In the decade after 2000, the amplitude of the neutron flux fluctuations in German 1300 MWe class PWR
built by Kraftwerk Union (KWU) increased and later decreased significantly. To further investigate the
hypothesis that changes in the mechanical properties of the fuel assemblies (FAs) and subsequent stronger
mechanical oscillations of fuel assemblies and other RPV internals are responsible for the observations
and to find possible excitation mechanisms, a simple mechanical model is developed. It describes
the dynamic motion response of the mechanically coupled system of Reactor Pressure Vessel (RPV), core
barrel and a row of fuel assemblies and takes reactive fluidic forces into account. To determine the components
to be considered in this model, RPV internals with an effect on neutron flux are identified in a
prior step.
The dynamic answers of the model to generic excitation scenarios and parametric studies reveal distinct
properties of the system. They indicate 1) that fluidic near-field coupling can equalize the fuel
assemblies’ reaction amplitudes within a certain region regardless of their individual stiffnesses, 2) that
the fluid must be part of the oscillation, 3) that a non-periodicity might stem from a superposition and
interaction of several oscillators, 4) that a local excitation source does not spread or synchronize over
the whole core, 5) that a neglection or separate consideration of the RPV and core barrel may be justified
in some cases. The findings altogether affirm that core-wide oscillations of the fluid flow leading to a
simultaneous oscillation of individual fuel assembly groups, possibly including bidirectional effects
between fluid and structure, in combination with the changes of the fuel assemblies’ mechanical properties,
might be responsible for the temporary increase of neutron flux fluctuations observed in PWR built
by KWU.
A variant of the mechanical model was studied in a full core investigation with coupled neutron kinetics
simulations. These simulations qualitatively reproduce important features of measurement data,
while being overall too small in amplitude to explain the observations quantitatively, which confirmed
the necessity of considering further effects.