International Journal of Fatigue 22 (2000) 41–52

www.elsevier.com/locate/ijfatigue

A method for the evaluation of the fatigue operational life of

submarine pressure hulls
a,*
L.B.R. Robles , M.A. Buelta b, E. Gonçalves c, G.F.M. Souza c

a
Department of Naval Architecture and Ocean Engineering, Polytechnic School, São Paulo University, Av. Prof. Mello Morais 2231, São
Paulo, SP 05508-900, Brazil
b
Department of Civil Engineering, Polytechnic School, São Paulo University, Av. Prof. Mello Morais 2231, São Paulo, SP 05508-900, Brazil
c
Department of Mechanical Engineering, Polytechnic School, São Paulo University, Av. Prof. Mello Morais 2231, São Paulo, SP 05508-900,
Brazil

Received 12 October 1998; received in revised form 9 February 1999; accepted 8 August 1999

Abstract

In this paper, a method is developed to calculate the fatigue lifetime of submarine pressure hulls, applying linear elastic fracture
mechanics, based on the Paris equation. It is known that this structure is submitted to high stresses caused by variable hydrostatic
pressure and residual stresses due to the welding process. The association of those variables may cause, after a certain period of
time, material fatigue and fracture. Through the application of Paris’ law, and based on an initial crack size, the crack size after a
given number of load cycles, defined by the number of submarine dives, is obtained, furnishing information for possible pressure
hull maintenance action, regarding crack repairs.  1999 Elsevier Science Ltd. All rights reserved.

Keywords: Fatigue; Fracture; Residual stress; Submarine; Weldments

1. Introduction to the technology of metallic structure construction dur-

ing recent years, it is not yet possible to eliminate the
In the area of naval construction, mechanical, and imperfections introduced into the structure by welding,
chemical industries, etc., most of the structural compo- such as residual stress and strain, the presence of cracks
nents are submitted to cyclic loads. Within these compo- or base metal thermally induced metallurgical changes.
nents the areas of high stress concentrations are more Fatigue can be defined as a process of successive frac-
sensitive to develop fatigue failure. In the specific case ture, which results from the repeated application of loads
of submarines, the connections between different or deformations, although the application of that cyclic
geometry areas, such as the connection of the cylinder load, just once, would not cause fracture [1]. For the
to the cap or the conical hull, are critical sections in occurrence of fatigue the following facts are necessary:
terms of fatigue failure analysis. In these cases, the study
of the fatigue lifetime should be very judicious, mainly if 1. Existence of cyclic stresses.
those areas are submitted to the weld process. Although 2. Existence of tension stresses.
welding presents advantages, when compared with the 3. Occurrence of local plastic deformations, due to shal-
traditional joining processes, such as hammer, riveting low roughness, metallurgical flaws and the presence
and foundry, its use in regions of high stress concen- of cracks.
trations increases the probability of fatigue failure dam-
age in these regions. Although the welding process of a submarine pressure
In spite of the countless and significant efforts applied hull is very well planned and executed, including many
weld tests to evaluate the weld quality, weld-induced
imperfections can be present, mainly flaws and tensile
* Corresponding author. Tel: +55 (0) 21 11 8185540; fax: +55 (0) residual stresses, which are essential parameters for the
21 11 8185717. occurrence of fatigue damage.
E-mail address: [email protected] (L.B.R. Robles) Therefore, the flaw propagation process for submarine

0142-1123/00/$ - see front matter  1999 Elsevier Science Ltd. All rights reserved.
PII: S 0 1 4 2 - 1 1 2 3 ( 9 9 ) 0 0 1 0 2 - 4
42 L.B.R. Robles et al. / International Journal of Fatigue 22 (2000) 41–52

