(2) bending stress in the central portion of a flat head shall be considered only from the fatigue standpoint

point and
due to pressure are therefore classified as peak stresses in Fig. KD-230.
(n) Secondary Stress. A secondary stress is a normal Examples of local thermal stress are:
stress or a shear stress developed by the constraint of (a) the stress in a small hot spot in a vessel wall
adjacent parts or by self-constraint of a structure. The basic (b) the difference between the actual stress and the
characteristic of a secondary stress is that it is self-limiting. equivalent linear stress resulting from a radial temperature
Local yielding and minor distortions can satisfy the condi- distribution in a cylindrical shell
tions that cause the stress to occur and failure from one (c) the thermal stress in a liner material that has a
application of the stress is not to be expected. Examples coefficient of expansion different from that of the base
of secondary stress are: metal
(1) general thermal stress [see KD-210(l)(I)]
(s) Stress Cycle. A stress cycle is a condition in which
(2) bending stress at a gross structural discontinuity the alternating stress difference (see Article KD-3) goes
(0) Shape Factor. The shape factor, Q', is defined as the from an initial value through an algebraic maximum value
ratio of the moment that produces a full plastic section (no and an algebraic minimum value, and then returns to the
strain hardening) to the bending moment that produces initial value. A single operational cycle may result in one
initial yielding at the extreme fibers of the section. Q' = or more stress cycles.
1.5 for a rectangular section.
(p) Shear Stress, 7. The shear stress is the component
of stress tangent to the plane of reference.
(q) Stress Intensity, S. The stress intensity is defined as EQUATIONS FOR CYLINDRICAL
twice the maximum shear stress. In other words, the stress AND SPHERICAL SHELLS
intensity is the difference between the algebraically largest
Below are equations for the limits of the design pressure.
principal stress and the algebraically smallest principal
The purpose of these requirements is to ensure adequate
stress at a given point. Tension stresses are considered
safety against collapse. These equations are only appro-
positive and compression stresses are considered negative.
priate for spherical and cylindrical shells remote froni. dis-
(r) Thermal Stress. A self-balancing stress produced by
continuities.
a nonuniform distribution of temperature or by differing
thermal coefficients of expansion. Thermal stress is devel-
oped in a solid body whenever a volume of material is
prevented from assuming the size and shape that it normally
should under a change in temperature. For the purpose of The shell shall have a diameter ratio Y (see KD-250)
establishing allowable stresses, two types of thermal stress which meets the requirements of KD-221.1, KD-221.2,
are recognized, depending on the volume or area in which KD-221.3, or KD-221.4, as applicable.
distortion takes place, as follows:
(I) general thermal stress, which is associated with KD-221.1 Cylindrical Monobloc Shells. The design
distortion of the structure in which it occurs. If a stress of pressure PD shall not exceed the limit set by the equation:
this type, neglecting local stress concentrations, exceeds
twice the yield strength of the material, the elastic analysis 2
PD = 3' (Sy)ln(Y)
may be invalid and successive thermal cycles may produce
incremental distortion. Therefore, this type is classified
KD-221.2 Cylindrical Layered Shells. For shells con-
as secondary stress in Fig. KD-240. Examples of general
sisting of n layers with different yield strengths, the equa-
thermal stress are:
tion in KD-221.1 is replaced by:
(a) stress produced by an axial temperature gradi-
ent in a cylindrical shell.
(b) stress produced by temperature differences
between a nozzle and the shell to which it is attached.
PD = :L 3'
n

j=l
2
(Sy)ln(lj)

(c) the equivalent linear stress produced by the

where Syj and 1) are the yield strength and diameter ratio
radial temperature gradient in a cylindrical shell. Equiva-
for each layer.
lent linear stress is defined as the linear stress distribution
which has the same net bending moment as ,the actual KD·221.3 Spherical Monobloc Shells. The design
stress distribution. pressure PD shall not exceed the limit set by the equation:
(2) local thermal stress, which is associated with
almost complete suppression of the differential expansion
and thus produces no significant distortion. Such stresses

lyright ASME International

vided by IHS under license with ASME
reproduction or networking permitted without license from IHS
 KD-221.4 Spherical Layered Shells. For shells con- ( 1) For location in the component subject to evalua-
sisting of n layers with different yield strengths, the equa- tion, determine the change in total equivalent plastic strain,
tion in KD-221.3 is replaced by: t:H=peq,k> for the k!h load step increment in the analysis.

