When placed on a vibrating oil bath, droplets have many wild behaviors, some of which mirror quantum mechanics. Even big droplets — bigger than 2 millimeters in diameter — can get in on the fun. This video shows several of these “jumbo superwalkers” in action, both singly and in groups. (Video and image credit: Y. Li and R. Valani; via GFM)
Glacier-fed rivers are often rich in colorful sediments. Here, photographer Jan Erik Waider shows us Iceland’s glacial rivers flowing primarily in shades of blue. While the wave action and diffraction in these videos is great, the real star is the turbulent mixing where turbid and clearer waters meet. Watch those boundaries, and you’ll see shear from flows moving at different speeds which feeds the ragged, Kelvin-Helmholtz-unstable edge between colors. (Video and image credit: J. Waider; via Laughing Squid)
For decades, biologists have focused on genetics as the key determiner for biological processes, but genetic signals alone do not explain every process. Instead, researchers are beginning to see an interplay between genetics and mechanics as key to what goes on in living bodies.
For example, scientists have long tried to unravel how an undifferentiated blob of cells develops a clear head-to-tail axis that then defines the growing organism. Researchers have found that, rather than being guided purely by genetic signals, this stage relies on mechanical forces–specifically, the Marangoni effect.
The image above shows a mouse gastruloid, a bundle of stem cells that mimic embryo growth. As they develop, cells flow up the sides of the gastruloid, with a returning downward flow down the center. This is the same flow that happens in a droplet with higher surface tension in one region; the Marangoni effect pulls fluid from the lower surface tension region to the higher one, with a returning flow that completes the recirculation circuit.
The same thing, it turns out, happens in the gastruloid. Genes in the cells trigger a higher concentration of proteins in one region of the bundle, creating a lower surface tension that causes tissue to flow away, helping define the head-to-tail axis. (Image credit: S. Tlili/CNRS; research credit: S. Gsell et al.; via Wired)
Back in May 2020, the Edenville Dam in Michigan failed dramatically, releasing flood waters that destroyed a downstream dam and caused millions of dollars of damage. In this Practical Engineering video, Grady deconstructs the accident, based on an interim report from the forensic team charged with investigating the failure. Along the way, he explains common causes of dam failures, what made the Edenville failure unusual, and how engineers build modern earthen dams to avoid this older design’s flaws. (Image and video credit: Practical Engineering)
As children, we’re taught that there are three distinct phases of matter–solid, liquid, and gas–but the reality is somewhat more complicated. In the right–often exotic–conditions, there are far more phases matter takes on. In a recent study, researchers described a metal that sits somewhere between a liquid and a solid.
In a liquid, atoms are free to move. During solidification, atoms lose this freedom, and their frozen positions relative to one another determine the solid’s properties. Atoms frozen into orderly patterns form crystals, whereas those frozen haphazardly become amorphous solids. In their experiment, researchers instead observed atoms in liquid metal nanoparticles that remained stationary throughout the transition from liquid to solid. The number and position of stationary atoms affected whether the final solid crystallized or not.
By tracking these stationary atoms and their influence, the team hopes to better control the material properties of the final solidified metal. (Image credit: U. of Nottingham; research credit: C. Leist et al.; via Gizmodo)
High-speed cameras are an amazing tool in fluid dynamics, but they come with a whole host of challenges. The camera and lighting have to be positioned to deal with reflections, the data sets are enormous, and post-processing all that data takes a long time.
Here, researchers experiment instead with studying a flow using an event-based camera, which records information only when and where the brightness changes. The images and videos look strange to our eyes, but, as the authors show, they work nicely for identifying flow features and extracting valuable data. (Video and image credit: D. Sun et al.)
Taken from a Cessna aircraft, photographer J. Fritz Rumpf’s image of a Brazilian landscape appears abstract. But it captures a serpentine river and surrounding dunes, dyed brown by decaying plant matter and sculpted by the forces of wind and current. This shot is part of a portfolio that won him the title of 2025 International Landscape Photographer of the Year. (Image credit: J. Rumpf; via ILPOTY)
Happy 2026! This will be a big year for me. I’ll be finishing up and turning in the manuscript for my first book — which flows between cutting edge research, scientists’ stories, and the societal impacts of fluid physics. It’s a culmination of 15 years of FYFD, rendered into narrative. I’m so excited to share it with you when it’s published in 2027.
As always, though, we’ll kick off the year with a look back at some of FYFD’s most popular posts of 2025. (You can find previous editions, too, for 2024, 2023, 2022, 2021, 2020, 2019, 2018, 2017, 2016, 2015, and 2014.) Without further ado, here they are:
What a great bunch of topics! I’m especially happy to see so many research and research-adjacent posts were popular. And a couple of history-related posts; I don’t write those too often, but I love them for showing just how wide-ranging fluid physics can be.
Interested in keeping up with FYFD in 2026? There are lots of ways to follow along so that you don’t miss a post.
And if you enjoy FYFD, please remember that it’s a reader-supported website. I don’t run ads, and it’s been years since my last sponsored post. You can help support the site by becoming a patron, buying some merch, or simply by sharing on social media. And if you find yourself struggling to remember to check the website, remember you can get FYFD in your inbox every two weeks with our newsletter. Happy New Year!
(Image credits: droplet – F. Yu et al., starlings – K. Cooper, espresso – YouTube/skunkay, fountain – Primal Space, Uranus – NASA, turbulence – C. Amores and M. Graham, capsule – A. Álvarez and A. Lozano-Duran, melting ice – S. Bootsma et al., puquios – Wikimedia, cooling towers – BBC, solar wind – NASA/APL/NRL, Lake Baikal – K. Makeeva, sprite – NASA, roots – W. van Egmond, sunflowers – Deep Look)
Engineering classes often discuss hydrostatics–the physics of non-moving water–before they cover fluid dynamics and its flows. But hydrostatics is plenty challenging on its own, as Steve Mould demonstrates in this video looking at how hydrostatic pressure depends on depth (and, not, as our intuition might suggest, on shape). As always, he has some nice countertop-scale demos to go with it. (Video and image credit: S. Mould)
When a fluid coats the inner walls of a cylinder, it can move downward in what’s called a collar flow. In our airways, a sinking collar flow can thicken as it falls, eventually blocking the airway completely.
In a Newtonian fluid, this thickening during motion is essentially unavoidable; any small disturbance to the fluid will make its thickness change. But in a viscoplastic fluid–one more akin to the mucus in our airways–researchers found that, below a critical film thickness, the collar flow won’t thicken to form a blockage. (Image and research credit: J. Shemilt et al.; via APS)