hull fatigue becomes more important, as submarine loads that accumulate during time, causing the pressure
structural projects consider the application of high resist- hull fatigue process.
ance steels for hull construction, requiring more sophisti- The typical structure of a submarine is a cylindrical
cated welding processes, taking into account the follow- pressure hull, reinforced internally with frames and bulk-
ing problems: heads, and closed with conical hull and caps. This struc-
ture should resist the hydrostatic pressure that it will be
1. The welding of high strength steels should be carried submitted to during underwater operation [2].
out with more care, considering the high degree of
difficulty in preventing the occurrence of imperfec- 2.1. Considerations about fatigue of the pressure hull
tions in the welded materials, when compared with
steels of less strength. Submarine pressure hull fatigue has been analysed
2. High strength steels usually present a reduction in the many times since the beginning of the 1960s, after the
fracture toughness with an increase of the mechanical development of submarines using nuclear power, due to
strength. This reduction can imply a decrease of the the fact that they could operate for longer periods of
estimated fatigue lifetime for the structure, from the time, and at greater depths underwater than diesel-elec-
point of view of the process of crack propagation tric submarines, usually used by navies at that time [3].
under the action of cyclic stress. Several researchers have studied fatigue-induced
crack propagation in submarine pressure hulls. Dunham
The fatigue lifetime calculation for a structure, during [4] and Kilpatrick [5] conducted experiments with mod-
its operational phases is usually executed with the aid els of pressure hulls, submitted to the action of oscillat-
of empirical or semi-empirical theoretical relationships ing external pressure, monitoring crack propagation in
that try to analyse the phenomenon of crack propagation. the structure. The authors concluded that cracks grow in
Linear–elastic or elastic–plastic fracture mechanics the pressure hull inner part, in the toe of the welds, and
allow the calculation of the critical dimensions of the propagate first around the circumference of the pressure
flaw or defect, which for a given material type and under hull, becoming a long circumferential crack, and then
specific mechanical loading, will originate an unstable propagating through the thickness of the pressure hull
propagation, inducing the fracture of the component. [6].
With regard to submarine pressure hull fatigue analy- These authors also concluded that pressure hull struc-
sis, in addition to the definition of the external and tural geometry, the magnitude of cyclic hydrostatic load-
internal cyclic loads acting on the structure due to ing and initial crack size are essential parameters for the
applied hydrostatic pressure and the action of residual evaluation of pressure hull fatigue damage.
stress, it is necessary to estimate the size and the
geometry of the defects introduced into the structure dur-
2.2. Crack growth propagation model
ing the manufacturing steps, and the material mechanical
properties such as Paris’ law constants and fracture
toughness, which represent the parameters that govern Linear elastic fracture mechanics describe the fracture
fatigue crack growth and unstable fracture. resistance of a structure in terms of a parameter named
In this paper, a method is proposed for calculation of stress intensity factor (K), which depends on the
the fatigue lifetime of submarine pressure hulls, taking geometry of the structure, the crack size and stress
into consideration the oscillating nature of the stress induced by internal or external loads.
induced by hydrostatic pressure during the immersion Paris and Erdogan [7] developed a semi-empirical
and submersion periods, the presence of physical defects relationship that relates the rate of crack growth (da/dN)
introduced by the welding process, such as cracks, and with the variation of the stress intensity factor range ⌬K,
also thermally induced residual stress. including a dependence relationship with a constant (C)
and a coefficient (m) obtained experimentally, which
characterize the material. This relationship is showed
2. Submarine structure below:
da
A submarine is by definition ‘A vessel that works or ⫽C(⌬K)m (1)
dN
usually operates below the surface of the water.’ In the
naval sense the submarine operates in great oceans and where da/dN is the fatigue-crack-growth rate,
seas in a salt water environment. When submerged, the ⌬K=Kmax⫺Kmin is the stress-intensity factor range, and
hull is under external hydrostatic pressure, which is Kmax and Kmin are the maximum and minimum values of
dependent on the depth at which the vessel is operating. K in cyclic loading, related to the maximum and mini-
So, the constant variations of depth, related to the mili- mum stresses in the load cycle.
tary aspects of the vessel’s mission, generate cycles of Linear elastic fracture mechanics considers a small
 L.B.R. Robles et al. / International Journal of Fatigue 22 (2000) 41–52 43

plastic deformation ahead of the crack tip [8,9]. If the Furthermore, in these areas, due to the hull construc-
plastic zone is large compared to the crack size, a correc- tion planning, there are several welding lines that can
tion in the stress intensity factor must be considered, in induce the presence of residual stress and cracks.
order to account for the effects of this plastic zone. Paris’ Therefore, regarding crack growth under cyclic load-
law is employed in the model of pressure hull fatigue ing, these areas can be considered to be the most critical
analysis. for fatigue analysed in this study, the most critical area
is a cone–cone intersection, near the stern of the submar-
ine.
3. Method of fatigue lifetime calculation The conclusion of this theorical study is similar to
the experimental results presented by Kilpatrick [5], and
As for fatigue analysis, the definition of the structural Dunham [4], where fatigue failure, defined by crack
detail that is submitted to the highest stress is very growth under cyclic loading, until hull perforation,
important, because it is the most probable region for always happened in the areas where there were intersec-
fatigue failure. tions of two axisymetrical solids, and close to welding
Considering the typical structure of a submarine press- lines.
ure hull, shown in Fig. 1, a great amount of stress analy-
sis was executed by the authors, employing analytical 3.1. Crack model
semi-empirical and numerical methods, such as the finite
element method, considering as external loading hydro- Considering the analysis related above, in this study
static pressure, in order to define the stress distribution we will consider a crack along the circumference of the
in several areas of the hull. hull, in the cone–cone intersection, close to the weld line
It was verified that the most critical areas for fatigue that joins the two solids. The growth of the crack depth
evaluation are the intersection between two axisymetr- through the thickness is considered to be constant along
ical solids, such as cone–cylinder, cone–cap, or between the circumference of the hull. The model of the crack
two cones with different conicity, due to the high stress assumes a plane strain condition, where the relationship
acting on these areas induced by hydrostatic loading. of the circumferential stress (sq) and the longitudinal