(2) Determine the limiting triaxial strain cL.k> for the

k!h load step increment using the equation below.
PD = ~ ~ (Sy) In(Y)
j=l

KD-221.5 Additional Loads. If the shell is subject to

~ (/T1'k + /TZ,k + /T3,k _ .!.) )
l+mz 3/Tek 3
cL,k = CLu e '
loading in addition to the internal pressure, the design shall
(
be modified as necessary so that the collapse pressure in
the presence of the additional load is greater than or equal
to 1.732 times the design pressure. where:
cpeq total equivalent plastic strain
e 2.7183, approximate value of the base ofthe natu-
ral logarithm
The shells shall have a diameter ratio that shall ensure
the same safety against collapse as in KD-221. This means
EI = minimum specified elongation, percent
CLk maximum permitted local total equivalent plastic
that the same equations are applicable for controlling the
strain at any point at the k!h load increment
diameter ratios for shells under external pressure as those
CLu maximum of m2, m3 and m4
given in KD-221 for shells under internal pressure, pro-
In natural logarithm
vided that all loadings are considered and the longitudinal
m2 value calculated from Table KD-230
stress remains the intermediate principal stress (i.e.,closed
m3 value calculated from Table KD-230
end case). Shells under external pressure shall also be
checked for safety against buckling. For the special case
m4 = value calculated from Table KD-230
ms value listed in Table KD-230
of cylindrical monobloc shells, the following equation shall
R Sy/Su
be used:
RA minimum specified reduction of area, percent
P D-
_ E(Y - 1)3 O"lk = principal stress in the "1" direction at the point
40(1 - pZ)y3 of interest for the k!h load increment
0"2k principal stress in the "2" direction at the point
but in no case shall exceed the value of PD given by the of interest for the k!h load increment
equation in KD-221.1.
0"3k principal stress in the "3" direction at the point
of interest for the k!h load increment
07 O"ek von Mises equivalent stress at the point of interest
A08
The equations for cylindrical and spherical shells in Sy yield strength at the analysis temperature (see Sec-
KD-220 need not be used, if an analysis is conducted using tion II, Part D, Table Y-l)
numerical methods such as elastic-plastic finite element Su tensile strength at the analysis temperature (see
or finite difference analysis, and the results satisfy the Section II, Part D, Table U)
requirements in KD-230(a) through (t). The material shall (3) Determine the strain limit damage for the k!h load
be assumed to be ideal elastic, perfectly plastic (nonstrain- step increment using the following equation:
hardening') with a yield strength equal to the value at
temperature from the appropriate table in Section II, Part D.
(a) The design margin against collapse load [see where:
KD-21O(e)(6)] shall be at least 1.732 in the case of the
most severe combination of specified loading.
De,k damage occuring during the h load step e
increment
(b) In addition, a design margin of at least 1.732 shall
D..cpeq,k change in total equivalent plastic strain drain
be applied to the load that causes the total accumulated
during the kth load step increment
damage at any point in the structure, De, to exceed 1.0
considering all loadings that will be applied to the compo- (4) Add the damage occuring during the k!h load step
nent and all of the load step increments in the analysis in increment, De,k> to the sum of the incremental damage
accordance with the following procedure: occuring at each previous increment to obtain the accumu-
lated damage, De.
1 A small amount of strain hardening may be used if necessary to (5) Repeat the process in subparas. (l) through (4)
achieve stability in the numerical solution. This strain hardening shall for all load step increments in the analysis.
not result in a calculated value of stress at any point, at the maximum
load applied during the analysis, which is more than 5% above the yield (6) If the component has been cold-formed without
strength. subsequent heat treatment, calculate the damage from

Iyright ASME International

l/ided by IHS under license with ASME
reproduction or netwol1<ing permitted without license from IHS