Validations, Verifications and Applications of the FEAT Code

Abstract

The FEAT code (finite-element analysis for temperature), a general-purpose two-dimensional finite-element computer code for heat-flow calculations in solids of arbitrary shapes, is frequently used in thermal design and assessment of CANDU fuel. Examples of applications include high-burnup pellets (two-dimensional heat conduction in short pellets with big chamfers), effect of end flux peaking on peak pellet temperature, graphite disc fuel (two dimensional heat transfer that is due to the existence of a highly conductive graphite disc between neighbouring pellets), pellet bottoming during load-following, sheath-bearing pad heat transfer analysis for designing a bearing pad that eliminates crevice corrosion, and effect of pellet grooves for instrumentation and gas storage.The FEAT code models both steady-state and transient two-dimensional heat conduction with internal heat generation; with user-specified boundary conditions (e.g.. prescribed boundary heat convection, prescribed boundary heat fluxes and prescribed boundary temperatures); with variable material properties such as temperature dependence of thermal conductivity, specific heat and density (nonlinear heat conduction); and with gaps between different materials (heat conduction in multiple bodies). A detailed validation and verification of the FEAT code was recently done using the validation matrix approach, which included the following activities: 1. creating validation matrices that include scenario-to-phenomenon table that specifies phenomena expected to occur during scenarios. and phenomenon-to-data set table that associates the phenomena to data sets that can be used to validate the modelling of phenomena; identifying all the features in the FEAT code that need to be tested; searching for cases to test features; forming a test matrix that consists of 40 test cases based on the phenomena modelled, for example, the convergence test cases, steady-state cases, transient cases, etc. All the features of the FEAT code were covered by the test matrix. 2. finding independent solutions for all the test cases from analytical solutions, other codes and experimental measurements. 3. comparing the FEAT predictions with the independent solutions for each of the test cases.This paper describes the results from this study. As well, some illustrative examples are given.The differences between FEAT predictions and results from analytical solutions or other independent codes are generally within 3.0 %. This result shows that the FEAT code correctly handles the fundamentals of heat transfer that the code simulates. Isotherms calculated by FEAT are consistent with a number of experimental observations including tear-drop-shaped voids that are due to end flux peaking, grain growth profiles in graphite disc fuel and in grooved pellets.and measured temperatures in the sheath with bearing pad. Also, the study confirms that the FEAT code converges rapidly to the true solution, both in space and in time.