On the bottom floor of a University of Chicago research laboratory, there is a tank of water that, several times a week, spontaneously contains a ball of pure chaos. The ball is not large, roughly the size of a small watermelon. It hangs, churning, in the middle of the tank without touching the sides, without being stirred by any paddle, and without drifting. Physicists call it The Blob. Creating it requires firing eight precisely choreographed water jets from the corners of the tank simultaneously, each one launching a vortex ring inward so that all eight rings arrive at the center at exactly the same moment and collide. What happens next is one of physics’ oldest unsolved problems, finally able to be watched in something like peace.
Turbulence is everywhere and almost nowhere understood. It governs ocean currents, hurricane clouds, blood moving through arteries, cream curling into coffee. It has resisted clean theoretical treatment for over a century, longer, really, than almost any other everyday physical phenomenon.
The difficulty, at bottom, is that you cannot study turbulence without making it, and making it always means interfering with it. Stick a paddle into a tank and the paddle changes how the fluid moves. Blow air through a pipe and the pipe’s walls shape the flow. Every attempt to isolate the phenomenon introduces the very complications you are trying to strip away. William Irvine, a physicist at UChicago who has worked on turbulence for years, framed it as a question that had never had a proper setting: what does turbulence actually do, he asked, when you simply let it loose?
The Blob is that clean setting. Or the closest thing to one yet achieved.
Irvine and then-graduate student Takumi Matsuzawa first created it about three years ago, and the system is, by experimental physics standards, almost deceptively simple-looking: a tank, eight vortex generators, a high-speed camera, tracer particles suspended in the water to catch the light. But what it produces is something researchers had genuinely not had before: a stationary, bounded-by-nothing ball of turbulence that could be watched from birth to death at whatever temporal resolution the camera allowed. The results, published in February in the Proceedings of the National Academy of Sciences, include several surprises about how turbulence spreads and how it dies.
The Spreading Problem
The first surprise concerned the shape of turbulence as it expands. Most people’s intuition about diffusion, trained on tea steeping or dye dissolving, is that things spread outward in a smooth gradient, becoming gradually less concentrated with distance. Turbulence does not do that. Instead, The Blob expands with a sharp front, a steep boundary between the churning region and the still water beyond it, almost like a slow-moving wall. The turbulent region and the quiet region do not blur into each other; they stay, for a while, emphatically separate. This effect had been glimpsed once before, in the 1990s, in experiments on superfluid helium conducted partly by theorist Nigel Goldenfeld, now at UC San Diego and a co-author on the new paper. But water is not superfluid helium, and the earlier experiments lacked the imaging resolution to trace the mechanism in detail. Watching The Blob with modern particle-image velocimetry, Matsuzawa and Irvine confirmed the same sharp-front behaviour in ordinary water for the first time, with enough spatial and temporal resolution to actually characterise what is happening.
Irvine noted that no other setup gives you turbulence “separated from the walls,” with its properties determined not by the container but by the method of creation.
Two Deaths in One Tank
The decay findings turned out to be stranger still. As The Blob wound down, the team tracked how its energy dissipated over time, and they noticed something odd: the energy did not drop according to a single consistent mathematical pattern. It dropped in one way at first, then switched to a different regime later on. Two distinct power laws, in sequence, in the same tank. To check whether this was some quirk of the blob-creation method, Matsuzawa set up a different approach: he placed a plastic mesh in the water and oscillated it to generate turbulence throughout the chamber at once. When that turbulence decayed, it followed a single law the whole way down. The difference, the team eventually worked out, came from the starting size of the eddies. When the blob is first made, its largest swirling structures are roughly as big as the blob itself, maybe a palm’s width across. As it expands, those eddies grow until they fill the whole container. By contrast, the grid method produces eddies that are already as large as the tank from the very beginning. “What this shows,” Matsuzawa said, “is that you can have two different laws of decay of turbulence in the same box.” Which sounds almost paradoxical until you realise that the box is not really what matters; what matters is how the turbulence started.
