randomly dispersed fully dense whiskers in a UO2 medium。

The thermal conductivity data on UO2 are available for var-

ious theoretical densities (TDs) [6]。 The thermal conductivity    Specimen wt% vol% % TD (microns)

curve for PIP SiC phase was constructed from one data point for 2 0。6 2。1 99。4 157

The mathematical thermal modeling of composite nuclear fuel is a key in the development of enhanced conductivity ox- ide fuels。 Finite element program ANSYS [9] has been used to develop thermal models and predict thermal conductivity of composite materials [10, 11]。 In the present study, the use of this method was expanded and applied to the modeling of com- posite nuclear fuels。 First, a thermal model was developed for the composite fuel using ANSYS, and the results were bench- marked with existing experimental data on UO2/BeO fuel。 Then the methodology was applied to UO2/SiC fuel。 The composite fuel microstructure was studied, and appropriate modification to the ANSYS thermal model was incorporated。 Thermal con- ductivity of the UO2/BeO composite fuel sample was measured with laser Flash method。 The UO2/SiC fuel thermal conductiv- ity measurements were not obtained due to problems in sample preparation。 Laser flash measurements on thermal conductivity of the UO2/BeO fuel were compared with model predictions, and these results were refined based on interface structure be- tween UO2 and BeO phases。

ANSYS THERMAL MODEL BENCHMARK

ANSYS was first benchmarked using grid pattern geome- try with a continuous high conductivity phase。 Ishimoto et al。 [12] fabricated, conducted thermal measurements, and modeled a composite of polycrystalline UO2 with an “almost” contin- uous BeO phase at the grain boundaries, and dispersed in the UO2 matrix in quantities of 1。1 to 4。2 vol %。 They found the continuous BeO phase yielded significantly higher conductivity。 The information from this paper [12] was used to benchmark the current FEM model versus measured data points for the composite。

First, the ANSYS Code was checked against closed form so- lutions on effective thermal conductivity for two phase geome- tries such as a parallel and series boundary。 Then, the Ishimoto et al。 data on UO2-BeO composite fuel [12] was used for bench- marking with ANSYS calculations。 The fabrication method em- ployed in [12] involves pellet formation from a mixture of the two components, which are then sintered above their eutectic temperature of 2423 K。 The result of holding the temperature of the pellet at 2473 K for one hour is the production of UO2 grains with a continuous phase of BeO segregated at the grain boundaries。 This lamellate type structure of BeO and UO2 re- semble the geometric configuration of the grid pattern used in previous ANSYS calculations。

The concentrations of BeO in the paper [12] include 0。6, 0。9, 1。2, and 13。6 weight percent, which produce microstruc- tures with BeO segregated at the grain boundaries of UO2 and a lamellar structure at high BeO concentrations。 These weight percentages were chosen to be compared to an equivalent AN- SYS model。 The thermal conductivity of the UO2 component [6] was corrected for the measured relative density of the spec- imens, 99。4%, 98%, 98。7%, and 99%, which was assumed to be the density of the UO2 portion of the composite。 The BeO thermal conductivity values were taken from Touloukian and Ho [13]。

The sizes of the UO2 grains are given in [12]。 The thickness of the BeO section was calculated by using the size of the UO2 grains and the corresponding total area of a grain surrounded by an even layer of the remaining area that is constituted of BeO。 Table 1 lists the volume percents, weight percents, and grain size of the different fuel pellets。