All relevant failure modes shall be considered in the design
for all relevant load scenarios and design conditions. The design
conditions are given in the earlier part of this chapter, and the
load scenarios are covered by 4.17.2.
4.18.1
Ultimate design condition
Structural capacity may be determined by testing, or by
analysis, taking into account both the elastic and plastic material
properties, by simplified linear elastic analysis or by the Code provisions.
4.18.1.1 Plastic deformation and buckling shall
be considered.
4.18.1.2 Analysis shall be based on characteristic
load values as follows:
| Permanent loads:
|
Expected values
|
| Functional loads:
|
Specified values
|
| Environmental loads:
|
For wave loads: most probable
largest load encountered during 108 wave encounters.
|
4.18.1.3 For the purpose of ultimate strength
assessment, the following material parameters apply:
-
.1.1
Re
= specified minimum
yield stress at room temperature (N/mm2). If the stress-strain
curve does not show a defined yield stress, the 0.2% proof stress
applies.
-
.1.2
Rm
= specified minimum
tensile strength at room temperature (N/mm2).
-
For welded connections where under-matched welds, i.e. where
the weld metal has lower tensile strength than the parent metal, are
unavoidable, such as in some aluminium alloys, the respective Re
and Rm
of the welds, after
any applied heat treatment, shall be used. In such cases, the transverse
weld tensile strength shall not be less than the actual yield strength
of the parent metal. If this cannot be achieved, welded structures
made from such materials shall not be incorporated in cargo containment
systems.
-
.2 The above properties shall correspond to the
minimum specified mechanical properties of the material, including
the weld metal in the as-fabricated condition. Subject to special
consideration by the Administration or recognized organization acting
on its behalf, account may be taken of the enhanced yield stress and
tensile strength at low temperature. The temperature on which the
material properties are based shall be shown on the International
Certificate of Fitness for the Carriage of Liquefied Gases in Bulk
required in 1.4.
4.18.1.4 The equivalent stress σC
(von
Mises, Huber) shall be determined by:
|
σx
|
= |
total normal stress in x-direction; |
|
σy
|
= |
total normal stress in y-direction; |
|
σz
|
= |
total normal stress in z-direction; |
|
τxy
|
= |
total shear stress in x-y plane; |
|
τxz
|
= |
total shear stress in x-z plane; and |
|
τyz
|
= |
total shear stress in y-z plane. |
The above values shall be calculated as described in 4.17.3.
4.18.1.5 Allowable stresses for materials other
than those covered by chapter 6 shall be subject to approval by the
Administration or recognized organization acting on its behalf in
each case.
4.18.1.6 Stresses may be further limited by fatigue
analysis, crack propagation analysis and buckling criteria.
4.18.2
Fatigue design condition
4.18.2.1 The fatigue design condition is the design
condition with respect to accumulated cyclic loading.
4.18.2.2 Where a fatigue analysis is required,
the cumulative effect of the fatigue load shall comply with:
|
ni
|
= |
number of stress cycles at each stress level during the life
of the tank; |
|
Ni
|
= |
number of cycles to fracture for the respective stress level
according to the Wohler (S-N) curve; |
|
nLoading
|
= |
number of loading and unloading cycles during the life of the
tank, not to be less than 1000footnote.
Loading and unloading cycles include a complete pressure and thermal
cycle;
|
|
NLoading
|
= |
number of cycles to fracture for the fatigue loads due to loading
and unloading; and |
|
Cw
|
= |
maximum allowable cumulative fatigue damage ratio. |
The fatigue damage shall be based on the design life of
the tank but not less than 108 wave encounters.
4.18.2.3 Where required, the cargo containment
system shall be subject to fatigue analysis, considering all fatigue
loads and their appropriate combinations for the expected life of
the cargo containment system. Consideration shall be given to various
filling conditions.
4.18.2.4.1 Design S-N curves used in the analysis
shall be applicable to the materials and weldments, construction details,
fabrication procedures and applicable state of the stress envisioned.
4.18.2.4.2 The S-N curves shall be based on a
97.6% probability of survival corresponding to the mean-minus-two-standard-deviation
curves of relevant experimental data up to final failure. Use of S-N
curves derived in a different way requires adjustments to the acceptable Cw
values specified in 4.18.2.7 to 4.18.2.9.
4.18.2.5 Analysis shall be based on characteristic
load values as follows:
-
| Permanent loads:
|
Expected values
|
| Functional loads:
|
Specified values or
specified history
|
| Environmental loads:
|
Expected load history, but
not less than 108 cycles
|
If simplified dynamic loading spectra are used for the estimation
of the fatigue life, they shall be specially considered by the Administration
or recognized organization acting on its behalf.
4.18.2.6.1 Where the size of the secondary barrier
is reduced, as is provided for in 4.4.3,
fracture mechanics analyses of fatigue crack growth shall be carried
out to determine:
-
.1 crack propagation paths in the structure;
-
.2 crack growth rate;
-
.3 the time required for a crack to propagate
to cause a leakage from the tank;
-
.4 the size and shape of through thickness cracks;
and
-
.5 the time required for detectable cracks to
reach a critical state.
The fracture mechanics are, in general, based on crack growth
data taken as a mean value plus two standard deviations of the test
data.
4.18.2.6.2 In analysing crack propagation, the
largest initial crack not detectable by the inspection method applied
shall be assumed, taking into account the allowable non-destructive
testing and visual inspection criterion, as applicable.
4.18.2.6.3 Crack propagation analysis under the
condition specified in 4.18.2.7: the simplified load distribution
and sequence over a period of 15 days may be used. Such distributions
may be obtained as indicated in figure 4.4. Load distribution and
sequence for longer periods, such as in 4.18.2.8 and 4.18.2.9 shall
be approved by the Administration or recognized organization acting
on its behalf.
4.18.2.6.4 The arrangements shall comply with
4.18.2.7 to 4.18.2.9, as applicable.
4.18.2.7 For failures that can be reliably detected
by means of leakage detection:
Predicted remaining failure development time, from the point
of detection of leakage till reaching a critical state, shall not
be less than 15 days, unless different requirements apply for ships
engaged in particular voyages.
4.18.2.8 For failures that cannot be detected
by leakage but that can be reliably detected at the time of in-service
inspections:
Predicted remaining failure development time, from the largest
crack not detectable by in-service inspection methods until reaching
a critical state, shall not be less than three times the inspection
interval.
4.18.2.9 In particular locations of the tank,
where effective defect or crack development detection cannot be assured,
the following, more stringent, fatigue acceptance criteria shall be
applied as a minimum:
Predicted failure development time, from the assumed initial
defect until reaching a critical state, shall not be less than three
times the lifetime of the tank.
4.18.3
Accident design condition
4.18.3.1 The accident design condition is a design
condition for accidental loads with extremely low probability of occurrence.
4.18.3.2 Analysis shall be based on the characteristic
values as follows:
-
| Permanent
loads:
|
Expected
values
|
| Functional
loads:
|
Specified
values
|
| Environmental
loads:
|
Specified
values
|
| Accidental
loads:
|
Specified
values or expected values
|
4.18.3.3 Loads mentioned in 4.13.9 and 4.15 need
not be combined with each other or with wave-induced loads.