NASA Probes the Drops That Slip Past Ice Protection in Its Icing Research Tunnel

When a Drop Breaks the Design Assumption

When aircraft designers plan for ice, they usually plan for very small water droplets. Most icing clouds are made of drops between 2 and 100 microns across, small enough that a wing's built-in protection systems are designed around them. But some clouds hold something different: water drops that are far larger than the design assumption, very cold, and still liquid. When a plane flies through them, those drops can behave in ways the standard models do not fully capture, freezing in places the aircraft's protection does not reach.

On October 5, 2026, NASA published results of a test campaign at the Glenn Research Center in Cleveland aimed at that exact problem, known as supercooled large droplet icing. The tests took place in June in the agency's Icing Research Tunnel, and NASA describes them as an important milestone for its Subsonic Flight Demonstrator project. The work is about measuring instruments as much as ice: NASA is building and calibrating the tools needed to generate and measure artificial icing clouds that reproduce large-drop conditions, so that industry's engineering tools can be checked against real physics.

The primary NASA announcement, an image article by Sarah Mann of NASA Armstrong Public Affairs, explains the hazard this way:

What NASA Said

The problem, known as "supercooled large droplet icing," occurs when aircraft fly through clouds containing unusually large drops of water that are very cold but remain liquid. When a plane passes through those drops, they can rapidly freeze to surfaces that are not protected from ice buildup.

Sarah Mann, NASA Armstrong Public Affairs Specialist, in "NASA Testing Aims at Supercooled Large Droplet Aviation Safety," published by NASA on October 5, 2026, at nasa.gov.

Supercooled Water, Explained: Why Size Changes the Physics

To understand why a bigger drop is a different problem and not just a bigger version of the same problem, it helps to start with what supercooled water is. Pure liquid water does not automatically freeze the moment it reaches 32 degrees Fahrenheit. In many clouds, droplets cool below freezing but remain liquid because there is nothing for ice to start growing on: no dust particle, no ice crystal, no surface roughness to act as a seed. This is a liquid below its freezing point, which is what scientists call supercooled. It is the same physics as freezing rain at ground level. If you have ever walked through a freezing rain event, you have encountered supercooled droplets; they freeze when they hit the cold pavement, your coat, or a power line, because that contact gives them the surface they were waiting for.

At altitude, the same droplets become an aviation problem because of size. Aircraft are designed, per NASA's article, for typical icing conditions with droplets from 2 to 100 microns in diameter, and for scale, a human hair is about 70 microns wide. In rarer clouds, supercooled drops can reach up to 2,000 microns, roughly the width of a grain of sand. That is a hundredfold range difference at the top end, and size changes physics.

A small droplet hitting a heated wing leading edge freezes where it lands, and an ice protection system designed for that spot handles it. A large droplet does something different: NASA says these much larger drops can hit or splash to the aft of an aircraft, including areas behind conventional ice protection systems. In plain terms, a large drop can break apart on impact and throw liquid water backward onto parts of the wing that were never designed to be deiced, because the designers assumed the water would stay where it landed. The aircraft's built-in protection is real, but it protects the wrong place for this kind of water.

This is a rare condition compared with ordinary icing clouds, and NASA describes it as a rare but persistent hazard. Rare is not the same as negligible: supercooled large droplet icing has been a known area of aviation safety study for decades, and it is one reason icing certification rules include large-drop conditions. NASA's article does not claim any new accident or incident, and this article does not make that claim either. What the article documents is a laboratory step: making the conditions measurable so engineers can design against them.

Why the Icing Research Tunnel Needs New Instruments

The aviation industry relies on engineering tools and models to design aircraft without having to fly prototypes through every dangerous cloud. NASA's article states that current tools work well for typical clouds, but that engineers have questions about how well those tools account for the physics of supercooled large drops. If the model is unsure about the physics, the certification analysis built on it inherits that uncertainty.

The way to answer such questions is to produce the conditions on demand, measure them carefully, and compare. That is what the Icing Research Tunnel at Glenn Research Center is for. It is a wind tunnel that manufactures icing clouds, using water sprays to create droplets of controlled sizes in a cold airstream, so researchers can expose test articles to repeatable icing conditions. For large-drop work, the tunnel itself needs upgrading: NASA states it is enhancing the equipment used to generate and measure experimental clouds in the tunnel. Before you can check a design tool against large drops, you need confidence that the cloud you made really contains the drop sizes you intended.

That measurement challenge is where the June campaign concentrated. NASA researchers tested new probes that calibrate the sizes of the drops in the tunnel's clouds. The new probes can "see" drops larger than 45 microns and perform a real-time analysis of their sizes, according to the NASA article. There is a division of labor in the plan: NASA says the results from these larger-drop probes will be mated to results from a different probe that measures droplets smaller than 45 microns, so that the complete droplet size spectrum of the tunnel cloud is known end to end. The article also references a laborious technique of post-processing droplet size image data from the tunnel, indicating that the real-time and post-processed measurements are meant to be compared against each other as a validation step.

The Subsonic Flight Demonstrator Connection

The campaign is a milestone for the Subsonic Flight Demonstrator project, part of NASA's Integrated Aviation Systems Program under the agency's Research and Technology Mission Directorate, and part of the broader effort to mature key airframe technologies such as new wing designs for the next generation of single-aisle airliners. Per NASA's project page, the SFD project exists to engage with industry and other government organizations to identify, select, and mature airframe technologies with a high probability of transitioning to the next-generation single-aisle-seat class airliner. New wing designs, including the truss-braced and thin-wing concepts NASA has been pursuing with industry partners, change where water lands and how it flows, which makes accurate large-drop characterization part of the safety case for those designs.

The connection is practical rather than rhetorical. If the next generation of airliners uses wings shaped differently from today's, their ice protection must be validated against the same hazard conditions, including supercooled large droplets. NASA's measurements of what its tunnel actually produces, drop by drop, feed the datasets that industry and regulators use for that validation.

What Is Verified and What Is Still Open

Some boundaries on what this article can claim are worth stating plainly. The tests ran June 8 through 11, 2026, in the IRT. The probes were tested and the data was collected. Detailed analysis of the collected data continues, per NASA, and the project team will share results with the broader aerospace community once complete. NASA has not published the analyzed results yet, has not stated a completion date, and has not stated when the enhanced tunnel capabilities will be fully certified for routine large-drop testing. Any statement about when industry will receive usable datasets, or when design tools will be updated, would be speculation.

It is also an observed fact, not a conclusion, that the probes "see" large drops in real time; whether that capability survives detailed validation against the post-processed image data is precisely the comparison NASA says it is undertaking. Readers should treat NASA's milestone language as the agency's assessment of its own program progress, which is a forward-looking framing typical of project communications, and treat the instrument specifications as claims that the continuing analysis is meant to confirm.

The Bottom Line for Readers

For a nontechnical reader, the core of this story is a mismatch between a design assumption and nature. Engineers assumed water arrives in drops up to about 100 microns, and built protection for that water. Nature occasionally delivers drops up to 2,000 microns, cold enough to freeze on contact and large enough to splash past the protected zones. Closing that gap does not start with a new airplane. It starts with a wind tunnel that can produce exactly the right cloud and instruments that can prove, drop by measured drop, what that cloud contains.

That is the step NASA reports as complete in principle and continuing in analysis. Once the drop size spectrum of the tunnel's experimental clouds is fully known, the numbers become the reference against which industry's engineering tools can be checked. If the tools and the tunnel agree, designers gain confidence. Where they disagree, the disagreement points to physics that needs better modeling, which is the kind of specific, actionable result a shared dataset can produce for the whole aerospace community.