🔬 physics

Fractal Scaling of Moffatt Vortices in Triangular Cavity Flow

This study numerically investigates the formation and fractal scaling of Moffatt vortices in triangular cavity flow, demonstrating that the resulting vortex cascade exhibits non-integer fractal dimensions systematically linked to vortex size and intensity, while confirming robust self-similarity across different geometries and Reynolds numbers.

Original authors: Rathindra Nath Basak, Sougata Biswas, Jiten C. Kalita

Published 2026-07-24

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Rathindra Nath Basak, Sougata Biswas, Jiten C. Kalita

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world where fluids don't just flow smoothly like water in a river, but instead get trapped in tiny, spinning whirlpools that nest inside one another like Russian dolls. This is the fascinating corner of science known as fluid dynamics, specifically looking at how thick, sticky liquids (like honey or oil) move when they are squeezed into tight spaces. For decades, scientists have been fascinated by a strange phenomenon predicted by a mathematician named Moffatt: if you push a fluid against a sharp corner, it doesn't just stop; it creates an endless chain of smaller and smaller spinning eddies, each one spinning the opposite way of the one before it. Think of it like a set of spinning tops, where the biggest one spins clockwise, the next one down spins counter-clockwise, the next one clockwise again, and so on, getting tinier and tinier until they vanish into the corner. Understanding this isn't just a game for mathematicians; it helps engineers design better machines, from tiny medical devices to massive industrial pumps, because these tiny swirls affect how energy is lost and how materials mix.

Now, picture a team of researchers taking a deep dive into this swirling mystery, but with a twist: they are looking at a triangular box instead of the usual square one. In their study, they simulated a slow-moving, sticky fluid inside a triangular cavity with a moving lid on top, like a conveyor belt dragging the fluid along. Using powerful computer simulations, they watched the fluid get dragged into the bottom corner. What they found was a beautiful, cascading sequence of seven distinct vortices (swirls), labeled V1 through V7, marching down toward the pointy tip. The biggest swirl, V1, was huge, but as they got closer to the corner, each subsequent swirl was about half the size of the previous one, and they spun with dramatically less energy. The researchers confirmed that their computer model matched real-world experiments done by others, proving that these "Moffatt vortices" really do exist in this triangular shape.

But the real magic of this paper lies in how they described the shape of these swirls. Usually, we think of shapes as having simple dimensions: a line is 1-dimensional, a square is 2-dimensional. However, these swirling chains are so complex and self-similar (looking the same whether you zoom in or out) that they don't fit neatly into those simple boxes. The authors used a clever math trick called the "area-perimeter method" to measure the "fractal dimension" of these vortices. Imagine trying to measure the length of a jagged coastline; the more closely you look, the longer it gets. Similarly, these vortices have a "fractal dimension" that is a number between 1 and 2, not a whole number. The study found that the largest vortex had a dimension of about 1.23, while the tiniest, most compressed vortex near the corner had a dimension of nearly 1.78. This means that as the vortices get squeezed into the tight corner, they become increasingly complex and "jagged" in their geometry.

The researchers also tested what happens when the fluid moves faster. They cranked up the speed (represented by the Reynolds number) from a slow crawl (Re = 1) to a moderate rush (Re = 500). Even though the main swirl moved slightly to the side and changed shape a bit, the smaller, nested vortices near the corner kept their self-similar, fractal pattern. To prove this wasn't just a quirk of the triangle, they compared it to a square cavity. In the square box, they found similar chains of vortices in the corners that behaved just like the ones in the triangle.

In short, this paper suggests that these corner vortices are not just random swirls but a genuine fractal cascade—a geometric pattern that repeats itself at every scale. The authors derived a new formula that allows scientists to estimate the fractal dimension of any vortex in the chain just by knowing the size of the first one. While the study relies on computer simulations and the finest details of the smallest vortices are still hard to capture perfectly due to grid limitations, the results strongly indicate that this fractal behavior is a robust feature of how viscous fluids behave in corners, regardless of whether the container is a triangle or a square. It's a reminder that even in the most confined, sticky corners of our world, nature loves to play with complex, repeating patterns.

Technical Summary: Fractal Scaling of Moffatt Vortices in Triangular Cavity Flow

Problem Statement
This study addresses the formation, quantification, and fractal characterization of corner vortices (Moffatt vortices) in slow, viscous, incompressible flow within a triangular cavity. While the existence of an infinite sequence of counter-rotating vortices in Stokes flow near a corner was theoretically established by Moffatt and Dean and Montagnon, and experimentally visualized by Taneda, rigorous numerical characterization of these structures in triangular geometries remains limited. Furthermore, the fractal nature of these specific vortex sequences has not been previously investigated using the area-perimeter method to uncover physical insights regarding their scaling behavior. The authors also seek to determine if these self-similar structures persist beyond the Stokes regime and across different geometries (triangular vs. square).

