WITHOUT FRICTION

Abstract

Classical fluid mechanics is frequently framed around historical paradoxes that arise from idealized continuum assumptions. Pascal’s hydrostatic paradox demonstrates that static fluid pressure depends exclusively on column height rather than container geometry, and Archimedes’ principle establishes that the buoyant force equals the weight of the displaced fluid. Conversely, d’Alembert’s paradox (1752) proves that steady, incompressible, inviscid potential flow past an arbitrary body yields identically zero drag force

(\mathbf{F}_{\text{drag}} \equiv 0).

In this paper, we resolve the apparent tension between idealized static equilibria and dynamic real-fluid interactions. First, we clarify the boundary between hydrostatics and hydrodynamics by formalizing buoyancy via Gauss’s divergence theorem applied to the Cauchy stress tensor. We show that d’Alembert’s paradox is an artifact of setting kinematic viscosity \nu = 0, which suppresses the no-slip Dirichlet boundary condition. By analyzing the Navier–Stokes system via the solenoidal Lamb vector

\mathbf{w} = \boldsymbol{\omega} \times \mathbf{u} and the non-local pressure Poisson equation, we demonstrate how boundary vorticity generation and flow separation break fore-aft pressure symmetry, producing form drag and dynamic modifications to static buoyancy. Then we link wake drag directly to the volumetric viscous dissipation rate \varepsilon = 2\nu S_{ij}S_{ij}.

1. Introduction: Disentangling Static and Dynamic Fluid Paradoxes

Much confusion in classical continuum mechanics stems from conflating static force balances with non-equilibrium dynamic flows. Three foundational principles govern these domains:

1. Pascal’s Hydrostatic Paradox (1663): In a static fluid (\mathbf{u} \equiv 0), the pressure at a given depth is independent of the cross-sectional geometry or total volume of the vessel, depending strictly on the gravitational head: p(z) = p_0 + \rho g z.

2. Archimedes’ Principle (c. 250 BCE): The net upward force exerted by a static fluid on an immersed body equals the weight of the displaced fluid:

\mathbf{F}_B = \rho_f g V_{\text{disp}} \, \hat{\mathbf{k}}

3. d’Alembert’s Paradox (1752): For an incompressible, inviscid (\nu = 0), irrotational (\nabla \times \mathbf{u} = 0) potential flow past an immersed obstacle \mathcal{B} \subset \mathbb{R}^3, the net hydrodynamic force acting on the body is identically zero:

\mathbf{F}_{\text{drag}} \equiv 0, \qquad \mathbf{F}_{\text{lift}} \equiv 0

d’Alembert’s mathematical proof within potential theory is unimpeachable, but its premise is physically ungrounded. In nature, fluids possess non-zero kinematic viscosity (\nu > 0). Viscous friction enforces the no-slip condition at solid boundaries, driving boundary-layer vorticity generation, flow separation, and asymmetrical wake formation.

2. Hydrostatic Equilibrium and the Archimedean Limit

Let \Omega \subset \mathbb{R}^3 be a domain occupied by an incompressible Newtonian fluid of constant density \rho_f. The Cauchy stress tensor is given by: \boldsymbol{\sigma} = -p \mathbb{I} + 2\rho_f \nu \mathbf{S} where p is the scalar pressure field, \nu is the kinematic viscosity, and \mathbf{S} = \frac{1}{2}\left(\nabla \mathbf{u} + (\nabla \mathbf{u})^T\right) is the rate-of-strain tensor.

The total force \mathbf{F} exerted by the fluid on the boundary \partial \mathcal{B} of an immersed obstacle \mathcal{B} (with outward normal \mathbf{n} pointing into the fluid) is:

\mathbf{F} = \oint_{\partial \mathcal{B}} \boldsymbol{\sigma} \cdot \mathbf{n} \, dS = -\oint_{\partial \mathcal{B}} p \mathbf{n} \, dS + 2\rho_f \nu \oint_{\partial \mathcal{B}} \mathbf{S} \cdot \mathbf{n} \, dS

Proposition 2.1 (Hydrostatic Exactness)

