-
1
- #1
TomBarsh
Structural
- Jun 20, 2002
- 1,003
COMPRESS provides an option labeled "Junctures act as lines of support" (see Set Mode Options dialog/Calculation tab). See Figure UG-28.1 for description of vessel configurations where cone-cylinder junctions are or are not treated as a line of support. What does this option accomplish for the design?
When the option is checked, cone-cylinder junctions will be treated as a line of support for external pressure. Normally, the junction must be reinforced to provide sufficient reinforcing area as per UG-33(f) and Appendix 1-8. But when the junction also acts as a line of support for external pressure these paragraphs require that the cone-cylinder junction must meet the stiffness requirements (moment of inertia) of Appendix 1-8. If the junction does not act as a line of support then ASME specifies additional requirements for minimum nominal thickness of the cone, knuckle, or toriconical section; see Figure UG-28.1, Note (1).
(Note that even if there is a knuckle at the junction the stiffness requirements must be met even though the reinforcing area requirements do not apply; see UG-33(g).)
The requirement for stiffness may dictate that the cone or cylinder thickness must be increased, or it may mean that a separate reinforcing ring be provided. Any of these options will increase cost. On the other hand, if the junction is not a line of support then the requirement of Figure UG-28.1, Note (1) may result in an increase in the thickness of the cone as well.
The particulars of this issue were first presented to me a number of years ago when a designer found that when the COMPRESS option was applied (junction was a line of support) then a stiffening ring was required, but when the option was not applied then no ring was required and the existing material thicknesses were sufficient as well; very surprising!
There was a difference in the analysis: when the option was active then the MAEP of the cone was greater than otherwise. However, the MAEP in that case was much higher than that of the other components and did not provide any benefit to the MAEP of the vessel. Thus there was no real benefit to specifying that the junction was to act as a line of support, and in fact it imposed the additional cost of providing the stiffening ring. This vessel on saddles was relatively small and the designers were using something like 1/4" plate across the board for practicality, consequently they were not penalized by the Note (1) requirements.
The second time I ran into this condition was for a very tall tower of relatively small diameter. Possibly around 120 feet tall with 4 foot diameter. The small diameter permitted the designers to omit vacuum rings without using excessive shell thickness (the thickness required by internal pressure was sufficient to handle the case of full vacuum). A transition to a larger diameter existed at the very bottom of the vessel. Again, the designers questioned why there was no change in plate thicknesses between the two cases but that a stiffener ring was required when the junction was a line of support.
This second example sheds some more light on the use of this "feature" of the ASME Code. In this design the effective length of the smaller diameter shell was little changed between the two cases. In one case Le ran from the top head to the upper cone-cylinder junction; in the second case it extended to the line at the bottom head and included the additional length of the transition and the short bottom cylinder. The length in the first case was so long already that the additional length to the bottom head was not sufficient to materially change the nominal thickness that was necessary. But by permitting (or requiring) the junctions to act as lines of support the upper junction was required to have a reinforcing ring. Again, there was no benefit to this design to permitting the junctions to act as lines of support.
Can we draw some general conclusions out of these examples? Perhaps.
In both examples the material thickness was governed by criteria other than external pressure (thinnest value for practical handling in first example, internal pressure in second example). The second example was such that there was little effective difference in the unsupported length of the long cylindrical shell section. Both of these principles might be considered when designing a vessel for vacuum. But if the transition exists at some intermediate elevation (unlike the second example) then the difference in unsupported length would be substantial (ie: the maximum difference would be by a factor of around 2.0) and there may then be some economy in permitting the junction to act as a line of support.
"Although this forum is monitored by Codeware it is not intended as a venue for technical support and should not be used as the primary means of technical support."
Tom Barsh
Codeware Technical Support
When the option is checked, cone-cylinder junctions will be treated as a line of support for external pressure. Normally, the junction must be reinforced to provide sufficient reinforcing area as per UG-33(f) and Appendix 1-8. But when the junction also acts as a line of support for external pressure these paragraphs require that the cone-cylinder junction must meet the stiffness requirements (moment of inertia) of Appendix 1-8. If the junction does not act as a line of support then ASME specifies additional requirements for minimum nominal thickness of the cone, knuckle, or toriconical section; see Figure UG-28.1, Note (1).
(Note that even if there is a knuckle at the junction the stiffness requirements must be met even though the reinforcing area requirements do not apply; see UG-33(g).)
The requirement for stiffness may dictate that the cone or cylinder thickness must be increased, or it may mean that a separate reinforcing ring be provided. Any of these options will increase cost. On the other hand, if the junction is not a line of support then the requirement of Figure UG-28.1, Note (1) may result in an increase in the thickness of the cone as well.
The particulars of this issue were first presented to me a number of years ago when a designer found that when the COMPRESS option was applied (junction was a line of support) then a stiffening ring was required, but when the option was not applied then no ring was required and the existing material thicknesses were sufficient as well; very surprising!
There was a difference in the analysis: when the option was active then the MAEP of the cone was greater than otherwise. However, the MAEP in that case was much higher than that of the other components and did not provide any benefit to the MAEP of the vessel. Thus there was no real benefit to specifying that the junction was to act as a line of support, and in fact it imposed the additional cost of providing the stiffening ring. This vessel on saddles was relatively small and the designers were using something like 1/4" plate across the board for practicality, consequently they were not penalized by the Note (1) requirements.
The second time I ran into this condition was for a very tall tower of relatively small diameter. Possibly around 120 feet tall with 4 foot diameter. The small diameter permitted the designers to omit vacuum rings without using excessive shell thickness (the thickness required by internal pressure was sufficient to handle the case of full vacuum). A transition to a larger diameter existed at the very bottom of the vessel. Again, the designers questioned why there was no change in plate thicknesses between the two cases but that a stiffener ring was required when the junction was a line of support.
This second example sheds some more light on the use of this "feature" of the ASME Code. In this design the effective length of the smaller diameter shell was little changed between the two cases. In one case Le ran from the top head to the upper cone-cylinder junction; in the second case it extended to the line at the bottom head and included the additional length of the transition and the short bottom cylinder. The length in the first case was so long already that the additional length to the bottom head was not sufficient to materially change the nominal thickness that was necessary. But by permitting (or requiring) the junctions to act as lines of support the upper junction was required to have a reinforcing ring. Again, there was no benefit to this design to permitting the junctions to act as lines of support.
Can we draw some general conclusions out of these examples? Perhaps.
In both examples the material thickness was governed by criteria other than external pressure (thinnest value for practical handling in first example, internal pressure in second example). The second example was such that there was little effective difference in the unsupported length of the long cylindrical shell section. Both of these principles might be considered when designing a vessel for vacuum. But if the transition exists at some intermediate elevation (unlike the second example) then the difference in unsupported length would be substantial (ie: the maximum difference would be by a factor of around 2.0) and there may then be some economy in permitting the junction to act as a line of support.
"Although this forum is monitored by Codeware it is not intended as a venue for technical support and should not be used as the primary means of technical support."
Tom Barsh
Codeware Technical Support