No abstract

We have developed a multicomponent model to describe the transient plastic deformation of Zircaloy fuel sheathing during postulated loss-of-coolant accidents (LOCA). From deformation maps, we identify three creep mechanisms, which in principle, occur in all three phase fields of Zircaloy-4 (α, α + β, β): diffusional creep, dislocation creep and athermal strain. Diffusional creep (phase boundary or grain boundary sliding) is modeled in the α and α + β phase fields. Dislocation creep is modeled in all three phase fields and controlled by a work-hardening/recovery expression. Athermal strain has practical significance only at lower temperatures (< 900 K). The sum of the strain rates of the three deformation mechanisms is the total plastic strain rate. At a particular point in the stress-temperature field, one mechanism usually dominates. Microstructural changes that affect deformation are also taken into account; these are changes in grain structure, recrystallization, and the α ⇌ (α + β ) ⇌ β phase transformation. When two components are present (α + β or recrystallized and unrecrystallized), they are assumed to deform with equal rates according to the appropriate single-phase model. The individual components of the model give excellent agreement with the isothermal data on which they are based, and the entire model gives very good agreement with transient data over a range of temperatures from 700 to 1600 K, a range of heating rates from 0 to 100 K/s, and a range of strain rates from 10-5 to 10-1 s-1. Several applications of the model in evaluating the history dependence of deformation are given.

A unified, microstructural creep law which simulates the transient creep deformation of Zircaloy at temperatures above 750 K has been used to follow the interaction of diffusional and dislocation creep with changes in material microstructure (grain size, recrystallization, phase fraction and anisotropy) under loss-of-coolant accident (LOCA) conditions. Comparison of a membrane sheath model using this creep law with a large number of tube tests (>700) in an inert environment has demonstrated good predictive capability. This model has been extended to cover other aspects affecting deformation and failure of CANDU (CANada Deuterium Uranium) fuel sheaths such as: nonuniform structure distribution resulting from temperature profiles along the sheath length, oxidation of the fuel sheaths, cracking of the oxidized layers and localized straining, large strain failure, and failure by beryllium-assisted cracking.

The Advanced CANDU Reactor (ACR™)* is an evolutionary advancement of the current CANDU 6® reactor, aimed at producing electrical power for a capital cost and unit-energy cost significantly less than that of current reactor designs. The ACR retains the modular concept of horizontal fuel channels surrounded by heavy water moderator, as with all CANDU reactors. However, ACR uses slightly enriched uranium (SEU) fuel, compared to the natural uranium used in CANDU 6. This achieves the twin goals of improved economics (e.g., via reductions in the heavy water requirements and the use of a light water coolant), as well as improved safety. This paper is focused on the double-ended guillotine critical inlet header break (CRIHB) loss-of-coolant accident (LOCA) in an ACR reactor, which is considered as a large break LOCA. Large Break LOCA in water-cooled reactors has been used historically as a design basis event by regulators, and it has attracted a very large share of safety analysis and regulatory review. The LBLOCA event covers a wide range of system behaviours and fundamental phenomena. The Phenomena Identification and Ranking Table (PIRT) for LBLOCA therefore provides a good understanding of many of the safety characteristics of the ACR design. The paper outlines the design characteristics of the ACR reactor that impact the PIRT process and computer code applicability. It also describes the LOCA phenomena, lists all components and systems that have an important role during the event, discusses the PIRT process and results, and presents the final PIRT summary table.

No abstract