When the fluid is in global thermodynamic and mechanical equilibrium, the velocity field vanishes identically: \mathbf{u}(x, t) \equiv 0. Consequently:

1. The viscous shear stress tensor vanishes identically: \mathbf{S} \equiv 0.

2. The momentum equation reduces to pure hydrostatic balance: \nabla p = -\rho_f g \hat{\mathbf{k}}

3. Integrating over the closed surface \partial \mathcal{B} and applying Gauss’s Divergence Theorem:

\mathbf{F}_{\text{static}} = -\oint_{\partial \mathcal{B}} p \mathbf{n} \, dS = -\int_{\mathcal{B}} \nabla p \, dV = \rho_f g \hat{\mathbf{k}} \int_{\mathcal{B}} dV = \rho_f g V_{\mathcal{B}} \hat{\mathbf{k}}

Remark 2.1: Hydrostatic buoyancy is mathematically exact for any continuous fluid at rest. It does not depend on turbulence, vortex stretching, or nonlinear advection, because all dynamic operators vanish when \mathbf{u} = 0.

3. The Hydrodynamic Transition: Vorticity and d’Alembert’s Resolution

When an obstacle moves through the fluid at relative velocity \mathbf{U}_\infty, the governing equations are the incompressible Navier–Stokes equations:

\partial_t \mathbf{u} + (\mathbf{u} \cdot \nabla)\mathbf{u} = -\frac{1}{\rho_f}\nabla p + \nu \Delta \mathbf{u} – g\hat{\mathbf{k}}, \qquad \nabla \cdot \mathbf{u} = 0 \tag{3.1} subject to the Dirichlet no-slip boundary condition:

\mathbf{u}\big|_{\partial \mathcal{B}} = 0, \qquad \lim_{|\mathbf{x}| \to \infty} \mathbf{u}(\mathbf{x}, t) = \mathbf{U}_\infty \tag{3.2}

3.1 The Solenoidal Lamb Vector

Using the convective vector identity:

(\mathbf{u} \cdot \nabla)\mathbf{u} = \frac{1}{2}\nabla |\mathbf{u}|^2 + \boldsymbol{\omega} \times \mathbf{u}

where \boldsymbol{\omega} \equiv \nabla \times \mathbf{u} is the vorticity vector, Eq. (3.1) becomes:

\partial_t \mathbf{u} + \boldsymbol{\omega} \times \mathbf{u} = -\nabla\left(\frac{p}{\rho_f} + \frac{1}{2}|\mathbf{u}|^2 + gz\right) + \nu \Delta \mathbf{u} \tag{3.3}

In inviscid potential flow (\nu = 0, \boldsymbol{\omega} \equiv 0), the term \boldsymbol{\omega} \times \mathbf{u} vanishes. The Bernoulli head \mathcal{H} = \frac{p}{\rho_f} + \frac{1}{2}|\mathbf{u}|^2 + gz is strictly harmonic. Fore-aft streamtube symmetry ensures that the surface integral \oint_{\partial \mathcal{B}} p \mathbf{n} \, dS cancels out completely in the direction of motion, producing d’Alembert’s paradox (F_{\text{drag}} = 0).

3.2 Non-Local Pressure Poisson Dynamics

Taking the divergence of Eq. (3.1) and applying the incompressibility constraint \nabla \cdot \mathbf{u} = 0 yields the elliptic Pressure Poisson Equation:

-\frac{1}{\rho_f} \Delta p = \nabla \cdot \bigl((\mathbf{u} \cdot \nabla)\mathbf{u}\bigr) = \partial_i u_j \partial_j u_i = S_{ij}S_{ij} – \frac{1}{2}|\boldsymbol{\omega}|^2 \tag{3.4}

Eq. (3.4) reveals the exact physical mechanism that resolves d’Alembert’s paradox:

1. Viscous Boundary Vorticity: The no-slip condition (\mathbf{u} = 0 on \partial \mathcal{B}) generates intense wall vorticity \boldsymbol{\omega}_{\text{wall}} = \mathbf{n} \times (\nabla \times \mathbf{u}).

2. Boundary Layer Separation: Adverse pressure gradients force the boundary layer to detach, shedding concentrated vortex sheets into the downstream wake.

