Section 1: 1.1 Nature of Ship Structures
Section 1: 1.1 Nature of Ship Structures
Section 1: 1.1 Nature of Ship Structures
Introduction
1.1 Nature of Ship Structures. The size and princi- influences on the structural arrangement and design. As
pal characteristics of a new ship are determined pri- an example, large oil tankers having fully loaded dis-
marily by its mission or intended service. In addition to placements exceeding 5978 MN (600,000 tons. Through-
basic functional considerations, there are requirements out this book tons indicate long ton-force, 1 ton = 2240
such as stability, low resistance, high propulsive effi- lbf) and dimensions of 400 m (1,312 ft) in length, 63 m
ciency, and navigational limitations on draft or beam, all (207 ft) in breadth, 35.9 m (118 ft) in depth, with a loaded
of which influence the choice of dimensions and form. draft of 28.5 m (94 ft), are currently in operation. Ships
Within these and other basic constraints, the ship’s struc- are among the most complex of structures and this is due
ture must be designed to sustain all of the loads ex- in part to their mobility. Good resistance and propulsive
pected to arise in its seagoing environment. As a result, characteristics dictate that the external surface of the
a ship’s structure possesses certain distinctive features hull or shell must be a complex three-dimensional curved
not found in other man-made structures. surface, and because the shell plating is one of the major
Among the most important distinguishing characteris- strength members the structural configuration may not
tics of ship structures are the size, complexity, and multi- always be chosen solely on the basis of optimum struc-
plicity of function of structural components, the random tural performance. Furthermore, the structural behavior
or probabilistic nature of the loads imposed, and the un- of the many geometrically complex members that consti-
certainties inherent in our ability to predict the response tute a ship’s hull is difficult to analyze, and the construc-
of the structure to those loads. In contrast to land-based tion of the vessel may be complicated because there are
structures, the ship does not rest on a fixed foundation few members having simple shapes.
but derives its entire support from buoyant pressures ex- 1.3 Multipurpose Function of Ship Structural Compo-
erted by a dynamic and ever changing fluid environment. nents. In contrast to many land-based structures, the
The methods of analysis employed by the naval archi- structural components of a ship are frequently designed
tect in designing and evaluating the structure of a ship to perform a multiplicity of functions in addition to that
must be selected with these characteristics in mind. Dur- of providing the structural integrity of the ship. For in-
ing the past few decades, ship structural design and anal- stance, the shell plating serves not only as the princi-
ysis have undergone far-reaching changes toward more pal strength member but also as a watertight envelope
rationally founded practices. In addition, the develop- of the ship, having a shape that provides adequate stabil-
ment of readily available computer-based analytical tools ity against capsizing, low resistance to forward motion,
has relieved the naval architect of much of the routine acceptable controllability, and good propulsive charac-
computational effort formerly involved in the analysis teristics.
of a ship’s structural performance. Nevertheless, many Internally, many strength members serve dual func-
aspects of ship structures are not completely amenable tions. For example, bulkheads that contribute substan-
to purely analytical treatment, and consequently the de- tially to the strength of the hull may also serve as
sign of the structure continues to involve a judicious and liquid-tight boundaries of internal compartments. Their
imaginative blend of theory and experience. locations are dictated by the required tank volume or
This section will deal in detail with the loads acting on subdivision requirements. The configuration of struc-
a ship’s hull, techniques for analyzing the response of its tural decks is usually governed by the arrangement of in-
structure to these loads, and both current and evolving ternal spaces, but they may be called upon to resist local
new methods of establishing criteria of acceptable struc- distributed and concentrated loads, as well as contribut-
tural design. A detailed description of ship structures and ing to longitudinal and transverse strength.
a discussion of the practical aspects of the structural de- Whereas in many instances structural efficiency alone
sign of ships as they are influenced by the combined ex- might call for beams, columns, or trusses, alternative
perience and analysis embodied in classification society functions will normally require plate or sheet-type mem-
rules is given in Chapters 17 and 18 of Lamb (2003). This bers, arranged in combination with a system of stiffen-
work should be treated as a complement to this chapter. ers, to provide resistance to multiple load components,
To aid in understanding the nature of the behavior of some in the plane of the plate and others normal to
ship structures, further details of some of their most im- it. An important characteristic of a ship structure is its
portant distinguishing will be given in the following sec- composition of numerous stiffened plate panels, some
tions. In some cases, it is helpful to compare the ship plane and some curved, which make up the side and
and its structure with other man-made structures and bottom shell, the decks, and the bulkheads. Therefore,
systems. much of the effort expended in ship structural analysis is
1.2 Size and Complexity of Ships. Ships are the concerned with predicting the performance of individual
largest mobile structures built by man, and both their stiffened panels and the interactions between adjoining
size and the requirement for mobility exert strong panels.