Fig. 1. Diagram for the calculation of stress due to hydrostatic pressure.

44 L.B.R. Robles et al. / International Journal of Fatigue 22 (2000) 41–52

stress (sL) acting on the pressure hull, induced by hydro-

static pressure, is given by:
sq⫽vsL (2)

This condition was verified by structural analysis of

this intersection, applying the finite element method.

3.2. Calculation of stress intensity factor Fig. 3. Model for moment bending [10].

The load applied to the structure of the submarine is sFw2

M⫽
composed of two types of loading: load due to hydro- 6
static pressure and internal load due to residual stress.
The pressure of the external water, which is variable and sF is the bending stress, w=2b is the thickness of
according to the operational depth of the submarine, the plate, a is the depth of the crack, g is the boundary
develops the fatigue phenomenon. For this load, the factor for K as a function of a and w. sF⬍0, because
stresses in the structure are defined using the finite the bending, as shown in Fig. 3, cause closure of the
element method. The distribution of stress in the thick- crack which initiates on the internal face of the hull.
ness of the plate is not constant. This can be divided into As for the residual stress acting on the pressure hull,
two components, normal (membrane) and bending. Fig. Souza [11] conducted a study to estimate the distribution
1 shows a scheme of the sequence for calculation of the of the residual stresses. In that study, he considered the
stresses in the pressure hull plate. total longitudinal residual stress as the composition of
Considering that the stress acting on the hull due to the residual stress in the longitudinal weld line, with the
hydrostatic pressure is linear elastic, the stress intensity same magnitude of the yield stress of steel and constant
factor related to this load can be defined through the along the thickness of the plate, with residual stress
application of the superposition principle, being com- induced by the fillet weld along the hull circumference,
posed of the sum of the stress intensity factor related to with magnitude equal to 75% of the yield stress in the
the normal stress and the stress intensity factor related internal part of the plate, with linear variation along the
to the bending stress. thickness of the plate, as shown in Fig. 4.
For the component of normal stress, the stress inten- For a linear opening mode stress distribution along the
sity factor Kn is represented by Fig. 2 and Eq. (3) [10]: crack surface, the stress intensity factor for residual
stress is given as, [12]:
冑
Kn⫽1.12sn pak(a/b) (3)
Kr⫽(1.1215P⫹0.4398Q) pa 冑 (5)
where k is the stress intensity factor boundary correction,
a is the depth of the crack, w=2b is the thickness of the where a is the depth of the crack, P is the stress at the
plate, l is the length of the hull circumference (plane crack tip, and P+Q is the stress at the surface of the
strain condition), 1.12 is the constant obtained as a func- opening of the crack (Fig. 5).
tion of the material and the opening mode of the crack, P and Q can vary as a function of the growth of the
and sn is the normal stress (sn⬍0 as shown in Fig. 2). crack length since the distribution of residual stress is
For the component of bending stress as shown in Fig. variable.
3, the stress intensification is [10]: The stress intensity factor (Kt) for a crack in the press-
ure hull, considering a given operational depth for the
6.M submarine, which corresponds to a specific hydrostatic
Kf⫽ g(a/w) (4)
(w−a)3/2 pressure, is defined by the sum of the stress intensity
factor related to the hydrostatic pressure induced normal
where M is the bending moment, given by: stress (sn) and bending stress (sf) and the stress inten-
sity factor related to the residual stress (sr), and is
expressed as:
Kt⫽Kn⫹Kf⫹Kr (6)

These values of stress intensity factors are corrected in

order to take into consideration the influence of the plas-
tic zone at the crack tip. This correction is made through
the application of the Dugdale model, described by
Fig. 2. Model for longitudinal load [10]. Broek [13].
 L.B.R. Robles et al. / International Journal of Fatigue 22 (2000) 41–52 45

Fig. 4. Longitudinal distribution of residual stress [11].

can be performed, as shown in Fig. 6.