3. Wake Pressure Deficit: In regions of high vorticity (|\boldsymbol{\omega}|^2 > 2 S_{ij}S_{ij}), Eq. (3.4) acts as an elliptic source, driving a substantial depression in the local pressure p.

4. Asymmetry and Form Drag: The low-pressure wake breaks the fore-aft pressure symmetry:

\mathbf{F}_{\text{pressure}} = -\oint_{\partial \mathcal{B}} p \, \mathbf{n} \, dS \ne 0 \implies F_{\text{drag}} > 0

4. Dynamic Buoyancy Modifications and Wake Dissipation

When an object maneuvers through a real fluid, the total vertical force \mathbf{F}_z is the superposition of static Archimedean buoyancy and dynamic vertical stress:

\mathbf{F}_z = \underbrace{\rho_f g V_{\mathcal{B}} \hat{\mathbf{k}}}_{\text{Static Archimedean Buoyancy}} + \underbrace{\oint_{\partial \mathcal{B}} \left( -p_{\text{dyn}} \mathbf{n} \cdot \hat{\mathbf{k}} + 2\rho_f\nu (\mathbf{S}\cdot \mathbf{n})\cdot \hat{\mathbf{k}} \right) dS}_{\text{Dynamic Normal / Lift Force}}

Vortical structures generated by body motion (e.g., oscillating fins in aquatic locomotion or wake downwash) generate dynamic pressure gradients that modify the apparent buoyant force.

Energy Cascade and Drag Equivalence

In steady turbulent flow, the mechanical work done to overcome drag is continuously transferred across scales via the convective energy cascade and dissipated into heat at the Kolmogorov scale \eta \equiv (\nu^3/\bar{\varepsilon})^{1/4}.

The rate of total mechanical energy dissipation throughout the fluid domain \Omega \setminus \mathcal{B} equals the product of drag force and free-stream velocity: \mathbf{F}_{\text{drag}} \cdot \mathbf{U}_\infty = 2\rho_f \nu \int_{\Omega \setminus \mathcal{B}} S_{ij} S_{ij} \, dV = \int_{\Omega \setminus \mathcal{B}} \rho_f \varepsilon(\mathbf{x}) \, dV where \varepsilon(\mathbf{x}) \equiv 2\nu S_{ij}S_{ij} is the local viscous dissipation rate.

5. Conclusion

1. Pascal’s Hydrostatic Paradox and Archimedes’ Principle describe exact force balances in fluids at rest (\mathbf{u} = 0), where viscous stress and non-linear convective transfer are identically zero.

2. d’Alembert’s Paradox is resolved by restoring kinematic viscosity (\nu > 0) and the no-slip boundary condition. This generates vorticity \boldsymbol{\omega}, drives boundary layer separation, and establishes a low-pressure downstream wake via the pressure Poisson equation.

3. The net drag force represents the macroscopic momentum deficit resulting from work dissipated across the boundary layer and turbulent wake:

F_{\text{drag}} U_\infty = \int \rho \varepsilon \, dV

Appendices

Appendix A: Rigorous Analytical Proof of d’Alembert’s Paradox in Potential Flow

Let \mathcal{B} \subset \mathbb{R}^3 be a compact, smooth, simply connected obstacle immersed in an infinite, incompressible, inviscid fluid governed by the Euler equations:

\frac{\partial \mathbf{u}}{\partial t} + (\mathbf{u} \cdot \nabla)\mathbf{u} = -\frac{1}{\rho_f} \nabla p, \qquad \nabla \cdot \mathbf{u} = 0 \tag{A.1} subject to the kinematic slip condition on the solid boundary \partial \mathcal{B}:

\mathbf{u} \cdot \mathbf{n} = 0 \quad \text{on } \partial \mathcal{B} \tag{A.2}

and the far-field condition \lim_{|\mathbf{x}| \to \infty} \mathbf{u}(\mathbf{x}) = \mathbf{U}_\infty = U_\infty \hat{\mathbf{i}}.

