When a viscous fluid flows over a solid surface, it cannot slip at the wall — the fluid layer in direct contact must match the surface velocity. This no-slip condition creates a thin region called the boundary layer in which the velocity transitions from zero at the wall to the freestream value U∞, and within which viscous forces are comparable to inertial forces. In 1908, Paul Richard Heinrich Blasius solved the governing Navier–Stokes equations for laminar flow over a flat plate analytically via a similarity variable η = y√(U∞/νx), yielding a universal velocity profile described by the ordinary differential equation f‴ + ½ff″ = 0. This elegant result underpins aircraft wing design, heat transfer calculations, and drag estimation on ships and submarines.
Adjust the freestream velocity U∞ and plate length L, and switch between water and air to see how kinematic viscosity ν shapes the boundary layer thickness δ₉₉ ≈ 5.0√(νx/U∞). The canvas shows the δ(x) edge curve growing with √x, velocity profiles plotted at four streamwise stations, and the skin-friction coefficient C̄f = 1.328/√Re_L. All displayed values update live as you change parameters.
What is the Blasius boundary layer and who derived it?
The Blasius boundary layer is the exact analytical solution to the laminar Navier–Stokes equations for flow over a semi-infinite flat plate with no pressure gradient, derived by Paul Blasius in 1908 as his doctoral dissertation under Ludwig Prandtl. Blasius introduced the similarity variable η = y√(U∞/νx), reducing the partial differential equations to a single third-order ODE: f‴ + ½ff″ = 0, with boundary conditions f(0)=0, f′(0)=0, f′(∞)=1. The shooting parameter f″(0) ≈ 0.4696 is determined numerically and completely characterises the universal velocity profile u/U∞ = f′(η).
How does boundary layer thickness δ grow with distance along the plate?
The 99% boundary layer thickness δ₉₉ — the height at which u/U∞ = 0.99 — grows as δ₉₉ ≈ 5.0√(νx/U∞) = 5.0x/√Re_x. This √x dependence arises from the balance between viscous diffusion (which spreads momentum laterally at rate ~√(νt)) and downstream convection. Doubling the plate length increases δ at the trailing edge by a factor of √2 ≈ 1.41. For air at 15 m/s over a 0.5 m plate, δ₉₉ at the trailing edge is roughly 5 mm — about the thickness of a credit card.
What is the skin friction coefficient and how is it calculated?
The local skin friction coefficient C_f(x) = τ_w/(½ρU∞²) = 0.664/√Re_x, where τ_w = μ(∂u/∂y)|_{y=0} is the wall shear stress. The plate-averaged value C̄_f = 1.328/√Re_L integrates this from leading to trailing edge. For a flat plate in air at Re_L = 10⁶, C̄_f ≈ 1.33×10⁻³ — a small but non-negligible drag source for aircraft. This laminar formula breaks down above Re_x ≈ 5×10⁵, where the boundary layer transitions to turbulence and C_f jumps significantly.
Transition occurs when small disturbances in the flow are amplified rather than damped. The critical Reynolds number Re_x,crit for a flat plate boundary layer is typically 3×10⁵–10⁶, depending on freestream turbulence intensity, surface roughness, and pressure gradient. Above this threshold, Tollmien-Schlichting waves grow exponentially, eventually breaking down into three-dimensional turbulent spots that merge and grow downstream. Transition dramatically increases skin friction (turbulent C_f ∝ Re_x^{-1/5} vs laminar Re_x^{-1/2}) but also enhances heat transfer and mixing — a mixed blessing for engineers.
Boundary layer thickness scales as δ ∝ √(ν/U∞), where ν is kinematic viscosity. Air at 20°C has ν ≈ 1.5×10⁻⁵ m²/s whilst water has ν ≈ 1×10⁻⁶ m²/s — roughly 15 times smaller. Therefore, at the same freestream velocity, water's boundary layer is √15 ≈ 3.9 times thinner than air's. This means water exerts higher shear stress per unit area than air at the same velocity (because τ_w = μU∞×0.332/√(νx)), even though water's dynamic viscosity μ = ρν is much larger.
The displacement thickness δ* = ∫₀^∞ (1 − u/U∞) dy ≈ 1.72√(νx/U∞) is the thickness of an imaginary layer of inviscid fluid that carries the same mass deficit as the real boundary layer. It represents how much the boundary layer effectively "pushes" the external streamlines away from the wall. For wing design, displacement thickness adds to the apparent aerodynamic shape of the wing surface — a 1 mm displacement thickness on a 200 mm chord aerofoil represents a ~0.5% effective thickness increase that must be accounted for in profile drag calculations.
