The surface tension of molten Zircaloy (Zry–2 and Zry–4), a Zr–1 at% Sn model alloy, and pure zirconium with varying hafnium content was measured using the oscillating droplet method with electromagnetic levitation in atmospheres with controlled oxygen partial pressure (𝑃𝑃O2 ). The surface tension of molten zirconium was not significantly affected by the presence of hafnium impurities in the 0.05–2.3 at% range. When the 𝑃𝑃O2 was buffered at low levels (≤ 10−4 Pa) using Ar He–H2–CO2 gas, oxygen gradually dissolved into the molten Zr–1 at% Sn alloy, resulting in a time-dependent decrease in surface tension. In contrast, under Ar–He gas (𝑃𝑃O2 ≈ 10−2 Pa), both oxygen content and surface tension remained nearly constant over time due to kinetic limitations in gas-phase transport. The surface tension of molten Zry–2 and Zry–4 decreased linearly with increasing temperature and was nearly identical to that of the molten Zr–1 at% Sn alloy under the Ar–He atmosphere. These values were consistently lower than those of pure molten zirconium. These experimental results, along with Butler model analysis, indicate that tin plays a dominant role in reducing surface tension through surface segregation. The maximum expanded uncertainty was within ±1%, confirming the reliability of the measurements.

View full abstract

A microgravity experiment to elucidate the cool flame dynamics in a multi-droplet system was conducted using the TEXUS-60 sounding rocket. The cool flame position was successfully observed through formaldehyde chemiluminescence. This allowed the study of cool flame propagation speed along n-decane droplet arrays, with droplet spacings of 8 mm (nine droplets) and 16 mm (five droplets). The spread speed was analyzed at an ambient temperature of 570 K and a pressure of 0.1 MPa, using air as the ambient gas. For both droplet configurations, the cool flame occurred from nearly the same position 4.7 s after the droplet was inserted and settled at the combustion position. The cool flame spread speeds decreased from around 250 mm/s to approximately 100 mm/s after the spontaneous ignition, and then increased. The maximum speed reached around 600 mm/s for the 16 mm droplet spacing case. Under these experimental conditions, the results suggest that in the first stage the cool flame spread follows the premixed propagation mode, while in the latter stage the acceleration may be driven by the sequential spontaneous ignition.

View full abstract

An upgraded partial-gravity generator based on the slope-sliding method proposed in our previous work (Nakamura and Sekimoto, 2014) was developed. This generator consists only of a sliding rack and variable-angle base with a low-friction surface, and the experimental rack is launched from the bottom of the slope. During the sliding operation, the rack experiences partial gravity in the normal direction of the sliding surface. The time for which partial gravity is achieved depends on the length of the slope: in the present test, it was at most 1.0 s, which is sufficient to capture the dynamic response of a flame to the gravity level. To check the performance, the g-sensor signals and behavior of a flame were analyzed. The precision achieved by this method was within 0.05G, which is quite satisfactory. Flame-image analyses revealed the expected features of partial gravity, suggesting that the present methodology is an effective and convenient means of realizing a partial-gravity environment.

View full abstract