A.1 Potential Representation

For irrotational flow (\boldsymbol{\omega} = \nabla \times \mathbf{u} \equiv 0), the velocity field is represented by a scalar potential \mathbf{u} = \nabla \phi. Incompressibility yields Laplace’s equation:

\Delta \phi = 0 \quad \text{in } \mathbb{R}^3 \setminus \mathcal{B}, \qquad \frac{\partial \phi}{\partial n} = 0 \quad \text{on } \partial \mathcal{B}, \qquad \lim_{|\mathbf{x}| \to \infty} \nabla \phi = \mathbf{U}_\infty \tag{A.3}

In steady flow, Bernoulli’s theorem relates the pressure field p(\mathbf{x}) directly to the kinetic energy density:

p(\mathbf{x}) = p_\infty + \frac{1}{2}\rho_f \left( U_\infty^2 – |\nabla \phi|^2 \right) \tag{A.4}

A.2 Multipole Potential Expansion

Because \Delta \phi = 0 outside \mathcal{B}, the velocity potential admits a multipole expansion in spherical coordinates (r, \theta, \varphi) centered at the body:

\phi(\mathbf{x}) = \mathbf{U}_\infty \cdot \mathbf{x} + \frac{\mathbf{d} \cdot \mathbf{x}}{4\pi |\mathbf{x}|^3} + \mathcal{O}\left(\frac{1}{|\mathbf{x}|^3}\right) \quad \text{as } |\mathbf{x}| \to \infty \tag{A.5}

where \mathbf{d} is the induced hydrodynamic dipole moment, and the monopole term vanishes identically due to zero net mass flux across \partial \mathcal{B} (\oint_{\partial \mathcal{B}} \nabla \phi \cdot \mathbf{n} \, dS = 0).

Differentiating Eq. (A.5), the perturbation velocity field \mathbf{u}’ = \nabla \phi – \mathbf{U}_\infty satisfies:

\mathbf{u}'(\mathbf{x}) = \mathcal{O}\left(\frac{1}{|\mathbf{x}|^3}\right) \implies |\nabla \phi|^2 = U_\infty^2 + 2 \mathbf{U}_\infty \cdot \mathbf{u}’ + \mathcal{O}\left(\frac{1}{|\mathbf{x}|^6}\right) \tag{A.6}

A.3 Control Volume Momentum Integration

Consider a large spherical control volume \mathcal{S}_R of radius R enclosing \mathcal{B}. By momentum conservation, the total force \mathbf{F} exerted by the fluid on the obstacle is:

\mathbf{F} = -\oint_{\partial \mathcal{B}} p \mathbf{n} \, dS = -\oint_{\mathcal{S}_R} \left[ p \mathbf{n} + \rho_f \mathbf{u} (\mathbf{u} \cdot \mathbf{n}) \right] dS \tag{A.7}

Substituting Eq. (A.4) and expanding on \mathcal{S}_R (where dS = R^2 d\Omega):

\mathbf{F} = \rho_f \oint_{\mathcal{S}_R} \left[ \left( \mathbf{U}_\infty \cdot \mathbf{u}’ \right) \mathbf{n} – \mathbf{U}_\infty (\mathbf{u}’ \cdot \mathbf{n}) – \mathbf{u}’ (\mathbf{U}_\infty \cdot \mathbf{n}) \right] dS + \mathcal{O}\left(R^{-1}\right) \tag{A.8}

Because \mathbf{u}’ \sim R^{-3}, the integrand decays as R^{-3}. Integrating over the surface area R^2 d\Omega yields:

\|\mathbf{F}\| \le \lim_{R \to \infty} \frac{C}{R} = 0 \tag{A.9} Thus, \mathbf{F} \equiv 0, identically yielding d’Alembert’s Paradox: neither drag nor lift can exist in a steady, unconfined potential flow.

Appendix B: Boundary Vorticity Generation and the Pressure Poisson Boundary Value Problem

In a viscous fluid (\nu > 0), the adherence condition \mathbf{u}|_{\partial \mathcal{B}} = 0 introduces boundary vorticity generation that breaks the potential flow symmetry.

