Masters Theses

Date of Award

8-2001

Degree Type

Thesis

Degree Name

Master of Science

Major

Nuclear Engineering

Major Professor

Laurence F. Miller

Committee Members

Ronald Pevey, Lawrence Townsend

Abstract

The High Flux Isotope Reactor (HFIR), located at the Oak Ridge National Laboratory, has seen occasional cladding damage (local swelling or blistering) in the reactor control elements over the past several years. The control elements are located in an annulus between the core and reflector, and contain three regions: a "black" region containing EU2O3 dispersed in an aluminum matrix; a "gray" region with Tantalum particles in an aluminum matrix; and a "white" region (or follower) of perforated aluminum. The cladding damage has been limited to the tantalum region, and it is expected that reduction in the tantalum fraction in this segment will reduce or eliminate the potential for clad damage. The purpose of this research is to determine the extent to which the tantalum fraction can be reduced, without unacceptable impacts on core power shape. A two dimensional R-Z geometry model of the HFIR has been created for use in DORT for neutronics calculations. Weighted cross sections are generated using the SCALE package by running BONAMI, NITAWL, and XSDRN-PM. XSDRN is used for cell weighting and then region weighting of the cross sections. These cross sections are mixed into the necessary mixtures for use in DORT by GIF. These cross sections are used in the DORT R-Z model to generate flux distributions and eigenvalues for different combinations of tantalum concentration and blade position. An initial reference case has been run using the HFIRCE-3 (Critical Experiment Series 3) core to determine the bias introduced by simplifying the HFIR to an R-Z model. The modeled critical configuration yielded an eigenvalue of 0.9907. This value is used later as a value that represents criticality in the core. This bias is introduced by azimuthal asymmetries not included in the R-Z model, by approximations in cross section generation, and by various other discretization of continuous variables. Spatial power distributions, however, are not affected by the bias. A set of runs has been used to determine an approximate relationship between k-effective and blade position for a series of different tantalum concentrations. The current initial loading of 38 volume percent and the target loading of 30 volume percent are the focus of this study. For each concentration, k is calculated with the control blades at heights of 14", 17.5", 20" and 25". These values are used to calculate an approximate critical blade position for each of the concentrations. The main concern with reduced tantalum loadings is an increase in power peaking in the core. This increase could occur because the control elements must be run closer together at beginning of cycle conditions, thus pinching the power shape with the large absorber regions closer together. The limiting thermal constraint is power peaking on the core exit of the inner fuel element. Peaking factors are calculated by dividing the fission neutron production rate in any interval by the average fission rate. Fission rates are assumed to be proportional to power. The results indicate that there is no difference in power shape between the initial loadings of 38 and 30 volume percent Ta. Such a modification is feasible since the peaking is not increased. A further decrease in loading would not be feasible, as peaking increases are observed for initial loadings below 25%.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS