Photo credit: www.sciencedaily.com
New Method to Simulate Turbulent Systems Developed by Oxford Researchers
A team of researchers at the University of Oxford has developed an innovative approach for simulating turbulent systems using probability-based methods. This groundbreaking research was published on January 29 in the journal Science Advances.
The challenge of predicting turbulent fluid flows has been a persistent goal for scientists and engineers. Despite advancements in computational technology, accurately simulating turbulent flows—except in the simplest cases—remains virtually impossible.
Turbulence is characterized by a complex interplay of eddies and swirls of varying shapes and sizes, which interact in unpredictable ways. This chaotic nature complicates modeling efforts, rendering even the most advanced supercomputers ineffective in simulating these fluctuations for applications in engineering or weather forecasting.
In collaboration with researchers from institutions in Hamburg, Pittsburgh, and Cornell, the Oxford team approached the problem by avoiding the direct simulation of turbulent fluctuations altogether. Instead, they conceptualized these fluctuations as random variables governed by a specific probability distribution function. By simulating these distributions, they successfully extracted essential quantities related to fluid flows, such as lift and drag, without needing to navigate the complexities of turbulence directly.
Traditionally, simulating turbulence through probability distributions involves solving high-dimensional Fokker-Planck equations. This task is typically infeasible with classical computational methods. However, the researchers harnessed a quantum-inspired computing technology pioneered at Oxford. This technique utilizes ‘tensor networks’ to represent turbulence probability distributions in a highly compressed format, facilitating more effective simulations.
In their study, the quantum-inspired algorithm, executed on a single CPU core, accomplished computations in mere hours—work that would typically require several days on a conventional supercomputer.
While this speed increase is significant, researchers believe that future advancements in dedicated computing hardware, such as tensor processing units and fault-tolerant quantum chips, could lead to even more substantial improvements.
According to the research team, this novel approach challenges the existing boundaries of turbulence simulation and may pave the way for modeling other chaotic systems that can be described through probabilistic methods.
Dr. Nikita Gourianov, the lead researcher from the Department of Physics at the University of Oxford, stated, “The demonstrated—and future—computational advantages not only grant access to previously uncharted territories of turbulence physics but also inspire the development of next-generation computational fluid dynamics codes. These innovations could enhance weather forecasting, improve vehicle aerodynamics, boost efficiency in the chemical industry, and much more.”
Source
www.sciencedaily.com