Steam explosions of single drops of pure and alloyed molten aluminum

Abstract

Studies of steam explosion phenomena have been performed related to the hypothetical meltdown of the core and other components of aluminum alloy-fueled production reactors. Our objectives were to characterise the triggers, if any, required to initiate these explosions and to determine the energetics and chemical processes associated with these events. Three basic studies have been carried out with 1–10 g single drops of molten aluminum or aluminum-based alloys: untriggered experiments in which drops of melt were released into water; triggered experiments in which thermal-type steam explosions occurred; and one triggered experiment in which an ignition-type steam explosion occurred. In untriggered experiments, spontaneous steam explosions never occurred during the free fall through water of single drops of pure A1 or of the alloys studied here. Moreover, spontaneous explosions never occurred upon or during contact of the globules with several underwater surfaces. When Li was present in the alloy, H2 was generated as a stream of bubbles as the globules fell through the water, and also as they froze on the bottom surface of the chamber. The triggered experiments were performed with pure A1 and the 6061 alloy. Bare bridgewire discharges and those focused with cylindrical reflectors produced a small first bubble that collapsed and was followed by a larger second bubble. When the bridgewire was discharged at one focus of an ellipsoidal reflector, a melt drop at the other focus triggered only very mildly in spite of a 30-fold increase in peak pressure above that of the bridgewire discharge without the reflector. Experiments were also performed with globules of high purity A1 in which the melt release temperature was progressively increased. Moderate thermal-type explosions were produced over the temperature range 1273–1673 K. At about 1773 K, however, one experiment produced a brilliant flash of light and bubble growth about an order of magnitude faster than normal; it destroyed the chamber. The exceptional vigor of this latter interaction was attributed to ignition of the metallic A1. In both the triggered steam explosions with drops of high purity A1 and untriggered experiments with drops of A1Li, only a few tens of cm3 of hydrogen per gram of metal were generated; the extent of metal-water reaction was only a few percent at most. The information obtained from these studies with triggered and untriggered drops of molten aluminum-based alloys can be used directly in certain safety studies where hypothetical core or component meltdown occurs in a production reactor. Moreover, this information is consistent with recent advances in the understanding of single-drop steam explosion triggering and concepts of the development of larger scale steam explosions.