B.1 The Boundary Pressure Gradient

Evaluating the momentum equation (3.1) directly on the rigid boundary

\partial \mathcal{B} (where \mathbf{u} = 0 and \partial_t \mathbf{u} = 0 insteady flow): \nabla p\big|_{\partial \mathcal{B}} = \mu \Delta \mathbf{u}\big|_{\partial \mathcal{B}} – \rho_f g \hat{\mathbf{k}} \tag{B.1}

Using the vector Laplacian identity \Delta \mathbf{u} = \nabla(\nabla \cdot \mathbf{u}) – \nabla \times (\nabla \times \mathbf{u}) = -\nabla \times \boldsymbol{\omega} (since \nabla \cdot \mathbf{u} = 0), the wall-normal pressure gradient satisfies:

\frac{\partial p}{\partial n} = \mathbf{n} \cdot \nabla p = -\mu \, \mathbf{n} \cdot (\nabla \times \boldsymbol{\omega}) – \rho_f g (\mathbf{n} \cdot \hat{\mathbf{k}}) \tag{B.2}

Eq. (B.2) couples the surface pressure distribution directly to the flux of vorticity generated at the wall.

B.2 Formal Green’s Function Inversion for the Dynamic Pressure

Taking the divergence of the Navier–Stokes momentum equation yields the inhomogeneous Poisson equation for dynamic pressure p_{\text{dyn}} = p + \rho_f g z:

-\Delta p_{\text{dyn}} = \rho_f Q(\mathbf{x}), \qquad Q(\mathbf{x}) \equiv \partial_i u_j \partial_j u_i = S_{ij}S_{ij} – \frac{1}{2}|\boldsymbol{\omega}|^2 \tag{B.3} subject to the Neumann boundary condition on \partial \mathcal{B}:

\frac{\partial p_{\text{dyn}}}{\partial n}\Bigg|_{\partial \mathcal{B}} = -\mu \, \mathbf{n} \cdot (\nabla \times \boldsymbol{\omega})\Big|_{\partial \mathcal{B}} \tag{B.4}

Let G(\mathbf{x}, \mathbf{x}’) denote the Neumann Green’s function for the domain \Omega = \mathbb{R}^3 \setminus \mathcal{B}, satisfying \Delta_{\mathbf{x}’} G(\mathbf{x}, \mathbf{x}’) = \delta(\mathbf{x} – \mathbf{x}’) with \frac{\partial G}{\partial n’} = 0 on \partial \mathcal{B}.

The exact non-local pressure field is:

p_{\text{dyn}}(\mathbf{x}) = \rho_f \int_{\Omega} G(\mathbf{x}, \mathbf{x}’) \left( S_{ij}S_{ij} – \frac{1}{2}|\boldsymbol{\omega}|^2 \right) dV’ – \mu \oint_{\partial \mathcal{B}} G(\mathbf{x}, \mathbf{x}’) \left[ \mathbf{n}’ \cdot (\nabla’ \times \boldsymbol{\omega}) \right] dS’ \tag{B.5}

Physical Mechanism:

1. In the wake downstream of flow separation, vorticity accumulates (|\boldsymbol{\omega}|^2 \gg 2 S_{ij}S_{ij}), making Q(\mathbf{x}’) < 0.

2. The negative source term inside the volume integral reduces the wake pressure relative to the stagnation pressure at the upstream face.

3. Integrating the resulting asymmetric pressure field over \partial \mathcal{B} yields non-zero form drag:

\mathbf{F}_{\text{form}} = -\oint_{\partial \mathcal{B}} p_{\text{dyn}} \mathbf{n} \, dS \ne 0

Appendix C: Control Volume Momentum Flux and Global Dissipation Equivalence

We prove that the mechanical work done to propel a body through a viscous fluid is identically equal to the total volumetric viscous dissipation rate across the fluid.

C.1 Momentum and Energy Balance

Let an obstacle \mathcal{B} translate at steady velocity -\mathbf{U}_\infty. In a body-fixed reference frame, the fluid enters at velocity \mathbf{U}_\infty and exits with a downstream deficit \mathbf{u}(\mathbf{x}) = \mathbf{U}_\infty + \mathbf{u}'(\mathbf{x}).

