What is the Reynolds Number?
The Reynolds number (Re) is a dimensionless quantity that characterises fluid flow behaviour by weighing the relative importance of inertial and viscous forces. When a fluid moves past a surface or through a confined space, two competing forces emerge: inertial forces that promote turbulence, and viscous forces that dampen it.
At its core, the Reynolds number tells you which force wins. Low values mean viscosity dominates, resulting in smooth, predictable laminar flow. High values mean inertia dominates, creating eddies, swirls, and chaotic turbulent flow. This single number is dimensionless—it has no units—which makes it universally applicable across different fluids, scales, and geometries.
Engineers use the Reynolds number to scale laboratory models to real-world conditions, to predict heat and mass transfer rates, and to design efficient pipe networks and aerodynamic structures.
Laminar vs. Turbulent Flow
Laminar flow occurs when viscous forces are dominant (typically Re < 2,100 in pipes). Fluid particles move in parallel layers without mixing. The flow is predictable, smooth, and efficient—ideal for applications requiring precise control like medical IV drips or laboratory microfluidics.
Turbulent flow takes over when inertial forces dominate (typically Re > 3,000 in pipes). Fluid particles move chaotically with eddies and vortices. Mixing is excellent, which benefits chemical reactors and heat exchangers, but friction losses increase.
Transition region (2,100 < Re < 3,000) is unstable. Flow oscillates between laminar and turbulent states. Engineers typically avoid this zone in design because behaviour is unpredictable.
The exact thresholds depend on geometry. A pipe has different critical values than flow around a sphere or over a flat plate.
Reynolds Number Formula
The Reynolds number can be calculated using two equivalent formulas, depending on which viscosity data you have available:
Re = (ρ × u × L) ÷ μ
Re = (u × L) ÷ ν
Re— Reynolds number (dimensionless)ρ— Fluid density (kg/m³)u— Fluid velocity (m/s)L— Characteristic linear dimension, such as pipe diameter or object length (m)μ— Dynamic viscosity (Pa·s or kg/(m·s))ν— Kinematic viscosity (m²/s), equal to μ ÷ ρ
Dynamic vs. Kinematic Viscosity
Dynamic viscosity (μ) measures the resistance to flow caused by internal friction. It depends on temperature and is expressed in Pascal-seconds (Pa·s) or kilograms per metre-second (kg/(m·s)). Honey has high dynamic viscosity; water has low dynamic viscosity.
Kinematic viscosity (ν) is the dynamic viscosity divided by density. It accounts for how "thick" a fluid feels relative to its mass. Kinematic viscosity is useful because it eliminates density from the calculation, which is why the second formula above is so convenient when you have kinematic viscosity data.
Both quantities are temperature-sensitive. As temperature rises, viscosity falls—a key consideration when calculating Reynolds numbers for heated or cooled systems.
Common Pitfalls and Caveats
Avoid these mistakes when calculating or interpreting Reynolds numbers.
- Using the wrong characteristic length — The choice of L (diameter, width, hydraulic radius, etc.) depends on geometry. Pipes use inner diameter; flat plates use length parallel to flow. If you pick the wrong dimension, your Re value and flow regime prediction will be incorrect.
- Ignoring temperature effects on viscosity — Viscosity changes dramatically with temperature. Water at 20 °C has μ ≈ 0.001 Pa·s, but at 60 °C it drops to 0.00047 Pa·s. Always check viscosity tables at the operating temperature. Neglecting this can shift Re by a factor of 2 or more.
- Confusing density with viscosity — Density and viscosity are independent properties. Heavy fluids (high density) are not necessarily viscous. Honey is dense and viscous; oil is dense but less viscous than honey; mercury is dense but has lower viscosity than water.
- Assuming the transition region is avoidable — In real systems, especially with vibration or surface roughness, turbulence may begin below Re = 3,000. The transition region is wider and more variable than theoretical predictions suggest. Design margins are critical.