A favourable pressure gradient (dP/dx < 0, flow accelerating) thins and stabilises the boundary layer by adding momentum to near-wall fluid, delaying transition to turbulence and preventing separation. An adverse pressure gradient (dP/dx > 0, flow decelerating) thickens the boundary layer, reduces near-wall momentum, and can cause flow separation when the wall shear stress drops to zero — the condition at which the Blasius solution breaks down. Separation occurs on the aft portion of wings at high angles of attack and causes stall — a critical phenomenon in aviation.
Boundary layer theory underlies the drag and heat transfer design of virtually every vehicle that moves through a fluid. Aircraft wings use carefully shaped aerofoils to manage the pressure gradient and delay separation; submarines control their boundary layers to reduce sonar-detectable noise; gas turbine blades must survive heat transfer rates calculated from boundary layer correlations; and the International Space Station reentry module uses boundary layer heating predictions to size its thermal protection system. The Blasius solution remains the starting point for all these analyses.
The momentum thickness θ = ∫₀^∞ (u/U∞)(1 − u/U∞) dy ≈ 0.664√(νx/U∞) measures the momentum deficit in the boundary layer and is directly related to drag via the von Kármán momentum integral equation: τ_w = ρU∞² dθ/dx. The total drag on one side of a plate of width b and length L equals ρU∞²bθ(L) — a beautiful result known as the von Kármán momentum thickness theorem that avoids integrating the local shear stress. For the Blasius profile, θ = 0.664δ₉₉/5.0 = 0.133δ₉₉.
The thermal boundary layer develops analogously to the velocity boundary layer, with thickness δ_T ≈ δ × Pr^{-1/3} where Pr = ν/α is the Prandtl number (ratio of momentum to thermal diffusivity). For air, Pr ≈ 0.71, so the thermal and velocity layers are nearly equal. For water, Pr ≈ 7, meaning heat is conducted much more slowly than momentum diffuses, producing a thermal layer roughly half as thick as the velocity layer. The local Nusselt number Nu_x = 0.332 Re_x^{1/2} Pr^{1/3} gives the heat transfer coefficient directly from these boundary layer parameters.
At the leading edge (x = 0), the Blasius solution predicts a singularity in the local skin friction coefficient C_f ∝ x^{-1/2} → ∞, and the boundary layer thickness δ → 0. Physically, the boundary layer at the very tip has zero thickness — the fluid first contacts the wall there. In practice, real plates have finite leading-edge thickness and bluntness effects create a small region where the Blasius solution is not strictly valid. For most engineering purposes this "leading-edge correction" is negligible because it affects only a tiny fraction of the plate length.
Near any solid surface, a moving fluid slows down due to friction, forming a thin region known as the boundary layer, where velocity rises from zero at the wall up to the full freestream speed. This page renders the Blasius solution — the exact analytical description of a laminar boundary layer developing over a flat plate with no pressure gradient. Watch how the layer thickens further downstream, how the velocity profile bulges outward, and how swapping fluids shifts the balance between viscous and inertial forces, summarised by the Reynolds number. The same physics governs drag on aircraft wings, ship hulls and turbine blades.
The canvas plots the boundary-layer edge δ₉₉ growing as √x along the plate, four velocity profiles u/U∞ at successive stations, and particles advecting at the local Blasius velocity. Live stat cards report ReL, δ₉₉, the average skin-friction coefficient C̄f, and the wall shear stress τw at the trailing edge.
Drag the Freestream U∞ slider to change flow speed and the Plate length L slider to change how far the fluid travels. Toggle the Water and Air buttons to swap kinematic viscosity ν, which reshapes how quickly the boundary layer grows and how much skin friction it produces.
Ludwig Prandtl introduced the boundary-layer concept in 1904; his student Paul Blasius solved it exactly for laminar flow in 1908. Shark skin and golf-ball dimples both exploit boundary-layer behaviour deliberately, encouraging an earlier, thinner turbulent layer that lowers overall drag.
Viscous diffusion spreads momentum away from the wall whilst the fluid simultaneously convects downstream, giving δ₉₉ ∝ √x. That is why doubling the plate length only increases the trailing-edge thickness by about 1.4 times, not twice.
The laminar Blasius profile modelled here is smooth and self-similar, described by u/U∞ = f′(η). A turbulent profile is fuller nearer the wall, closer to a one-seventh power law, because turbulent mixing transports momentum towards the surface far more effectively than viscosity alone.
Air's kinematic viscosity is roughly fifteen times that of water. Since thickness scales as δ ∝ √ν, air's boundary layer at the same freestream speed ends up around four times thicker than water's.
ReL = U∞L/ν compares inertial to viscous forces. Low values keep the flow laminar, as modelled here, whilst real flows typically transition towards turbulence once the local Reynolds number climbs past roughly 500,000.
The no-slip condition forces fluid touching the plate to match its zero velocity, creating a steep velocity gradient and wall shear stress τw. Averaging this stress along the plate gives the skin-friction coefficient C̄f = 1.328/√ReL.