Taking the inner product of the steady Navier–Stokes equations with \mathbf{u}:

\nabla \cdot \left[ \mathbf{u} \left( p + \frac{1}{2}\rho_f |\mathbf{u}|^2 \right) – 2\mu \mathbf{u} \cdot \mathbf{S} \right] = -2\mu \mathbf{S} : \mathbf{S} \equiv -\rho_f \varepsilon(\mathbf{x}) \tag{C.1} where \mathbf{S} : \mathbf{S} = S_{ij}S_{ij} and \varepsilon(\mathbf{x}) = 2\nu S_{ij}S_{ij} is the local rate of viscous dissipation.

C.2 Surface Integration

Integrate Eq. (C.1) over the volume \Omega_V bounded internally by the body surface \partial \mathcal{B} and externally by a far-field cylindrical control surface \mathcal{S}_{\text{out}}:

\oint_{\partial \mathcal{B} \cup \mathcal{S}_{\text{out}}} \left[ \left( p + \frac{1}{2}\rho_f |\mathbf{u}|^2 \right) \mathbf{u} – 2\mu \mathbf{u} \cdot \mathbf{S} \right] \cdot \mathbf{n} \, dS = -2\mu \int_{\Omega_V} S_{ij}S_{ij} \, dV \tag{C.2}

On the body surface \partial \mathcal{B}, the no-slip condition enforces \mathbf{u} \equiv 0, causing the inner surface integral to vanish identically.

On the far-field boundary \mathcal{S}_{\text{out}}, decomposing \mathbf{u} = \mathbf{U}_\infty + \mathbf{u}’, neglecting terms quadratic in the small perturbation \mathbf{u}’, and utilizing the momentum balance \mathbf{F}_{\text{drag}} = \oint_{\mathcal{S}_{\text{out}}} \left[ \rho_f \mathbf{u}’ (\mathbf{U}_\infty \cdot \mathbf{n}) + p’\mathbf{n} \right] dS yields:

\mathbf{F}_{\text{drag}} \cdot \mathbf{U}_\infty = 2\mu \int_{\Omega} S_{ij} S_{ij} \, dV = \int_{\Omega} \rho_f \varepsilon(\mathbf{x}) \, dV \tag{C.3}

This identity proves that drag is strictly an energetic phenomenon. An object will not experience hydrodynamic drag without continuously generating velocity gradients that dissipate kinetic energy into heat.

Appendix D: The Singular Inviscid Limit (\nu \to 0^+) and Kato’s Criterion

A central question in fluid mechanics is why solutions of the Navier–Stokes equations do not converge to solutions of the Euler potential flow equations as \nu \to 0^+.

D.1 Singular Perturbation and the Loss of Boundary Conditions

The Navier–Stokes momentum equation features viscosity as the coefficient of the highest spatial derivative: \nu \Delta \mathbf{u} = \nu \nabla^2 \mathbf{u}

Setting \nu = 0 reduces the spatial order of the partial differential equation from 2 to 1. Consequently, the inviscid Euler system can only satisfy the single scalar kinematic condition \mathbf{u} \cdot \mathbf{n} = 0, abandoning the tangential no-slip conditions:

\mathbf{u} \cdot \mathbf{t}_1 = 0, \qquad \mathbf{u} \cdot \mathbf{t}_2 = 0 \quad \text{on } \partial \mathcal{B}

This reduction produces a singular boundary layer of thickness \delta \sim \sqrt{\nu L / U_\infty} (Prandtl, 1904). Within this layer, tangential velocity gradients scale as:

\|\nabla \mathbf{u}\| \sim \frac{U_\infty}{\delta} \sim mathcal{O}\left(\nu^{-1/2}\right)

The volumetric dissipation rate within the boundary layer volume V_{\text{BL}} \sim A_{\text{body}} \delta scales as:

\int_{V_{\text{BL}}} \nu \|\nabla \mathbf{u}\|^2 \, dV \sim \nu \left( \frac{U_\infty}{\delta} \right)^2 (A \delta) \sim \nu \left(\frac{U_\infty^2}{\nu}\right) \delta \sim U_\infty^2 \sqrt{\nu} \xrightarrow[\nu \to 0^+]{} 0

However, when the boundary layer separates, it injects an active sheet of concentrated vorticity into the outer flow, triggering wake turbulence where dissipation becomes independent of viscosity (anomalous dissipation).

