UG-23(b) - Minimum (Sa, B) or Minimum (Sa*E, B) ?

[AWDMIKE]

In reviewing UG-23(b) after a recent discussion I overheard, I have found that different people have interpreted UG-23(b) in at least two different ways. I'm curious of what some of you think about this matter. The two different ways are as follows:

Minimum (Sa, B),
Minimum (Sa*E, B),

where

Sa is the allowable tensile stress,
B is the allowable longitudinal compressive stress,
E is the joint efficiency.

UG-23(b) states that the maximum allowable longitudinal compressive stress for cylinders shall be the lesser of (1) the maximum allowable tensile stress or (2) the value of factor B and that "The joint efficiency for butt welded joints shall be taken as unity." UG-23(a) covers the maximum allowable tensile stress value, but makes no mention of the joint efficiency.

I have found no interpretations or examples produced by ASME (including PTB-4-2013) that suggest that you should include the joint efficiency for tension in the determination in the lesser of the two ('Minimum' statement number two), however, the snippet in UG-23(b) mentions joint efficiency for compression and that itself could suggest that a joint efficiency (if any) should be accounted for in tension.

The issue arises when you take the lesser of Sa or B, and B governs, but when you take the lesser of Sa*E or B, and Sa*E governs.

 

[SnTMan]

AWDMIKE, UG-23(b) is limited to seamless or butt-weld construction. It just says that either may use joint efficiency of 1. I take it at its face, E = 1.0 for butt joints.

I suppose you could have a lap welded cylinder for which you wanted to find allowable compressive stress per UG-23(b) methods. In that case I would use E per Table UW-12.

Regards,

Mike