We use the 2-D steady boundary-layer momentum equation at the wall.
Step 1 — Start from the x-momentum equation inside the thin layer:u∂x∂u+v∂y∂u=−ρ1dxdp+ν∂y2∂2uWhy this form? In a thin boundary layer pressure does not vary across it, so p=p(x) is imposed by the outer flow.
Step 2 — Evaluate AT the wall, y=0.
The no-slip condition gives u=0 and v=0 at the wall, so both convective terms vanish:
(u∂x∂u+v∂y∂u)y=0=0Why this step? No-slip kills u and v exactly on the wall, leaving a clean balance.
Step 3 — What's left is the wall-curvature law. Multiply the leftover viscous balance −ρ1dxdp+ν∂y2∂2u=0 by ρ and use μ=ρν:
μ∂y2∂2uy=0=dxdpWhy this matters: the curvature of the velocity profile at the wall is set entirely by the pressure gradient.
Step 4 — Define separation precisely. Separation begins where the wall shear (and thus the near-wall slope) first drops to zero:
Recall What single quantity equals the wall curvature of the velocity profile?
∂y2∂2uwall=μ1dxdp — the pressure gradient (over μ).
Recall State the exact mathematical condition for the separation point.
τw=μ∂u/∂y∣y=0=0; downstream the near-wall slope goes negative (backflow).
Recall Why does a turbulent boundary layer separate later than a laminar one?
Turbulent mixing brings high-momentum outer fluid to the wall, so it withstands the adverse pressure gradient longer.
Recall Explain to a 12-year-old (Feynman)
Imagine you're cycling and the road starts going uphill. A strong friend zooms over the top; a tired kid runs out of energy halfway, stops, and rolls backward. The air next to a wing or ball is the "tired kid" — viscosity made it slow. When the pressure starts rising (the uphill), that slow air stops and rolls back, so the smooth flow peels off the surface and makes a messy swirling wake. That peeling-off is separation.
Dekho, boundary layer separation samajhna ho to ek simple picture rakho: wall ke paas wali air viscosity ki wajah se already slow ho chuki hai — uska "fuel" (kinetic energy) kam hai. Ab agar flow direction me pressure badhne lage, yaani dp/dx>0 (isko adverse pressure gradient kehte hain), to ye rising pressure us slow air ko peeche dhakelta hai. Free stream wali tez air to pressure ki "uphill" cross kar leti hai, par wall ke paas wali thaki hui air ruk jaati hai aur ulti (backflow) ho jaati hai. Bas yahीं se layer surface se alag (separate) ho jaati hai aur peeche ek ghoomta hua wake ban jaata hai.
Maths ka magic ye hai: wall pe no-slip ki wajah se u=v=0, to momentum equation se sirf bachta hai μ∂2u/∂y2∣wall=dp/dx. Matlab profile ki curvature wall pe sirf pressure gradient se decide hoti hai. Adverse gradient me curvature positive ho jaati hai, jisse profile me inflection point aata hai — yahी unstable aur reversal-prone banata hai. Separation point wahan hai jahan wall shear τw=μ∂u/∂y∣0=0 ho jaata hai; uske aage flow ulta behne lagta hai.
Real life me iska matter kyun karta hai? Diffuser (widening pipe) me area badhta hai, U girta hai, pressure badhta hai — adverse gradient — isliye diffusers easily "stall" karte hain. Cylinder ya ball ke peeche bada wake banta hai to form drag badh jaata hai. Aur mazedaar baat — golf ball ke dimples layer ko turbulent bana dete hain; turbulent layer outer high-momentum air ko wall tak mix karti hai, isliye adverse gradient ko zyada der jhel leti hai, separation late hota hai, wake patla hota hai aur drag kam — ball door jaati hai. Yaad rakho: villain speed nahi, balki adverse pressure gradient hai jo slow wall-fluid ko peeche dhakelta hai.