D.2 Kato’s Theorem (1984)

Kato rigorously formalized the criterion under which a viscous flow converges to an inviscid Euler flow:

Theorem (Tosio Kato, 1984): Let \mathbf{u}^\nu be a family of Leray–Hopf weak solutions of the Navier–Stokes equations in a domain \Omega with solid boundary \partial \Omega, and let \mathbf{u}^0 be a smooth solution of the Euler equations with \mathbf{u}^0 \cdot \mathbf{n} = 0.

Then \mathbf{u}^\nu \to \mathbf{u}^0 in L^\infty([0, T]; L^2(\Omega)) as \nu \to 0 if and only if:

\lim_{\nu \to 0^+} \nu \int_0^T \int_{\Gamma_{c\nu}} \|\nabla \mathbf{u}^\nu\|^2 \, dx \, dt = 0 \tag{D.1} where

\Gamma_{c\nu} = \{ x \in \Omega : \operatorname{dist}(x, \partial \Omega) \le c\nu \} is an inner boundary strip of width proportional to \nu.

Implication for d’Alembert’s Paradox:

In real boundary-layer flows with adverse pressure gradients, flow separation causes the energy dissipation within the near-wall strip \Gamma_{c\nu} to remain strictly bounded away from zero (\mathcal{O}(1)). Thus, the limit \nu \to 0^+ violates Kato’s criterion, proving that the inviscid Euler limit is mathematically discontinuous from real viscous hydrodynamics. Potential flow can never be recovered for separated flows around blunt bodies.

References

1. d’Alembert, J. le Rond (1752). Essai d’une nouvelle théorie de la résistance des fluides. David l’aîné, Paris.

2. Pascal, B. (1663). Traitez de l’équilibre des liqueurs et de la pesanteur de la masse de l’air. Guillaume Desprez, Paris.

3. Prandtl, L. (1904). Über Flüssigkeitsbewegung bei sehr kleiner Reibung. Verhandlungen des III. Internationalen Mathematiker-Kongresses, Heidelberg, pp. 484–491.

4. Batchelor, G. K. (1967). An Introduction to Fluid Dynamics. Cambridge University Press, Cambridge.

5. Landau, L. D., & Lifshitz, E. M. (1987). Fluid Mechanics (Course of Theoretical Physics, Vol. 6, 2nd ed.). Pergamon Press, Oxford.

6. Schlichting, H., & Gersten, K. (2017). Boundary-Layer Theory (9th ed.). Springer-Verlag, Berlin.

7. Lamb, H. (1932). Hydrodynamics (6th ed.). Cambridge University Press, Cambridge.

8. Kato, T. (1984). Remarks on zero viscosity limit for incompressible Navier–Stokes equations. Seminar on Nonlinear Partial Differential Equations (Chern, S. S., ed.), Mathematical Sciences Research Institute Publications, Vol. 2, Springer, New York, pp. 85–98.

9. Beale, J. T., Kato, T., & Majda, A. (1984). Remarks on the breakdown of smooth solutions for the 3-D Euler equations. Communications in Mathematical Physics, 94(1), 61–66.

10. Saffman, P. G. (1992). Vortex Dynamics. Cambridge University Press, Cambridge.

11. Onsager, L. (1949). Statistical hydrodynamics. Il Nuovo Cimento (1943–1954), 6(Suppl. 2), 279–287.

12. Eyink, G. L., & Sreenivasan, K. R. (2006). Onsager and the theory of hydrodynamic turbulence. Reviews of Modern Physics, 78(1), 87–135.

13. Isett, P. (2018). A proof of Onsager’s conjecture. Annals of Mathematics, 188(3), 871–963.

14. Pope, S. B. (2000). Turbulent Flows. Cambridge University Press, Cambridge.

15. White, F. M. (2016). Fluid Mechanics (8th ed.). McGraw-Hill Education, New York.