2 THE PRINCIPLES OF NAVAL ARCHITECTURE SERIES
1.4 Probabilistic Nature of Ship’s Structural Loads. stress analysis is usually carried out on an idealization
The loads that the ship structure must be designed to of the real structure. For example, beam theory may
withstand have many sources. There are static compo- be used to predict the stress distribution in part or the
nents, which consist principally of the weight and buoy- whole of the hull girder, even though it is known that the
ancy of the ship in calm water. There are dynamic com- ship geometry may not follow exactly the assumptions of
ponents caused by wave-induced motions of the water beam theory.
around the ship and the resulting motions of the ship Second, the actual properties of the materials of con-
itself. Other dynamic loads, usually of higher frequency struction may not be exactly the same as those assumed
than the simple wave-induced loads, are caused by slam- by the designer. As delivered from the mill, steel plates
ming or springing in waves and by the propellers or pro- and shapes do not agree precisely with the nominal di-
pelling machinery. These sometimes cause vibrations in mensions assumed in the design. Similarly, the chemical
parts or in the entirety of the ship. Finally, there may be and physical properties of the materials can vary within
loads that originate due to a ship’s specific function, such certain tolerance limits. The rules of classification soci-
as ice breaking, or in the cargo it carries, as in the case eties specify both physical and chemical standards for
of thermally induced loads associated with heated or re- various classes of shipbuilding materials, either in the
frigerated cargoes. form of minimum standards or in a range of acceptable
An important characteristic of these load components values. The materials that are actually built into the ship
is their variability with time. Even the static weight and should have properties that lie within these specified
buoyancy vary from voyage to voyage and within a voy- limits, but the exact values depend on quality control
age, depending upon the amount and distribution of in the manufacturing process and are not known in ad-
cargo and consumables carried. To design the structure vance to the designer. Furthermore, there will inevitably
of the ship for a useful life of 20 years or more, this time be some degradation of material physical properties, for
dependence of the loading must be taken into considera- example, caused by corrosion over the lifetime of the
tion. ship.
Like the sea itself, the loads imposed by the sea are Third, the integrity of ship construction contains a sig-
random in nature, and can therefore be expressed only in nificant element of skill and workmanship. When per-
probabilistic terms. Consequently, it is generally impos- forming a stress analysis, the designer may assume
sible to determine with absolute certainty a single value perfect alignment and fit of load-carrying members and
for the maximum loading that the ship structure will be perfectly executed welds. This ideal may be approached
called upon to withstand. Instead, it is necessary to use by the use of a construction system involving highly
a probabilistic representation in which a series of loads skilled workmen and high standards of inspection and
of ascending severity is described, each having a proba- quality control. Nevertheless, an absolutely flawless
bility corresponding to the expected frequency of its oc- welded joint or a plate formed precisely to the intended
currence during the ship’s lifetime. When conventional shape and fabricated with no weld-induced distortion or
design methods are used, a design load may then be cho- joint misalignment is a goal to strive for but one that is
sen as the one having an acceptably low probability of never attained in practice.
occurrence within a stated period (Section 2.3). In more It will be obvious that the uncertainties involved in
rigorous reliability methods (Section 5), the load data in the determination of both the loads and the structural
probability format can be used directly. responses to these loads make it difficult to establish
1.5 Uncertainty Associated with Ship’s Structural Re- criteria for acceptable ship structures. In the past, allow-
sponse. As a consequence of the complexity of the able stress levels or safety factors used by the designer
structure and the limitations of our analysis capabilities, provided a means of allowing for these uncertainties,
it is seldom possible to achieve absolute accuracy in pre- based upon past experience with similar structures. In
dicting the response of the structure even if the load- recent years, reliability principles have been applied,
ing were known exactly. In the case of the uncertain- using probability theory and statistics, to obtain a more
ties present in the predictions of structural loading, it is rational basis for design criteria. In the reliability ap-
necessary for the designer to consider the probable ex- proach to design, structural response data as well as
tent and consequences of uncertainties in the structural strength data can be expressed and used in probability
response prediction when making a judgment concern- format. These principles are discussed in Section 5.
ing the overall acceptability of the structure. One of the 1.6 Modes of Ship Strength and Structural Failure.