With this graphical representation, a mathematical
relationship can be established between the stress
intensity range, induced by a given operational depth
range, and the crack depth. The authors of this paper
propose that these curves can be represented by poly-
noms, which can be fitted to the date applying
Fig. 5. Residual stress in the tip of the crack [12]. regression techniques.
So, for a given operational depth range, the stress
intensity factor can be obtained as a function of crack
depth, ⌬Kk(a).
3.3. Algorithm for fatigue lifetime calculation 3. The fatigue crack growth in the hull of a submarine,
in the presence of a crack has a behavior of the type:
The fatigue lifetime calculation is based on the Paris (da)/(dN)=C(⌬K1)m, where C and m are of the press-
model for crack growth prediction and must be perfor- ure hull material.
med according to the following algorithm: Considering a specific operational depth range,
which defines a given stress intensity range as a func-
1. Calculation of the stress intensity factor (KI) as a tion of the crack depth, as shown in (2), the number
function of the depth of the crack (a) and the thick- of cycles for a given crack, with an initial depth
ness of the pressure hull plate [2], for several load (ainitial), reach a final depth (afinal), is defined by:
conditions equivalent to many operational depths (hn),

冕
(afinal)k
and considering many crack depths. These results will
1
be obtained according to Eq. (6), as a combination of Nk⫽ (⌬K(a)k)−mda (8)
the stress intensity factors calculated for the effects C
(ainitial)k
of the external load, as well as for the effect of the
residual stress, as a function of crack length. These where (ainitial)k is the initial value of the crack depth
values will be presented in the form of tables and in the instant that the submarine enters the range of
graphs. depth k, (afinal)k is the final value of the crack depth
2. Compute the stress intensity range, considering a in the instant that the submarine leaves the range of
given crack depth and a given stress range, equivalent the depth k, Nk is the number of cycles that the sub-
to a variation of the operational depth of the submar- marine has to execute in the range of the depth k,
ine. This stress intensity range is defined as: varying among the quotas i and j, until the crack
reaches the (afinal)k.
(⌬KI)k⫽(KI)i⫺(KI)j (7)
For a group of specific operational depth ranges,
where i and j are indicative of the depth quotas pre- considering a given initial crack depth, the number of
established by the analyst, in function of the probable cycles necessary for the crack to grow to a given final
missions of the submarine, and k is the index that crack depth can be defined through the application of
identifies the range of the variation of the referring Eq. (8), and the date can be shown in a graphical
stress to the quotas of the indexes i and j. form, as shown in Fig. 7.
Considering a given crack depth, the stress inten- 4. Define the operational profile of the submarine, con-
sity factor range can be related to that crack depth, sidering the number of dives that the submarine will
according to a given submarine depth operational do, classifying the variation in operational depth
range. So, a graphical representation of the stress ranges in each of these dives according to the depth
intensity factor range according to a given crack depth operational range defined in (2). Put the classified
46 L.B.R. Robles et al. / International Journal of Fatigue 22 (2000) 41–52

Fig. 6. Stress intensity range (⌬K) as a function of the operational depth range and crack depth (a).

Fig. 7. Number of cycles in a given depth operational range for the crack growth to a specific depth.

dive sequence in a given order, such as, the submarine for the dive sequence defined in item (4), it is possible
will first dive according to operational range 1, then to define the expected final crack depth at the end of
according to operational range 4, and so on. This this operational sequence.
diving sequence is defined by the submarine designer 7. The maximum crack depth admissible for the submar-
or by the submarine operator or according to a deal ine structure is defined through the application of the
between them. linear elastic fracture mechanics concepts, where
5. Considering the given initial crack depth, and the first there will be an unstable fracture when the stress
sequence of dive, for example N1 diving cycles in intensity factor defined by Eq. (6) equals the material
operational range 1, it is possible to define, through critical stress intensity factor (KIC). This maximum
the graphical representation shown in Fig. 8, the final crack depth is defined as:
crack depth at the end of this diving cycle.
KT(acritical)⫽KIC (9)
6. The afinal crack depth, defined in (5), will be used as
the ainitial crack depth for the second sequence of dive 8. If the crack depth defined at the end of step 1 is bigger
in a given operational range, and repeating the pro- than the critical crack depth, the pressure hull struc-
cedure shown in (5), the crack depth at the final dive ture will present unstable fatigue fracture during the
of the sequence is defined. Applying this procedure previously predicted operational life. So the oper-
 L.B.R. Robles et al. / International Journal of Fatigue 22 (2000) 41–52 47

