Thermo-mechanical contributions to pellet-clad interaction (PCI) in advanced gas-cooled reactors (AGR) are modelled in the ABAQUS finite element (FE) code. User supplied subroutines permit the modelling of the non-linear behaviour of AGR fuel through life. Through utilization of ABAQUS's well-developed pre-and post-processing ability, the behaviour of the axially constrained steel clad fuel was modelled. The 2D axisymmetric model includes thermomechanical behaviour of the fuel with time and condition dependent material properties. Pellet cladding gap dynamics and thermal behaviour are also modelled. The model treats heat up as a fully coupled temperature-displacement study. Dwell time and direct power cycling was applied to model the impact of online refuelling, a key feature of the AGR. The model includes the viscoplastic behaviour of the fuel under the stress and irradiation conditions within an AGR core and a non-linear heat transfer model. A multiscale fission gas release model is applied to compute pin pressure; this model is coupled to the PCI gap model through an explicit fission gas inventory code. Whole pin, whole life, models are able to show the impact of the fuel on all segments of cladding including weld end caps and cladding pellet locking mechanisms (unique to AGR fuel). The development of this model in a commercial FE package shows that the development of a potentially verified and future-proof fuel performance code can be created and used.
Keywords: Fuel Performance, ABAQUS, Finite Element, AGR
Introduction
Pellet-cladding interactionNumerous authors have investigated the phenomenon of PCI in the past [1][2][3][4][5][6][7][8][9][10]. These studies have included the AGR system. The AGR studies from the early 90s offer insight into the effective bonding forces between AGR cladding and fuel, albeit that these studies were for a very simplified geometry and material properties. Walker's models were two dimension r-T segments of a pellet and cladding in a coupled thermo-mechanical transient analysis. This analysis showed that the expected bond between pellet and clad will fail and cause the debonding of about 6% of the pellet-clad surface producing sufficient stress relief to prevent further debonding. Walker also shows that a failure of this size was sufficient to cause a two orders of magnitude reduction in the number of pellet cracks to propagate in the adjacent cladding. Most work in recent years has focussed on water cooled systems [2][3][11][12][13]. Marchal [11] showed through 3D modelling of PCI, in the CAST3M FE code, that the while pellet fracture did produce stress concentrations in the cladding there was also a significant stress remaining in the fuel fragment. These pellet fragments had sufficient remaining stresses that later interaction with the cladding through friction contact produced further radial cracking. Marchal's smeared crack model emphasises the non-linear behaviour of PCI cracking and that it is strongly dependent on fuel-cladding friction.
*Manuscript Click here ...