most important tasks facing the engineer is to properly Avoidance of structural failure is an overriding goal of
balance the acceptable level of uncertainty in their struc- all structural designers. To achieve this goal, it is nec-
tural response predictions and the time and effort that essary for the naval architect to be aware of the pos-
must be expended to achieve a higher level of accuracy. sible modes of failure and the methods of predicting
The existence of this uncertainty is then acknowledged their occurrence. The types of failure that can occur
and must be allowed throughout the design. in ship structures are generally those that are charac-
In ship structural performance prediction, there are at teristic of structures made of stiffened plate panels as-
least three sources of uncertainty. First, the designer’s sembled through the use of welding to form monolithic
STRENGTH OF SHIPS AND OCEAN STRUCTURES 3
structures with great redundancy (i.e., having many al- meet all the criteria. Although this ideal has not yet been
ternative paths for lines of stress). attained, steady progress is being made in that direction.
It should be noted that structural failure might occur The original set of requirements imposed upon the
in different degrees of severity. At the low end of the fail- ship will include the functional requirements of the
ure scale, there may be small cracks or deformations in owner and, in addition, institutional requirements such
minor structural members that do not jeopardize the ba- as those established by government and other regulatory
sic ability of the structure to perform its function. Such bodies concerned with safety, navigation, pollution pre-
minor failures may only have aesthetic consequences. At vention, tonnage admeasurement, and labor standards.
the other end of the scale is total catastrophic collapse The methods of selecting the overall dimensions and
of the structure, resulting in the loss of the ship. There the arrangement of the ship to meet these requirements
are several different modes of failure between these ex- have been dealt with in Lamb (2003). Thus, when design-
tremes that may reduce the load-carrying ability of in- ing the principal members of the ship structure, it may
dividual members or parts of the structure but, because be assumed that the overall dimensions of the ship and
of the highly redundant nature of the ship structure, do the subdivisions of its internal volume occupied by bulk-
not lead to total collapse. Such failures are normally de- heads, decks, and tank boundaries have already been de-
tected and repaired before their number and extent grow termined to meet these various requirements. The prob-
to the point of endangering the ship. lem of structural design then consists of the selection of
Four principal mechanisms are recognized to cause material types, frame spacing, frame and stiffener sizes,
most of the cases of ship structural failure, aside from and plate thickness that, when combined in this geomet-
collision or grounding. These modes of failure are as fol- ric configuration, will enable the ship to perform its func-
lows: tion efficiently for its expected operational lifetime.
r Buckling due to compressive or shear instability
At this point, to select the criteria to be satisfied by the
r Excessive tensile or compressive yield
structural components of the ship, the designer must rely
r Fatigue cracking
on either empirical criteria, including factors of safety
r Brittle fracture.
and allowable stresses, or on the use of reliability princi-
ples discussed in Section 5. The term synthesis, which is
The first three modes of failure are discussed in more defined as the putting together of parts or elements so as
detail in Section 4. The last one, brittle fracture, was to form a whole, is often applied to the process of ship
found to play a major role in the failure of many of the structural design.
emergency cargo ships built during World War II. The However, an additional element is needed to complete
causes of these failures ultimately were traced to a com- the design synthesis: finding the optimal combination
bination of factors associated with the relatively new of the various elements. Due to the complexity of ship
techniques of welded construction employed in build- structures, as well as the probabilistic nature of available
ing the ships. The solution to the problem was obtained information needed for certain vital inputs to the design
through the development of design details that avoided process, it is usually impossible to achieve an optimum
the occurrence of notches and other stress concentra- solution in a single set of calculations. Instead, some sort
tions, together with the selection of steels having a high of iterative procedure must be employed. The traditional
degree of resistance to the initiation and propagation method of ship structural design, involving the extrapo-
of cracks, particularly at low temperatures. Features lation of previous experience, can even be thought of as
termed crack arrestors were incorporated to provide an iterative process in which the construction and opera-
fail-safe designs by limiting the extent of propagation of tional experience of previous ships form essential steps.
