Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Local Stresses in Vessels-Notes on the Application of plicabillty of the cylindrical vessel portion of the bulletin is
WRCI07 and WRC291 limited to'rigid inserts. The Bulletin does not recommend any
specific method in analyzing an actual nome connection in
the cylindrical vessel. It is left largely to the designers to make
their own judgment.
L.C. Peng1 WRC107 presents detalled tabular forms for calculating
stresses at four major axis locations. Stresses at both inside
and outside surfaces on these locations can be readily
The welding Research Council (WRC) Bulletin No. 107 [l] calculated following the' stepby-step procedure outlined in the
and BuJJetin No. 297 [2] are two or the most important design form_ The final results are the total skin stress imensities at
guides ever published for the design of pressure vessels. Wide- these four locatlons in the shell. No separate membrane stress
ly used in the design of attachments and ocule connections. intensity is given, nor is the stress in the nozzle calculated.
the,' have become indispensable tools in rccent years. WRC Bulletin No. 297. WRC297 was published in
WRC-I07 and WRC297 are invaluable due to their com~ August 1984 [2]. It is a supplement to WRC107 and is
bined consideration of theory, experimental data, and specifically applicable to cylindrical nozzles in cylindrical
engineering judgment. In some iostances the theoretical values vessels. This bulletin was based on Professor Steele's
have been adjusted several hundred percent upward to match theoretical work. It gives data for larger DIT ratios than in
the available experimental results. They are reliable tools WRC107, and also provides better readabiHtLfor small
without which the so-called "Design by Analysis" approach values of d/Dby plotting the curves using }..=dNDT as the
would have been impractical. Unfortunately many designers abscissa. Most importantly, the new theory considers the
have misapplied the data presented, thus resulting in inconsis- opening on the shell together with the restraining effect of the
tem designs. This micle describes the data available in these Do:u.le wall. This is a better model than the unperforated shell
two bulletins and explains the nature of inconsistency incurred used in WRC107 for simulating the no:u.le connections in
in some designs. A supplemental formula is then developed cylindrical vessels.
for calculating the combined maximum stress intensity to be Because most pressure vessel and piping codes have dir
used in designs.
ferent allowable stress criteria for different stress categories.
\YRe Bulletin No. 107. WRC107 was first published in WRC297 emphasizes the separation of membrane stress and
August 1965 [11. It was based on Professor Bijlaard's skin bending stress. Two examples are given outlining the
theoretical work, with some adjustments made based on detailed procedures for calculating membrane and total stress
available experimental data. A few revisions have been made intensities at points located in the longitudinal and the
since its first publication. The latest revision made in March transverse planes. Both shell and nome stresses are
1979 relabeled some "'of the curves. WRC1Q7 has been used calculated.
widely in the design of vessel nozzles and attachments. It was
one of the major driving forces in promoting the "Design by
Analysis" plulosophy. The bulletin covers both spherical and Location oJ Maximum Stress
cyiir.drical vessels. The fact that the examples given in both WRC-107 and
In sphericai vessels, the original theoretical work was based WRC297 specificalJy outline the procedures for calculating
on round rigid inserts and round nome connections. Square stresses at the four major axis corners, has led designers to
inserts or connections can be analyzed using an equivalent think that the maximum stress in the connection must be one
round attachment having a diameter equal to 8/7 of the at of those stresses. This presumption introduces inconsistency
tachment width. On the other hand. the original theoretical and nonconservatism in the design of nome and attachment
work for the cylindrical vessel was based on square and rec- connections. The maximum stress is not normany located at
tangular-shaped uniform loads acting on unperforated vessels. these corners. Although the calculation involves only the
Round attachments can be analyzed using an equivalent secondary stress which itself involves various uncertainties. a
square having the width of 7/8 of the attachment diameter.
Because of the assumption of the unperforated shell, the ap-
Contributed by lne Pre"ur. Veue\s and Piping Divi.:lon for publication In lb.
Jorn..>;.u OT PUS5lJtUl V<:S5EL TECHNOLOGY. Manus<;ripl received by lhe pyp
Dil'liion. Dl;ember 2t, 1987.
,J'tl,
"~;-> ~ , I ,
a
Fig. 1 Vessel aUllchmen15
m'
"'~)
55 = shear stress
.
The applied loading can generally be divided into four com-
I, ponent groups. They are radial force, bending mOUlent, tor-
i a) l bJ (el sional moment, and shear force. In the following discussion
Flg.2 Sire," on spheneal,hell the stress COntributed by each component group is described
first. They are then combined to become the maximum stress.
The discussion follows the WRC~297 stress orientation of
radial and circumferential directions with respect to the noz-
zle. Because of the different stress otientation adopted in the
certain amount of deviation is always expected. For instance, cylindrical shell ponion of the WRC-107, reorientation of the
as much as a lOpcrccnt difference may be made just from WRC-I07 data is required for the cylindrical vessel The pro-
reading the chart by different persons. H ... .vever, just because cedure shows the method for calculating the stress in the shell.
of its inherent uncertainty, effort is needed to make it as con- The same procedure can be used for calculating the stress in
sistent as possible. Anything that can be done to improve its the nozzle.
certainty should be done. In cases when deviation is
unavoidable, it is preferrcd to deviate on the conservative side. Stress due /0 P. In a spherical vessel, it is obvious that the
To demonstrate when an inconsistency may occur, the at- stresses created by the radial load are uniform around the eu-
tachment on a spherical shell can be used as an example. tire attachment circumfereuce. In a cylindrical vessel, though
Figure 2(a) shows a bending moment, M, acting on a nozzle the stresses differ from location to location, it can also be
connection at a spherical vessel. By choosing a coordinate regarded as uniform taking the maximum around the attach-
system as shown in Fig, 2(b). the stresses at the four major axis ment circumference as the unifonn value. TIlls assumption in~
points are SA = S. 5B"" S, SC "" O,and SD "" O. However. if a troduces some conservatism but is not overly conservative.
designer happens to have the coordinate system set up as WRCw297 has already adopted this approach in developing the
shown in Fig. 2(c), the applied moment will be decomposed in- design curves. In reference to the attachment orientation, the
to M) ""'M2 =O.707IM" two components. The stresses at the stress created by the radial load can be written symbolically as
four major axis poims, in this case, are SA=SB=
Membrane stress: S,m (Pl, S~m (P)
SC"'SD=O.7071S. This stress is about 40 pen:ent below the (1)
expected maximum stress. The same nozzle connection and Bending stress: Srb (P), S~b (P)
the same applied moment, yet the calculated stresses are They are constant around the entire attachment cir-
substantially different depending solely on how the coordinate cumference.
system is set up. From the foregoing demonstration. it is clear
that in Fig. 2(c), the maximum stress is not located at the ma~ Stress due to Me and M L Stresses created by Me and M L
jor axis points, but at the off-axis points P and N. In general, are not independent as assumed by some designers. In a
if the moments acting around both coordinate axes are spherical vessel the Me andML can be conveniently combined
nonzero, the maximum stress is not located at the major axes. as
This warning was properly stated in Paragraph 3.3.5, (2)
WRC-I07, which said, "However, in the general case of ar w