抄録
The KIWI-TNT experiment was analyzed with a new two-dimensional excursion code MARS to ascertain the validity of the method used in calculating the energy release in the hypothetical accident analysis on fast reactors. In calculating the total energy release, adopting the nominal temperature coefficient, the present calculation was found to underestimate the total energy release by 9 % in reference to the experimental value of 8.8× 109 J. The calculated power pulse width was 1.1 msec, which is about half that was observed in the experiment. Adopting a temperature coefficient reduced to 0.5× 10-5Δk/Δθ, the calculated energy release agrees with that of the experimental value, but then the discrepancy in power pulse width became the largest.
Assumption that the expansion occurs in the axial direction only approaches the calculated energy release quite close to the values from experiment, although it still leaves a considerable disagreement in power pulse width.