any cracks that might actually have occurred. In each new design, the naval architect considers this
Because the control of brittle fracture is accomplished past experience and modifies the new design intuitively
principally through detailed design and material selec- to achieve an improved configuration. The successful de-
tion, it is only considered briefly in this chapter. Informa- signer is one whose insight, understanding, and mem-
tion on these topics may be found in Lamb (2003), Chap- ory, along with skill in methods of structural analysis,
ters 17 and 20. resulted in consistently improved previous designs in
1.7 Design Philosophy and Procedure. The develop- successive ships.
ment of completely rational structural design proce- Even when the most advanced methods are used,
dures is being pursued in several disciplines, including much of structural design consists of a stepwise process
civil, aeronautical, and mechanical engineering, as well in which the designer develops a structural configuration
as in naval architecture. Using such procedures, a set on the basis of experience, intuition, and imagination,
of requirements or criteria to be met by the structure then performs an analysis of that structure to evaluate
should first be formulated, then through the applica- its performance. If necessary, the scantlings are revised
tion of fundamental reasoning and mathematical analy- until the design criteria are met. The resulting configu-
sis, augmented by the introduction of certain empirical ration is then modified in some way that is expected to
information, it should be possible to arrive at a structural lead to an improvement in performance or cost, and the
configuration and a set of scantlings that simultaneously analysis is then repeated to re-ensure that the improved
4 THE PRINCIPLES OF NAVAL ARCHITECTURE SERIES
configuration meets the design criteria. Thus, a key ele- Formally, the final optimization step consists of a
ment in structural design is the process of analyzing the search for the best attainable (usually minimum) value
response of an assumed structure. The process of find- of some quantity such as structural weight, construction
ing a structural configuration having the desired perfor- cost, overall required freight rate for the ship in its in-
mance by synthesis is the inverse of analysis, and is not tended service (see Lamb 2003), or the so-called total ex-
nearly so straightforward, especially in the case of com- pected cost of the structure. The last of these quantities,
plex structures. Consequently, it is only after completing as proposed by Freudenthal (1969), consists of the sum
several satisfactory design syntheses that the process of of the initial cost of the ship (or other structure), the an-
optimization can take place. ticipated total cost of complete structural failure multi-
In summary, five key steps can be identified to charac- plied by its probability, and a summation of lifetime costs
terize the structural design process, whether it be intu- of repair of minor structural damages (see also Lewis
itive or mathematically rigorous: et al. 1973).
The search is performed in the presence of constraints
(a) Development of the initial configuration and scant-
that, in their most elementary form, consist of the re-
lings.
quirement that each member of the structure does not
(b) Analysis of the performance of the assumed de-
fail under the expected loadings—Steps (b) and (c).
sign.
Such an optimization procedure forms the basis for
(c) Comparison with performance criteria.
a sound economical design, whether it be carried out
(d) Redesign the structure by changing both the con-
automatically, using one of the formal mathematical op-
figuration and scantlings in such a way as to effect an
timization schemes, or manually, with or without ma-
improvement.
chine computational assistance for some parts of the
(e) Repeat the above as necessary to approach an op-
process.
timum.
Section 2
Ship Structural Loads
2.1 Classification of Loads. It is convenient to divide changing with time. They may be broken down into the
the loads acting on the ship structure into four categories following components:
as follows, where the categories are based partly upon r Wave-induced hull pressure variations
the nature of the load and partly upon the nature of the r Hull pressure variations caused by oscillatory ship
ship’s response.
motions
2.1.1 Static Loads. Static loads are loads that vary r Inertial reactions resulting from the acceleration of
slowly with time and change when the total weight of
the mass of the ship and its contents.
the ship changes, as a result of loading or discharge of
cargo, consumption of fuel, or modification to the ship 2.1.3 High-Frequency Dynamic Loads. High-
itself. Static loads are influenced by: frequency dynamic loads are time-varying loads of
sufficiently high frequency that they may induce a vibra-
rWeight of the ship and its contents tory response in the ship structure. Some of the exciting
rStatic buoyancy of the ship when at rest or moving loads may be quite small in magnitude but, as a result
rThermal loads resulting from nonlinear temperature of resonant amplification, can give rise to large stresses
gradients within the hull and deflections. Examples of such dynamic loads are the
r Concentrated loads caused by dry-docking and following:
grounding.
r Hydrodynamic loads induced by propulsive devices
2.1.2 Low-Frequency Dynamic Loads. Low- on the hull or appendages
frequency dynamic loads are loads that vary in time
r Loads imparted to the hull by reciprocating or un-
with periods ranging from a few seconds to several balanced rotating machinery
minutes, and therefore occur at frequencies that are
r Hydroelastic loads resulting from the interaction of
sufficiently low, compared to the frequencies of vibra- appendages with the flow past the ship
tory response of the hull and its parts, that there is
r Wave-induced loads primarily due to short waves
no appreciable resonant amplification of the stresses whose frequency of encounter overlaps the lower natu-
induced in the structure. The loads are called dynamic ral frequencies of hull vibration and which therefore may
because they originate mainly in the action of the waves excite an appreciable resonant response, termed spring-
through which the ship moves, and therefore are always ing.