Navier-Stokes Equations: Derivation from Newton's Second Law

Boundary conditions are essential: at walls, the no-slip condition applies (fluid velocity equals wall velocity). For Couette flow, integrate d²u/dy² = 0 to get u(y) = C₁y + C₂. Apply no-slip conditions: u(-h/2) = 0 and u(h/2) = U_wall. Solving gives C₁ = U_wall/h and C₂ = U_wall/2. The velocity profile is u(y) = (U_wall/h)(y + h/2), confirming the linear profile expectation. This solution satisfies both boundary conditions, providing a sanity check.

Couette flow requires specific boundary conditions: (1) No-slip condition at stationary plate (y=0): velocity = 0, (2) No-slip condition at moving plate (y=h): velocity = U. The plates are assumed infinitely long and very thin, allowing two-dimensional analysis. The flow is unidirectional with velocity only in the x-direction. These conditions, combined with the continuity equation, determine the velocity profile.

Two essential boundary conditions determine the integration constants in the velocity profile: (1) No-slip condition at the pipe wall: u_z(R) = 0, meaning the fluid velocity matches the wall velocity (which is zero for fixed walls); (2) Finite velocity at the centerline: The natural logarithm term in the general solution must vanish to prevent infinite velocity at r=0, requiring C₁ = 0. These conditions ensure physically realistic solutions that match the physical constraints of the problem.

Couette flow is a fundamental fluid dynamics phenomenon where fluid flows between two parallel plates, with one plate stationary and the other moving horizontally, creating a boundary layer where velocity increases linearly from zero at the stationary surface to a maximum at the moving surface; this flow configuration demonstrates how boundary conditions (no-slip at the stationary wall and prescribed velocity at the moving wall) uniquely determine the velocity profile through the Navier-Stokes equation solution, showing that when the pressure gradient is zero (simple Couette flow), the velocity varies linearly with distance from the stationary wall, while non-zero pressure gradients produce parabolic velocity distributions.

The no-slip boundary condition states that fluid velocity at a solid boundary equals the velocity of that boundary. For a rotating cylinder of radius A with angular velocity ω, the tangential velocity is ωA. At the cylinder surface (x = ±D/2), v = ±ωA. Applying these conditions to v(x) = Ax + B and solving the system yields A = 0 and B = ωA, giving v(x) = ωA. This represents plug flow where all fluid moves at the same velocity as the boundary.

Nondimensionalization of the Navier-Stokes equations involves introducing characteristic scales (length L, velocity V, frequency f, pressure difference ΔP) to transform dimensional equations into dimensionless form, revealing four key dimensionless parameters: Reynolds number (inertial/viscous forces), Froude number (inertial/gravitational forces), Euler number (pressure/inertial forces), and Strouhal number (unsteady/convective acceleration). This process enables comparison of term magnitudes to identify and eliminate negligible terms, forming simpler approximate equations like the Euler equation when viscous terms are negligible, though care must be taken as viscous effects remain important near boundaries even when bulk flow neglects them.

The dimensional momentum equation is ∂(ρuᵢ)/∂t + ∂(ρuⱼuᵢ)/∂xⱼ = -∂P/∂xᵢ + μ∂²uᵢ/∂xⱼ∂xⱼ. After non-dimensionalization, it becomes ∂(ρ̂ûᵢ)/∂τ̂ + ∂(ρ̂ûⱼûᵢ)/∂x̂ⱼ = -∂P̂/∂x̂ᵢ + (1/Re)∂²ûᵢ/∂x̂ⱼ∂x̂ⱼ, where Re = ρ∞U∞L/μ∞ is the Reynolds number. The Reynolds number represents the ratio of inertial to viscous forces. When Re is large, viscous effects become negligible, and the flow behaves as inviscid. This is why non-dimensional solvers fix the Reynolds number as a key parameter. The energy equation ∂(ρe)/∂t + ∂(ρuⱼe)/∂xⱼ = -∂(Puⱼ)/∂xⱼ + ∂(qⱼ)/∂xⱼ + ∂(τᵢⱼuⱼ)/∂xⱼ becomes ∂(ρ̂ê)/∂τ̂ + ∂(ρ̂ûⱼê)/∂x̂ⱼ = -∂(P̂ûⱼ)/∂x̂ⱼ + (1/RePr)∂²T̂/∂x̂ⱼ∂x̂ⱼ + (1/Re)∂(τ̂ᵢⱼûⱼ)/∂x̂ⱼ, where Pr = μCp/K is the Prandtl number.

The Reynolds number emerges naturally from nondimensionalizing the Navier-Stokes equation for an incompressible, Newtonian fluid by scaling velocity with a characteristic velocity U_s, position with a characteristic length L_s, and time with T_s = L_s/U_s; this process reveals that the Reynolds number represents the ratio of inertial to viscous forces, while other dimensionless numbers like the Euler number (pressure-to-inertial force ratio) and Froude number (gravity-to-inertial force ratio) also appear as parameter groups.

Applying derivative transformation rules systematically to all terms yields the complete nondimensionalized equation: u*du*/dx* + v* dv*/dy* = -dp*/dx* + (μ/ρv∞L)(d²u*/dx*² + d²v*/dy*²). The coefficient μ/(ρv∞L) simplifies to 1/Re, where Re is the Reynolds number. This demonstrates that the entire equation is now dimensionless, with all dimensional quantities absorbed into the dimensionless parameter 1/Re. This approach reveals the relative importance of different physical mechanisms in fluid flow, enabling qualitative analysis and identification of dominant balances without solving the full dimensional equations.