Methodology
The research employs a numerical approach using the finite-volume-based commercial solver ANSYS Fluent (Release 21.1). The governing Navier–Stokes equations are solved in their non-dimensional, primitive variable form for a steady, two-dimensional, incompressible flow inside an isosceles triangular cavity (altitude twice the base).

Numerical Scheme: A pressure-based coupled solver is utilized. Convective terms are discretized via a second-order upwind scheme, and diffusive terms via a central difference scheme.
Grid Resolution: Four distinct grid resolutions (ranging from 2,028 to 127,308 elements) are generated to perform a grid independence study. The finest grid (element size 0.005) is used for primary flow quantification, while data from all grids are analyzed to evaluate grid-scaling effects on fractal characteristics.
Fractal Analysis: The fractal dimension ( $D$ ) of the vortex sequences is calculated using the area-perimeter method. The outermost streamlines of the vortices are approximated as semi-ellipses. The perimeter ( $P$ ) is estimated using Ramanujan's approximation, and the area ( $A$ ) is calculated geometrically. The fractal dimension is derived from the power-law relationship $P \propto A^{D/2}$ .
Validation: Results are validated qualitatively and quantitatively against Taneda's experimental flow visualization and Moffatt's theoretical predictions. Comparative analysis is also conducted with square cavity flows at various Reynolds numbers ($Re$).

Key Results

Vortex Quantification: At $Re=1$, the simulation successfully resolves a cascade of seven corner vortices ( $V_1$ to $V_7$ ) near the bottom apex. The size ( $r$ ) and intensity ( $\psi$ ) of successive vortices decrease rapidly, following geometric progressions with asymptotic ratios of $r_{n+1}/r_n \approx 0.49$ and $\psi_{n+1}/\psi_n \approx 0.0012$ . These ratios align closely with Moffatt's theoretical predictions.
Fractal Dimensions: The computed fractal dimensions for the vortices lie between 1 and 2, confirming their non-integer, fractal nature. An inverse relationship is observed: the largest primary vortex ( $V_1$ ) has the smallest fractal dimension ( $D \approx 1.23$ ), while the smallest resolved vortex ( $V_7$ ) approaches $D \approx 1.78$ . This trend is attributed to the increasing spatial confinement as the vortices approach the corner.
Scaling Law: An empirical relation is proposed to estimate the fractal dimension of any successive vortex ( $n$ ) based on the geometric scaling of the primary vortex:
$D(n) \approx 2 \left[ \frac{\log P_1 - (n-1)\log 2}{\log A_1 - 2(n-1)\log 2} \right]$
This relation is generalized to arbitrary grid resolutions. A strong linear correlation ( $R^2 \approx 0.99$ ) is found between $\log P$ and $\log A$ , confirming discrete scale invariance.
Reynolds Number Influence: The study examines $Re = 100$ and $Re = 500$. While the primary vortex center shifts toward the right wall and the secondary vortex expands, the higher-order vortices ( $V_3, V_4$ , etc.) preserve their characteristic shapes and alternating vorticity signs. The geometric self-similarity of the vortex sequences persists across these moderate Reynolds numbers.
Geometric Universality: A comparative analysis with a square cavity reveals that corner vortices in square geometries also exhibit self-similar sequences and fractal characteristics (e.g., $D \approx 1.56$ for the primary corner vortex in a square cavity). This suggests that the fractal organization is a robust feature of confined viscous flows rather than a geometry-specific artifact.

Significance and Claims
The paper claims to bridge the gap between classical fluid topology and fractal geometry by rigorously characterizing Moffatt vortices in a triangular cavity. Its primary contributions include:

Numerical Validation: Providing a high-resolution numerical confirmation of the Moffatt vortex cascade in a triangular geometry, resolving up to seven distinct eddies.
Fractal Characterization: Establishing that these corner vortices possess non-integer fractal dimensions that are systematically linked to vortex size and intensity.
Methodological Framework: Introducing a practical, empirical relation to estimate the fractal dimension of successive vortices based on the primary vortex's geometry, applicable across arbitrary grid resolutions.
Robustness: Demonstrating that the self-similar, fractal nature of corner vortices is not limited to Stokes flow or triangular geometries but persists in moderate Reynolds number regimes and square cavities.

The authors modestly note that the approach is limited by the mesh sensitivity of the area-perimeter method, particularly for the smallest, deepest vortices, which require sufficiently fine spatial resolution for reliable dimension extraction. They conclude that these findings provide a basis for future investigations into multiscale vortex structures in more complex geometries and higher Reynolds number regimes.

More like this