The team’s theoretical framework extends classical work by the Soviet mathematicians A. N. Kolmogorov and G. I. Barenblatt, whose model of freely decaying turbulence has been around for decades but never quite verified in a clean experiment. Minhui Zhu, then a graduate student at the University of Illinois at Urbana-Champaign who handled much of the theoretical side, was struck by how much explanatory power a fairly minimal model could provide: “Even a single isolated blob of turbulence is a complex system. It’s amazing that a minimal theoretical picture can still capture the essential behaviors observed in the experiment.” The same analysis, she noted, ruled out several previously proposed theories, which is perhaps just as important as confirming the one that survived.
A Footprint That Outlasts the Chaos
Among the subtler findings is this: the turbulent cascade, the process by which energy flows from large eddies down through progressively smaller ones until it is finally lost as heat, leaves what the paper calls an “indelible footprint” in the fluid long after the turbulence itself has quieted. The fluid does not simply return to the state it was in before. Something of the cascade persists, detectable in the structure of the flow for far longer than intuition would suggest. In the experiment, Matsuzawa tracked particles in the tank for up to seventeen minutes after The Blob was created, and even at the end of that period the water had not quite settled. The ghost of the turbulence lingers. Goldenfeld, reflecting on the seven-year arc from initial conception to publication, described the work as an example of “how fundamental aspects of one of the most complex physical phenomena can be explored scientifically through innovative experiments and imaginative theory.” He also acknowledged, in a way that feels honest rather than ceremonial, that the collaboration required scientists “to improvise and persevere on a very challenging problem.”
The practical stakes are not small. Turbulence modelling underpins the design of aircraft, wind turbines, jet engines, and the plasma-confining chambers of fusion reactors, where swirling instabilities are among the central obstacles to generating useful power. Better theoretical foundations for how turbulence decays and spreads could propagate outward into any of those fields. For now, though, the team is content to keep summoning The Blob and watching what it does. There are more questions than answers still, which is roughly as it should be.
Frequently Asked Questions
What exactly is “The Blob” and why is it useful for studying turbulence?
The Blob is a ball of turbulence created at the centre of a water tank by firing eight precisely timed vortex rings inward from the tank’s corners. Because the turbulence forms away from the walls and without any paddle or obstacle maintaining it, it evolves freely without interference from boundaries. That makes it about the cleanest experimental setting researchers have yet managed for watching turbulence in its natural, unforced state.
Why has turbulence been so hard to understand compared with other fluid phenomena?
Creating turbulence to study it always involves some kind of intervention, and that intervention changes what you see. A paddle, a pipe wall, a pump, a grid: all of these imprint their own structure onto the flow and make it harder to extract the intrinsic behaviour of turbulence itself. Most laboratory experiments have also struggled to isolate turbulence from boundary effects, which is why The Blob’s wall-free design represents a meaningful step forward.
What does it mean that turbulence leaves a “footprint” after it decays?
As turbulent energy cascades down from large swirling structures to smaller and smaller ones before dissipating as heat, it leaves a signature in the structure of the fluid that persists even after the visible churning has stopped. In the experiments, water that appeared to have settled still showed evidence of the earlier turbulent cascade for up to seventeen minutes. The fluid essentially remembers, in a statistical sense, the way it was disturbed.
Could this research affect the design of fusion reactors?
Potentially, yes. Turbulence is one of the central obstacles in magnetic confinement fusion, where swirling plasma instabilities cause energy to leak out of the containment field. Better theoretical foundations for how turbulence spreads and decays, even in ordinary water, could help inform models used to design and control plasma behaviour in tokamaks and similar devices. The physics principles translate, even if the working fluid is very different.
Why did the turbulence in the blob decay according to two different mathematical laws, while grid-generated turbulence followed only one?
The difference comes down to the initial size of the eddies. When the blob first forms, its largest swirling structures are roughly the size of the blob itself, and they grow as the turbulence expands to fill the tank. That growth produces an early decay regime that eventually transitions to a second one once the eddies have reached the container walls. Grid-generated turbulence, by contrast, immediately fills the tank with large eddies, so there is no transition: only a single decay pattern from the start.
ScienceBlog.com has no paywalls, no sponsored content, and no agenda beyond getting the science right. Every story here is written to inform, not to impress an advertiser or push a point of view.
Good science journalism takes time — reading the papers, checking the claims, finding researchers who can put findings in context. We do that work because we think it matters.
If you find this site useful, consider supporting it with a donation. Even a few dollars a month helps keep the coverage independent and free for everyone.