Fig. 8. Crack depth growth in N1 cycles in operational depth range 1.

ational dive sequence must be changed or the struc- slint(hmax)=⫺301.7 MPa, which is the longitudinal nor-
tural design must be reviewed, in order to fulfil the mal stress in the internal face of the plate at the
requirements of the operational life of the submarine. maximum operation depth.
Applying the integral defined in Eq. (8), for the slext(hmax)=⫺18.82 MPa, which is the longitudinal nor-
operational range, in the device sequence where the mal stress in the external face of the plate at the
final crack depth is expected to be bigger than the maximum operation depth.
critical crack depth, and defining afinal equal to acritical,
it is possible to define the maximum number of diving Based on these results, knowing that the distribution of
cycles for this operational range, in order to prevent the membrane stress and bending stress are dependent
unstable fracture; this number of cycles is named Np. on the operational depth (h), the following equations are
9. The total number of diving cycles for the submarine obtained, to express the hydrostatic pressure-induced
could be defined as the sum of the number of diving stress acting on the submarine pressure hull:
cycles defined for each operational range, the number
of cycles (Nv) is N1+N2+…Np, where Np was defined h
slext⫽slext(hmax) (10)
in item (8). hmax
h
slint⫽slint(hmax) (11)
hmax
4. Application of the method of calculation slint−slext h
sF(h)⫽ (12)
2 hmax
In this section, the methodology for pressure hull
fatigue evaluation will be applied to the submarine, the sn(h)⫽slint⫺sF(h) (13)
structural plan of which is presented in Fig. 1. The struc-
ture of this submarine is constructed with HY-80 high
strength structural steel, and the analysis will be made
for the cone–cone connection. slext(hmax), longitudinal normal stress in the external
part of the plate at the maximum depth.
4.1. Stress due to hydrostatic pressure slint(hmax), longitudinal normal stress in the internal
part of the plate at the maximum depth.
The stress is calculated using a hydrostatic pressure slint, longitudinal normal stress in the internal part of
of 2.5 MPa (250 m of maximum depth of the submarine) the plate at an operational depth h.
applied on the external face of the pressure hull plate, slext, longitudinal normal stress in the external part of
in the cone–cone connection region, as shown in Fig. 1. the plate at an operational depth h.
Thus, employing the finite element program Algor sF(h), stress that generates bending at an operational
[14] to model this structural region, the following results depth h.
were obtained for the stress action on the pressure hull sn(h), stress that generates compression at an oper-
at the cone–cone connection: ational depth h.
48 L.B.R. Robles et al. / International Journal of Fatigue 22 (2000) 41–52

4.2. Residual stress where:

Q⫽0 P⫽551 MPa.
The stress distribution is defined taking the yield stress
of HY-80 steel as 551 MPa, based on the hypothesis and acritic=5.45 mm, the maximum depth of the crack,
presented in Fig. 4. considering the influence of plasticity at the crack tip.
Fig. 9 shows a scheme of the area studied, the location Therefore, defining the stress intensity factors Kn, Kf
of the crack and the resulting stresses acting on the struc- and Kr as a function of the depth of the crack (a) con-
ture, due to hydrostatic loading and welding residual sidering the influence of plasticity at the crack tip, it is
stress. possible to define the stress intensity factor for the crack,
as shown in Fig. 10.
4.3. Stress intensity factor calculation For the calculation of the stress intensity factor, the
authors used the possible operational depths for the sub-
Starting from the characteristics equations for the dif- marine equal to 0, 50, 100, 200 and 250 m.
ferent loading types (Equations (3), (4) and (5)), the Based on the results in Fig. 10, and applying item (2)
stress intensity factors are determined as a function of of the algorithm presented in Section 3.3, the authors
the crack (a) length for each depth defined for submar- defined 15 operational depth ranges for cyclic load esti-
ine operation. mation. The stress intensity ranges defined for these
Barson and Rolfe [10] present the behavior of the operational depth ranges for a given crack depth are
fracture toughness KIC, for HY-80 steel as a function graphically presented in Fig. 11.
of temperature. For martensitic steels, the critical stress
intensity factor is: 4.4. The fatigue lifetime calculation
KIC⫽80 ksi ⵪ 88 MPa for a temperature of T⫽⫺175°C
Having obtained the stress intensity factor for each
operational depth range (⌬KI) as a function of the depth
Considering the stress intensity factors independently of the crack (a), the number of cycles or the lifetime in
calculated as a function of the crack depth (a), it can be fatigue of the structure can be determined, following
seen that the crack is subjected to larger stress when the steps (3) to (9) of the algorithm presented in Section
submarine is on the surface, depth equal to 0 m. So, the 3.3, taking the maximum admissible crack depth equal
maximum crack depth is obtained through the equality to 5.45 mm.
between KIC and Kr, as shown below: For martensitic steels (HY-80 steel) the coefficients
of the Paris law are equal to C=1.3534×10⫺7 and m=2.25
冑
(1.1215P+0.43Q) pacritic for stress in MPa and crack depth in meters.
Kr⫽ ⫽KIC (14)
1000 Furthermore, considering the structural welding plan

