Page 356 - Practical Ship Design

Freeboard and Subdivision 313

resulting in the flooding of one, two or any number of adjacent compartments and
of penetrating or not penetrating longitudinal bulkheads and watertight decks or
flats. The probability of the ship having sufficient residual buoyancy and stability
to survive in each case of damage is assessed and the summation of all positive
probabilities gives an “attained Subdivision Index” which must be greater than a
“Required Subdivision Index” which is based on ship’s length and complement for
passenger ships and on ship’s length only for cargo ships.
Some of the pros and cons of deterministic and probabilistic methods can be
illustrated by a few examples. These are based on the passenger ships rules, as of
course there are no deterministic rules for cargo ships.
In the deterministic rules the statutory maximum transverse extent of damage is
set at B/5 measured inboard at any point from the half breadth at the subdivision
waterline. This makes the precise positioning of a longitudinal watertight bulkhead
critical: if it is inboard of the B/5 criterion by a few millimetres the space inboard of it
becomes “intact” buoyancy for the purpose of considering compliance with the
prescribed damage stability criteria and may also be used to increase the permissible
length; if it is outboard of the B/5 criterion then from a subdivision point of view it
does not exist, although the rules require that damaged stability calculations be done
on the assumption it is not breached if this results in a more onerous situation.
Under the probabilistic regulations the precise positioning ceases to matter and
damage calculations are made with and without the bulkhead being breached.
The probabilistic rules take a somewhat similar approach to the water tightness
of decks with calculations being made on the alternative assumptions that these are
and are not breached.
The probabilistic method takes into account not only safety against flooding as
in the deterministic method, but also safety against capsize. In doing this the
probabilistic rules takes account of the fact that it is usually the damaged stability
that is the ultimate determining factor in the deterministic approach. A ship can
pass the subdivision rules but fail on damaged stability, but a ship meeting the
damaged stability requirements will also pass the subdivision ones.
The arguments clearly favour the rational of the probabilistic rules and it is thought
likely that the deterministic rules will be phased out in the foreseeable future.
There would seem to be two main objections to the probabilistic rules. The first
of these is the extremely large amount of calculations required, which although
acceptable in this computer age is scarcely to be welcomed. The other objection is
the lack of guidance that it gives to a designer, who may even be driven to
continuing use of the deterministic method in initial design, changing to the
probabilistic later - and hoping this does not entail major changes! It must be
admitted that although the design guidance given by the deterministic rules eased
the designer’s task, some of its features, as instanced above, did not lead to
optimisation in the interests of safety.