A series of TROI steam explosion experiments was performed using various prototypic melts. The melt was pure zirconia, eutectic corium (70: 30 weight percent of UO 2 : ZrO 2 ) or iron-added eutectic corium. In this series, an experiment with pure zirconia, two experiments with eutectic corium, and three experiments with iron-added corium were carried out. A steam explosion was found to be somewhat related to the melt composition and an external triggering. As with most of the previous tests, zirconia melt led to a steam explosion again. It is quite certain that a zirconia melt will more than likely result in an energetic steam explosion. Meanwhile, eutectic corium led to an energetic steam explosion by applying an external trigger, but it had a weak steam spike without an external trigger. The explosivity of eutectic corium cannot be ignored, as an external trigger led to an energetic steam explosion. Iron-added corium did not lead to an energetic steam explosion. The reason for this is likely to be a relatively low melt temperature (superheat) when compared to zirconia melt or oxidic corium melt, resulting from the melting method used in the TROI experiments, an induction heating applied to a cold crucible. The iron-added corium at a low temperature was solidified easily so a steam explosion did not occur.

TROI experiments1) have been performed to reveal unsolved issues of a steam explosion by using real core material at the Korea Atomic Energy Research Institute (KAERI). One of the findings from the TROI experiments is that the results of a fuel coolant interaction (FCI) are strongly dependent on the composition of corium, which is composed of UO 2 , ZrO 2 , Zr, and steel. The TEXAS-V simulation for the TROI experiments indicated that a relatively low void fraction seems to have resulted in a strong steam explosion, and the low-voided mixture seems to have been induced by large-sized particles. The particle sizes in the nonexplosive TROI tests were analyzed because the explosive tests do not represent the particles during mixing. The analysis of particle sizes indicated that the debris size reflected the material difference, and the order of the particle size for each melt material was the same as that in the TEXAS-V simulation. The TEXAS-V calculation for the alumina/water system indicated that thermal conductivity is also related to the material effect on the FCI result. A heat loss evaluation using a single-sphere, filmboiling model showed that reliable values for thermal conductivity and particle size provide a reasonable estimation for the FCI result. The steam explosion in corium/water interactions is suppressed by a smaller particle size and an induced larger heat loss during mixing.

The suppression of a vapor explosion is reviewed from a void fraction point of view from previous research results and the results of an experiment and analysis for TROI using a prototypic reactor material. In a tin-water system, a high fraction of air which played the role of steam reduced the peak pressure of a steam explosion. According to the sensitivity analysis that was carried out with an increase in vapor volume fraction, an energetic vapor explosion hardly took place in a mixture with a high void fraction. Under higher vapor fraction conditions ( v > 0:3), the vapor explosion was very weak. The prototypic corium showed a relatively high void fraction compared to ZrO 2 , which is known as an explosive material, because this corium system generates many smaller particles compared to the ZrO 2 system. Also this corium system showed a relatively low explosivity compared to the ZrO 2 system because the high void fraction of the corium system prevents contact between water and hot melt drops in the triggering stage. When considering the experimental results for the role of air instead of steam, an air supply system to provide a high volume fraction during a premixing process can radically prevent and/or mitigate a steam explosion.

The TEXAS-V code tuned for TROI-13 was used for analyzing the parametric findings in TROI experiments. The calculations on the melt composition are relatively similar to the TROI experimental results. The water depth effect in TEXAS-V code seems to be consistent with TROI experiments in some degree. The water area effect of TEXAS-V calculations seems not to be harmonious to that in TROI experiments. This seems to indicate that TEXAS-V as 1-dimensional code or as the numerical steam explosion has a limitation on estimating area effect. Thus, TEXAS-V tuned for TROI-13 seems to have an ability to estimate the parametric effect of TROI experiments. The evaluated TEXAS-V was used for estimating the ex-vessel steam explosion load. The calculated explosion pressure and load were about 40 MPa and 75 kPa.sec, which are not much threatening level for containment integrity. 서 론 용융물-냉각수 반응은 때때로 급격한 열전달로 인한 파괴적인 충격파를 동반하는 증기폭발을 발 생시킬 수가 있다.(1) 그래서, 증기폭발에 대한 많 은 실험적 연구 및 해석적 연구가 진행되어 왔 다.

No abstract