Fig. 9. Distribution of stress and location of the crack in the plate.

 L.B.R. Robles et al. / International Journal of Fatigue 22 (2000) 41–52 49

Fig. 10. Stress intensity factor Vs crack depth, with correction due to plasticity.

Fig. 11. Variation of the stress intensity factor (⌬K) Vs depth of crack.

and non-destructive inspection applied to the welds, the

initial crack depth defined for the structure is 0.2 mm. Applying the equation, and following step (3) of the
So Eq. (8) applied for the definition of number of cycles fatigue lifetime algorithm, the number of cycles and
necessary to induce a given crack growth is written as: crack lengths for each operation range ares defined, as
shown in Fig. 12, taking a maximum crack depth of 6.5

冕
a
107 mm just for numerical analysis.
N⫽ (⌬K(a))−2.25da ⌬KI⫺MPa. In Fig. 13, schematically representing Fig. 12, is
1.3534
0.2 shown the process of calculation of the total number of
50 L.B.R. Robles et al. / International Journal of Fatigue 22 (2000) 41–52

Fig. 12. Lifetime of the fatigue in cycles (N) Vs depth of the crack (a).

cycles for fatigue occurrence, being the supposed differ- crack depth is smaller than the maximum allowed crack
ent numbers of cycles, Ni, being realized in different depth, the procedure is repeated for more operational life
operational depth ranges and in a certain sequence, cycles, for one and a half years, always considering the
defined by the typical mission of the submarine through- final crack depth at the end of one operational life cycle
out its lifetime. In order to define the crack growth as the initial crack depth at the beginning of another
throughout the submarine’s operational life, an oper- operational life cycle. The results of this calculation pro-
ational depth classification was developed, taking into cedure are shown graphically in Fig. 14, for the typical
consideration the maneuvers executed by a submarine operational profile that was adopted, where the expected
similar to the analysis in this study, during one and half crack depth is presented as a function of the years pre-
years. The operational depth ranges, in meters, employed dicted for the operational life of the submarine.
in this study are: (50–0), (150–0), (100–0), (200–50), Based on the results presented in Fig. 14, it can be
(250–150), (250–50), (250–0). seen that, for the typical operational depth cycle
Applying items (5) to (9) of the fatigue lifetime algor- employed in this paper corresponding to operational
ithm, the behavior of the crack growth is a function of cycles of one and a half years, the maximum crack depth
the number of cycles of dives, for several operational allowable for the pressure hull structure will be achieved
ranges. Thus, given an operational range when the sub- after 40 years of the submarine’s operational life, that is
marine goes from a depth of 0 to 250 m (⌬H=250–0), almost twice the usual life of military submarines.
with an initial crack depth (a(initial)1), diving N1 cycles in Furthermore, at the beginning of the operational life
this range, a depth for the final crack a2 will be obtained, the crack growth is very small, becoming greater after
which will be the initial crack for the following oper- the 20 operational years.
ation. If the submarine dives, for example, in another This fact strengthens the necessity for detailed struc-
operational depth range ⌬H=250–50, etc. a new dimen- tural inspections of the submarine pressure hull during
sion of the crack at the end of this dive cycle will be its operational life, especially when it is close to its orig-
obtained. This procedure is repeated, successively, to inal predicted operational lifetime, because the crack can
other operational depth ranges, covering the typical dive grow very fast after this period of time.
profile for one and half years of the submarine’s life. Finally, it must be remembered that the crack growth
At the end of this procedure, the crack depth at the history presented is dependent on the dive cycle esti-
end of this operational life cycle is defined. As the final mated for the submarine. Any change in this cycle will