

It would be unrealistic to give a formula that would be exact for both o/d and i/d cylinders. Generally its better to work with i/d but you may want to use o/d for pipe, since pipe outside diameters are fixed. If using pipe you should make an allowance for mill under tolerance, usually 12 1/2% of nominal wall thickness, then deduct any corrosion allowance.

metalonis (Mechanical) · Answer 5 · 2004-08-10 09:37:33

Sorry guys, I'm sure that these two formulas are the same, one written in terms of the ID and one in terms of the OD. Any differences might be the result of rounding. Otherwise I don't see how the results could be different.

Christine74 (Mechanical) · Answer 6 · 2004-08-10 14:19:20

metalonis,

The formula for the required thickness in terms of the ID is

T = PRi/(SE-0.6P)

The formula for the required thickness in terms of the OD is

T = PRo/(SE+0.4P)

These two formulas give different results unless T = Treq (or P = MAWP). This difference is NOT due to rounding errors.

prex (Structural) · Answer 7 · 2004-08-11 03:31:18

Well, of course Christine74, the two formulae may give consistent results only if you use consistent dimensions, that is R_o=R_i+T

prex

http://www.xcalcs.com
Online tools for structural design

metalonis (Mechanical) · Answer 8 · 2004-08-11 08:30:50

Exactly, prex.

If Ro does not equal Ri+t, then the two equations HAVE to give DIFFERENT results.

Christine74 (Mechanical) · Answer 9 · 2004-08-11 11:39:48

Since those formulas are solved for the required thickness, you can't just plug in a value for t in one part of the equation and then leave t as a variable in another part of the equation. You could try substituting Ro = Ri + Treq, but that obviously isn't correct, just a rough approximation.

The result is that you end up with inconsistent results from the two formulas.

IRstuff (Aerospace) · Answer 10 · 2004-08-11 17:33:29

Have you tried to replace Ro with Ri+T in the second equation?

I get the first equation back, so the equations appear to be very consistent.

TTFN

codeeng (Petroleum) · Answer 11 · 2004-08-11 18:19:42

This is quite the debate however you can't go changing anything. You have to use one or the other and there will always be that discrepancy. The formula for OD in the Appendix is given for convenience and is sufficiently accurate when considering the factor of safety in the allowable stress. There are other Code requirements based on rules of thumb or experience which could be more questionable.

JoeTank (Structural) · Answer 12 · 2004-08-16 11:07:50

Christine74,
Just for the record, what are the thickness requirements when calculated both ways. I suspect that both answers are very close.

Steve Braune
Tank Industry Consultants
www.tankindustry.com

Christine74 (Mechanical) · Answer 13 · 2004-08-16 11:51:54

Steve,

Yes, for most of the cylinders I've calculated, the required thicknesses are VERY close, within a few thousandths of an inch.

However, in running nozzle reinforcement calculations, it is not uncommon to use a nozzle that is significantly thicker than the required thickness for internal pressure (as per UG-45). With a large difference in the required vs. actual thickness, the results of these two formulas diverge enough to affect the MAWP for the reinforcement calculations by a few psi.

I plan to stick with the more accurate I.D. formula from now on.

-Christine

prex (Structural) · Answer 14 · 2004-08-16 13:08:13

I agree, Christine74, as both formulae are allowed, there is no reason for not using the more economical one, or the R_i based formula.
However it is not a matter of accuracy. The point is that pressure is of course assumed to act at R_i in both, but in one R_i is given, in the other one must be inferred in a safe way. And as the formula doesn't know the actual thickness, it is inevitable to assume pressure is acting at R_o-T (if you inspect closely the formula, this would be in fact R_o-0.4T, but that's another story).
Anyway I agree with you: the R_o based formula is useless.

prex

http://www.xcalcs.com
Online tools for structural design

Come Join Us!

Posting Guidelines

Why Don't The Required Thickness Don't Match?
2