Nondimensionalization of the Navier-Stokes equations involves scaling velocity, pressure, and time using characteristic parameters (density, viscosity, characteristic length, characteristic velocity, and characteristic time) to reveal two dimensionless groups—the Reynolds number (Re = ρUL/μ) and the Strouhal number (St = τ/(L/U))—which determine the relative magnitudes of different terms in the equations; for microfluidic and nanofluidic applications where length scales are small, the Reynolds number is typically very low, allowing neglect of inertial terms and reducing the full Navier-Stokes equations to the simpler Stokes equations that govern viscous-dominated flows.

CFD (Computational Fluid Dynamics) is a numerical approach for solving fluid dynamics and heat transfer problems. Three fundamental approaches exist: (1) Experimental approach - oldest method using physical models and laboratory experiments; (2) Analytical approach - uses governing equations for exact solutions; (3) Computational approach (CFD) - uses numerical methods and computers for approximate solutions. Fluid dynamics problems are governed by conservation laws (mass, momentum, energy) expressed as partial differential equations. The generic conservation equation has four terms: unsteady, convective, diffusion, and source. The convective term is nonlinear (involving products like ρu·u), making analytical solutions impossible for 99.99% of real problems. The continuum assumption requires fluid properties to be continuous functions of space and time. Analytical solutions are only possible when convective terms are zero (e.g., Poiseuille flow). For high Reynolds number flows where convective terms dominate, CFD becomes essential. CFD emerged in the 1940s-1950s when computers became available, enabling solutions for problems that could not be solved analytically or experimentally.

The fundamental Navier-Stokes equations governing fluid dynamics have no general analytical solution, making computational methods essential. CFD solves these equations numerically by dividing objects into millions of tiny cells forming a mesh. Engineers set boundary conditions while computers perform brute-force iteration to balance all cell interactions. This generalized algorithm works for any geometry. Advantages include lower costs than experimental facilities, detailed results explaining why answers are good or bad, and the ability to simulate real-world physics without approximations or physical models.

Computational methods in fluid dynamics (CFD) are numerical solutions to the Navier-Stokes equations used to model complex fluid behavior when analytical solutions are impossible due to discontinuous boundaries and nonlinear phenomena; these methods discretize fluid domains using either mesh-based approaches (finite differences, finite elements, finite volumes) or particle-based methods (SPH), and are essential for applications ranging from aerodynamics and heat transfer to biomedical flows and environmental modeling, though they produce approximations rather than exact simulations due to inherent stochasticity and discretization errors.

CFD solutions are inherently approximate due to two fundamental limitations: First, the transport equations (Navier-Stokes) contain non-linear terms that require numerical discretization methods since analytical solutions are impossible. Second, computational resources (processing power and memory) limit the complexity of models, forcing replacement of some terms with empirical models, particularly for turbulent flows where fluctuations make accurate modeling extremely challenging.

This section introduces why numerical methods are essential in CFD. Governing equations like Navier-Stokes are partial differential equations that provide mathematical descriptions but require numerical interpretation to reveal physical flow behavior. Two fundamental categories exist: deterministic methods (always produce identical outputs for same inputs, e.g., ball bouncing from fixed height) and stochastic methods (statistically-based, produce slightly varying outputs). The three primary numerical methods for CFD are introduced: Finite Difference Method (oldest, simplest for simple geometries), Finite Element Method (handles arbitrary geometries with variational principles), and Finite Volume Method (focuses on flux conservation). The section establishes that numerical methods convert differential equations into numerical values at discrete grid points, making complex fluid dynamics understandable for engineering analysis.

The Navier-Stokes equations are one of the seven Millennium Prize problems established by the Clay Mathematics Institute, offering a $1 million prize for a correct solution. The specific problem asks for a proof of one of four statements regarding the existence and smoothness of solutions. A smooth solution mathematically means the solution is differentiable everywhere. The challenge lies in proving whether solutions always exist and remain smooth for all time, or whether singularities (discontinuities) can develop, which would indicate chaotic or turbulent behavior.

The Navier-Stokes equations, which describe the motion of viscous fluids, remain one of mathematics' greatest unsolved problems despite over 200 years of research; the fundamental question is whether smooth initial data always produces smooth solutions for all time (global existence) or whether singularities (points where the equations break down) can develop, with recent advances including computer-assisted proofs of finite-time blowup for 3D Euler equations in cylindrical domains and non-uniqueness results for Leray-Hopf solutions, though the original Millennium Prize question for Navier-Stokes remains unresolved.

The Navier-Stokes equations are partial differential equations that describe how viscous fluids move. Named after French engineer Claude-Louis Navier and Irish mathematician George Gabriel Stokes, these equations model the motion of liquids and gases. The Navier-Stokes existence and smoothness problem, which is the fourth Millennium Prize Problem, asks whether smooth solutions exist for all initial conditions in three-dimensional space. Specifically, given an initial velocity field determining how particles flow at the starting time, one must prove that there exists a smooth vector velocity field and scalar pressure field that solve the Navier-Stokes equations everywhere, or provide an example where this is impossible.

The Navier-Stokes equations, which describe fluid motion in nature (air, water, blood), consist of two fundamental laws: conservation of mass and Newton's second law applied to fluids. Despite being used successfully in physics and engineering, these equations present a major unsolved mathematical problem: it remains unproven whether, given realistic initial conditions in three dimensions, solutions always exist, are unique, and remain smooth (no singularities). This is one of the six unsolved Millennium Problems, offering a prize of one million dollars and eternal fame to whoever solves it.