Abstract

Pumice often floats on water, and there is a large variation in the length of time it floats. We investigated the factors that influence the flotation time of pumice from Sakurajima volcano, Japan, and how water infiltrates in the pumice. Five air-fall samples of different sizes and porosities were immersed in water and the time they floated was evaluated. The flotation time varied from almost immediately sinking to more than 10 days, and a clear correlation between flotation time and sample size was not observed within the range of 2–39 cm3. From the measured values of the total porosity, connected porosity, transport porosity, dead-end porosity, and isolated porosity of each sample, it was found that there was a strong correlation between the flotation time and total porosity, connected porosity, and dead-end porosity. In order to investigate the size of the pores and the order in which water enters them, the amount of water in each pore radius was measured under three conditions: when the sample was immersed in water for 10 min or 3 days, and when all the pores were filled with water. Samples with a low porosity sank rapidly because enough water entered the interior to lose buoyancy due to the initial rapid capillary imbibition. At this stage, water was found to enter the large-radius pores preferentially. Samples with a high porosity did not sink by initial process, but eventually sank. Water infiltration during this period was found to be more pronounced in dead-end pores than in transport pores. In the experimental method adopted in this study, water present in the corners and walls of the pores was regarded as water in the dead-end pores, and it is presumed that an increase in such water and the dissolution of entrapped air occurred in the later stage.