Comparison of total drag coefficient (CD) versus Reynolds number for a smooth surface (Plain) and DMR-coated surfaces (DMR1, DMR2). DMR achieves up to 43.6% drag reduction in the transitional regime and maintains lower drag than the smooth surface up to the highest tested Reynolds number.
For more than 80 years, a fundamental principle of fluid dynamics assumed that smoother surfaces produce less aerodynamic drag. However, a research group led by Associate Professor Aiko Yakeno at the Institute of Fluid Science, Tohoku University, has overturned this long-standing assumption. By applying Distributed Micro-Roughness (DMR) - irregular microscale surface textures - to a test model, the team achieved a world-first experimental demonstration of up to 43.6% aerodynamic drag reduction. By reducing drag in this innovative new way, we may be able to reduce fuel consumption and CO₂ emissions across aviation, automotive, marine, and rail transportation in the future.
The key to this breakthrough was the institute's 1-m Magnetic Suspension and Balance System (MSBS), one of the largest of its kind in the world. By levitating the streamlined model body (a test model that looks like a long, sideways tear drop) using electromagnetic force, the MSBS eliminates all support interference that would otherwise mask minute drag changes from the DMR. This advanced system allows for the observation of incredibly accurate, realistic free-flight conditions that could not be achieved with a traditional wind tunnel.
In addition to experimental validation in the lab, the research team used wall-resolved Large Eddy Simulation (LES) to computationally visualize the flow of air and establish a mechanism to explain the drag reduction. The results show that drag reduction is driven by suppression of skin-friction drag - a mechanism fundamentally different from the flow separation caused by the dimples of golf balls.
LES further revealed that the DMR roughness height corresponds to k⁺ ≈ 1.2-1.7 in viscous units, well below the hydraulically smooth threshold (k⁺ < 5). A surface that fluid dynamics classifies as smooth produced a dramatic 43.6% drag reduction - challenging decades of design orthodoxy. Moreover, the effect persisted up to the highest tested Reynolds number (Re = 3.6×10⁶), suggesting drag reduction may extend beyond the transitional regime into the turbulent flow domain.
"When air passes over an airplane wing, for example, it moves in a smooth flow called laminar flow and transitions into a much less ordered flow called turbulent flow," explains Aiko Yakeno, Associate Professor at Tohoku University. "By reducing this chaotic turbulent energy, we can also reduce friction drag."
In lieu of changing more complicated aspects requiring moving parts or power, a (bumpy) makeover may be the most practical way to design fuel-efficient vehicles. The team is currently working hard to elucidate the detailed friction drag reduction mechanism, with international collaboration underway with a group of researchers lead by Professor Jonathan Morrison at Imperial College London.
The findings were published in the Journal of Fluid Mechanics on May 7, 2026.
