10th June 2026
A miniature aircraft looks like its full-sized counterpart, but the air does not know that it is supposed to behave like a scaled photograph. Shrink the wing while keeping the wind speed unchanged, and you can change the balance of forces in the flow. A beautifully made model may therefore answer the wrong experimental question.
The solution is the Reynolds number, a dimensionless quantity which compares inertial and viscous effects. For a wing, a common definition is: Re = UL/ν, where Re is the Reynolds number, U is the incoming flow speed, L is a characteristic length such as the wing chord, and ν, ‘nu’, is the fluid’s kinematic viscosity.
Consider an illustrative wing with a chord of one metre moving through air at three metres per second. Taking the kinematic viscosity as 0.000015 square metres per second gives a Reynolds number of 200,000. A one-tenth-scale model in the same air would need a speed of thirty metres per second to match it. Simply shrinking the wing while keeping the original speed would reduce its Reynolds number tenfold.
This suggests a surprisingly useful freedom. If the testing equipment cannot provide the required speed, changing the fluid or its properties may help. A water experiment can sometimes reproduce a dimensionless flow regime that would be awkward to achieve in air. What matters is the relevant combination of quantities, rather than whether the apparatus resembles the eventual machine. [1]
There are limits. Matching Reynolds number alone does not guarantee that every feature matches. Compressibility introduces Mach number, the ratio of flow speed to sound speed. Surface roughness, incoming turbulence and the relative dimensions of the test section may also matter. The important similarity conditions depend on the physical question. [2]
[1] MIT OpenCourseWare — Dimensional analysis and dynamic similarity
Leave a comment