The following is a qualitative account of
some of the details of the steel used to
construct the Sizewell 'B' reactor pressure
vessel, and how such steel behaves under extreme
stress levels. This account is based upon
published accounts of work in the field across
the world, and over the years. It is neither
exhaustive nor definitive. But it is food for
thought.
back to index
The Sizewell 'B' reactor pressure vessel is
made from "low alloy steel type A355B". Steel was
invented in the nineteenth century by reducing
and controlling the amount of carbon included
within it. The strength of steel can be much
increased by small additions of some other metals
during manufacture.
back to index
A533B steel has density 7.83 grams per cubic
centimetre, and nominal composition:-
| ingredient | atoms percent |
| Iron | 97.02 |
| Nickel | 0.52 |
| Manganese | 1.31 |
| Carbon | 1.15 |
back to index
Some impurities present in steel can have
adverse effects on the physical properties and
need to be tightly controlled below maximum
permissible values. These include hydrogen and
copper. The effects of fusion welding, neutron
irradiation, stress cycling, stress at high
temperatures, and corrosion on steel are known to
affect the composition and structure of steel
adversely. Permanent deformation at low
temperatures, known as 'cold work' also makes it harder and less ductile.
back to index
Iron melts around 1800 degrees centigrade and
on solidification and cooling it has a system of
allotropic forms. If the temperature is reduced
very slowly to room temperature it is found to
consist of tightly packed metallic grains each
with a body centred cubic 'crystalline'
structure, known as ferrite. When other elements
are present the structure is more complicated.
Carbon and iron tend to combine in different ways
in different temperature ranges, and the changes
involved in going from one stable form to another
take time to occur, so that in addition to the
chemical composition, the rate of temperature
change is an important factor in determining the
physical properties of the end product. For
example rapid cooling from a temperature of 1000
degrees centigrade produces a hard brittle
product with the structure which existed at the
high temperature. This procedure is called
quenching or hardening. Once in the hard state,
the properties can be controlled by 'soaking' the
steel at intermediate temperatures and subjecting
it to cooling at carefully controlled rates of
temperature reduction. The properties of the
steel are measured by mechanical procedures
which, because of the inhomogeneous nature of the
product, give enormously variable results. For
example, the shear strength of the steel can be
calculated theoretically by reference to the
known properties of the ferrite crystal atomic
structure. However, the value measured in tests
is only one thousandth of the theoretical value
indicating the extent to which it is weakened by
internal irregularities and disorder. The
granular structure of the steel can be examined
by the standard crystallographic polishing and
etching techniques. Fusion welding thick steel
sections together is achieved by filling a gap
between the sections with 'passes' of deposited
weld metal. For each 'pass' during which one
strip of fusion weld metal is deposited, the
adjacent parent metal is heated up to the melting
point and down again. The Sizewell 'B' vessel
cylindrical wall is 215 millimetres thick,
necessitating about sixty passes to deposit the
required amount of weld metal. The welding is
performed in sequence from both the inside and
the outside of the vessel, with the 'root' of the
inside welds cut away before application of the
outside weld. The parent steel in the vicinity of
the weld is said to contain a 'heat affected
zone', abbreviated to HAZ. The physical
properties of the weld metal need to be
controlled independently of the parent material
in order to ensure so far as possible that
failure of the weld does not occur
preferentially. Uneven cooling of steel from high
temperature causes variation in the crystal
structure within the component. Local variations
in density which result, mean that stress fields
exist within the component which, if they remain
in service, will modify the actual stress
distribution in service from that assumed in the
design procedure. It is not unusual for quite
large cracks to form during poorly conducted
temperature control.
back to start?