Documentation

TauCeti.Analysis.PDE.EnergyLowerBounds

Pointwise diagonal lower bounds for divergence-form energy integrands #

The pointwise energy integrand of a divergence-form operator acts on jets in ℝ × EuclideanSpace ℝ n. With principal lower bound λ, drift bound β, and mass floor μ, its diagonal satisfies the Young-inequality estimate (λ - ε)‖∇u‖² + (μ - β²/(4ε))|u|² ≤ energyIntegrand A b c U U for every ε > 0. Taking ε = λ/2 gives the explicit product-norm bound min (λ/2) (μ - β²/(2λ)) · ‖U‖² ≤ energyIntegrand A b c U U when λ > 0. The mass floor may have either sign; when it dominates the drift defect, the bound is nonnegative.

These estimates feed the integrated inequalities in TauCeti.Analysis.PDE.EnergyForm.Integrated.Basic. Coercivity for Lax--Milgram requires bounds for the integrated form on a complete inner-product space.

The theorems take a single principal coefficient A, drift coefficient b₀, and mass coefficient c₀, together with their pointwise bounds. For a coefficient field satisfying UniformlyEllipticOn Ω a λ Λ, use the pointwise specializations in TauCeti.Analysis.PDE.Ellipticity.Energy at a point x ∈ Ω. Symmetry of the zero-drift integrand is recorded separately in TauCeti.Analysis.PDE.SymmetricEnergy.

The estimates follow the standard Young-inequality absorption argument in the energy method, as in Evans, Partial Differential Equations, Chapter 6.

Main declarations #

@[instance_reducible]

Local classical decidable equality for finite coordinate indices in lower-bound proofs.

Equations
Instances For
    theorem TauCeti.PDE.garding_energyIntegrand_self_of_mass_lower_bound_of_bounds_with_parameter {n : Type u_1} [Fintype n] {lam mu beta eps : ℝ} (heps : 0 < eps) {A : Matrix n n ℝ} {b₀ : EuclideanSpace ℝ n} {c₀ : ℝ} (hA : ∀ (ξ : EuclideanSpace ℝ n), lam * ‖ξ‖ ^ 2 ≤ A.toQuadraticForm' ξ.ofLp) (hb : ‖b₀‖ ≤ beta) (hc : mu ≤ c₀) (U : ℝ × EuclideanSpace ℝ n) :
    (lam - eps) * ‖U.2‖ ^ 2 + (mu - beta ^ 2 / (4 * eps)) * U.1 ^ 2 ≤ ((energyIntegrand A b₀ c₀) U) U

    A mass floor and a free positive Young parameter give the diagonal lower bound (λ - ε)‖∇u‖² + (μ - β²/(4ε))|u|². Both coefficients may have either sign.

    theorem TauCeti.PDE.garding_energyIntegrand_self_of_mass_lower_bound_of_bounds {n : Type u_1} [Fintype n] {lam mu beta : ℝ} (hlam : 0 < lam) {A : Matrix n n ℝ} {b₀ : EuclideanSpace ℝ n} {c₀ : ℝ} (hA : ∀ (ξ : EuclideanSpace ℝ n), lam * ‖ξ‖ ^ 2 ≤ A.toQuadraticForm' ξ.ofLp) (hb : ‖b₀‖ ≤ beta) (hc : mu ≤ c₀) (U : ℝ × EuclideanSpace ℝ n) :
    lam / 2 * ‖U.2‖ ^ 2 + (mu - beta ^ 2 / (2 * lam)) * U.1 ^ 2 ≤ ((energyIntegrand A b₀ c₀) U) U

    Pointwise lower bound for the energy integrand with bounded drift and a mass lower bound.

    If the principal part has quadratic lower bound λ‖ξ‖², the drift satisfies ‖b₀‖ ≤ β, and the mass coefficient satisfies μ ≤ c₀, then the diagonal of the jet form is bounded below by (λ/2)‖∇u‖² + (μ − β²/2λ)|u|².

    theorem TauCeti.PDE.min_diagonal_lower_bound_mul_norm_sq_le_energyIntegrand_self {n : Type u_1} [Fintype n] {lam mu beta : ℝ} (hlam : 0 < lam) {A : Matrix n n ℝ} {b₀ : EuclideanSpace ℝ n} {c₀ : ℝ} (hA : ∀ (ξ : EuclideanSpace ℝ n), lam * ‖ξ‖ ^ 2 ≤ A.toQuadraticForm' ξ.ofLp) (hb : ‖b₀‖ ≤ beta) (hc : mu ≤ c₀) (U : ℝ × EuclideanSpace ℝ n) :
    min (lam / 2) (mu - beta ^ 2 / (2 * lam)) * ‖U‖ ^ 2 ≤ ((energyIntegrand A b₀ c₀) U) U

    The mass-floor Gårding lower bound implies the explicit diagonal estimate with constant min (λ / 2) (μ - β² / (2λ)), allowing the second coefficient to have either sign.

    theorem TauCeti.PDE.min_lam_mass_mul_norm_sq_le_energyIntegrand_zero_drift_self {n : Type u_1} [Fintype n] {lam : ℝ} {A : Matrix n n ℝ} {c₀ : ℝ} (hlam : 0 ≤ lam) (hA : ∀ (ξ : EuclideanSpace ℝ n), lam * ‖ξ‖ ^ 2 ≤ A.toQuadraticForm' ξ.ofLp) (U : ℝ × EuclideanSpace ℝ n) :
    min lam c₀ * ‖U‖ ^ 2 ≤ ((energyIntegrand A 0 c₀) U) U

    Zero-drift diagonal lower bound from a nonnegative principal quadratic lower bound and an arbitrary mass coefficient.

    theorem TauCeti.PDE.mul_norm_snd_sq_le_energyIntegrand_zero_drift_self {n : Type u_1} [Fintype n] {lam : ℝ} {A : Matrix n n ℝ} {c₀ : ℝ} (hA : ∀ (ξ : EuclideanSpace ℝ n), lam * ‖ξ‖ ^ 2 ≤ A.toQuadraticForm' ξ.ofLp) (hc : 0 ≤ c₀) (U : ℝ × EuclideanSpace ℝ n) :
    lam * ‖U.2‖ ^ 2 ≤ ((energyIntegrand A 0 c₀) U) U

    A zero-drift energy density dominates the squared gradient component when the principal quadratic form has lower bound λ and the mass coefficient is nonnegative.

    Explicit diagonal lower bound for the shifted Laplacian jet form with arbitrary mass.