Why does a floating droplet freeze differently from a puddle on Earth?
On Earth, gravity does two things that dominate freezing: it drives natural convection inside the liquid, and it pulls the liquid against a container wall. In a free-floating droplet in microgravity, both are largely removed. The China Space Station experiments maintained a long-term stable microgravity condition of about 10⁻⁵ g₀, which is roughly one hundred-thousandth of Earth's gravity [1]. Under those conditions, strong natural convection inside the liquid alloy is suppressed, so the solidification process is no longer masked by gravity-driven stirring [1].
The absence of container walls matters too. In a container, the walls act as surfaces where crystals can nucleate and where liquid can be drawn from as the solid shrinks. A containerless droplet has no such reservoir. The paper on solidification shrinkage describes this as a fundamental alteration: both the suppression of natural convection and the absence of container walls change how the solidifying alloy contracts and how liquid flows to fill the gaps [4].
What does the frozen droplet actually look like inside and on its surface?
When a liquid alloy freezes, it contracts. In a free-floating droplet, that contraction can leave behind internal voids called freezing shrinkage cavities [1]. The 2026 study quantified this by comparing two competing effects: the volume contraction from the liquid-to-solid phase transition and the ability of the remaining liquid to flow and fill the shrinking regions. By comparing these, the researchers built a dimensionless criterion that predicts where shrinkage risk is high, and it matched their experimental observations across different undercooling levels [4].
The surface also develops a distinctive pattern. The microgravity solidification produced surface dendritic microstructures — branching, tree-like crystal growths — that grew not only along the tangential direction (parallel to the surface) but also along the normal direction (perpendicular to the surface) [1]. The researchers attribute this combined growth direction to the interplay between shrinkage dynamics and liquid surface flow in the space environment [1]. In other words, as the droplet contracts and liquid moves across its surface, the dendrites are pushed or drawn in more than one direction as they grow.
How cold can the liquid get before it freezes, and does that change the outcome?
A liquid can be cooled below its normal freezing point without solidifying — a state called undercooling. In the space station experiments, the refractory alloys Nb₈₂.₇Si₁₇.₃ and Zr₆₄V₃₆ reached substantial undercoolings far from a thermodynamically stable state, at temperatures up to 2265 K [1]. That means the liquid was held well below its equilibrium freezing temperature before solidification began.
The degree of undercooling matters for shrinkage. The shrinkage criterion developed in the 2026 study successfully predicted shrinkage risk across different undercooling levels, suggesting that how far the liquid is cooled before freezing influences how much void forms and where [4]. The experiments combined measured thermophysical properties with calculated internal flow velocities to quantify both the contraction and the liquid's ability to feed the shrinking regions [4].
What happens if a liquid is suddenly exposed to the vacuum of space?
A separate concern is what happens when a liquid is exposed almost instantaneously to vacuum, as can occur during launch or space vehicle operations [2]. The 1964 study examined the rate at which freezing takes place under vacuum conditions, using liquid nitrogen as a test case [2]. The abstract is fragmentary, but it establishes that sudden vacuum exposure is a distinct freezing scenario relevant to space technology, separate from the slow solidification of a droplet held in microgravity [2].
This is different from the containerless alloy experiments, which were conducted under controlled microgravity with the droplet held in place rather than exposed to open vacuum [1][4]. The vacuum-exposure case involves rapid evaporation and cooling at the liquid surface, which can drive freezing from the outside in at a much faster rate. The supplied evidence does not specify how the two processes compare in detail, so the vacuum case should be treated as a related but separate phenomenon.
Sources used in this answer
Freezing Shrinkage Dynamics and Surface Dendritic Growth of Floating Refractory Alloy Droplets in Outer Space
Microgravity solidification of refractory alloy droplets aboard the China Space Station produced freezing shrinkage cavities and surface dendrites that grew both tangentially and normal to the droplet surface, attributed to combined shrinkage dynamics and liquid surface flow.
Freezing of liquids on sudden exposure to vacuum
Liquids exposed almost instantaneously to vacuum, as in launch and space vehicle operations, freeze at a measurable rate, studied using liquid nitrogen.
Computer simulations of freezing and supercooled liquids
Computer simulations of freezing and supercooled liquids address the thermodynamics and kinetics of phase transitions but do not address microgravity or space conditions.
Space manufacturing-oriented prediction and control strategy of solidification shrinkage for ultrahigh temperature alloys
Containerless solidification experiments on the China Space Station quantified phase-transition volume contraction and liquid feeding capability, yielding a dimensionless shrinkage criterion that predicts local shrinkage risk across undercooling levels.
Structural precursor to freezing in the hard-disk and hard-sphere systems
Model hard-disk and hard-sphere fluids show a structural precursor to freezing, with the liquid structure factor near 2.85 at the freezing line, but the study does not address microgravity or space environments.