Documentation

LeanPool.EscauriazaSereginSverak.PartV.ForcedHeatCriticalGN

Real-valued interpolation and the regularized critical profile #

The Gagliardo–Nirenberg inequality of lem:u-ten-thirds of the CKN manuscript in real integral form for continuously differentiable functions, and the elementary bounds and limits of the regularized critical profile heatRegG η and entropy heatRegEnergy η as η → 0, as used in lem:pv-stokes.

The whole-space Gagliardo–Nirenberg inequality in real integral form: ∫ |f|^{10/3} ≤ K² (∫ f²)^{2/3} ∫ |∇f|², with K the real Sobolev constant of CKN.h1_gagliardoNirenberg_tenThirds.

theorem ESS.heatRegG_nonneg_le_rpow {η : ℝ} (hη : 0 < η) (v : CKN.Foundation.Parabolic.Vec3) :
0 ≤ heatRegG η v ∧ heatRegG η v ≤ (∑ i : Fin 3, v i ^ 2) ^ (3 / 4)

The regularized critical profile lies between 0 and (∑ v_i²)^{3/4}.

theorem ESS.heatRegG_tendsto_rpow (v : CKN.Foundation.Parabolic.Vec3) :
Filter.Tendsto (fun (η : ℝ) => heatRegG η v) (nhdsWithin 0 (Set.Ioi 0)) (nhds ((∑ i : Fin 3, v i ^ 2) ^ (3 / 4)))

As the regularization vanishes, the regularized critical profile tends to (∑ v_i²)^{3/4} = |v|^{3/2}.

theorem ESS.heatRegEnergy_tendsto_rpow (v : CKN.Foundation.Parabolic.Vec3) :
Filter.Tendsto (fun (η : ℝ) => heatRegEnergy η v) (nhdsWithin 0 (Set.Ioi 0)) (nhds (1 / 3 * (∑ i : Fin 3, v i ^ 2) ^ (3 / 2)))

As the regularization vanishes, the regularized entropy tends to ⅓ (∑ v_i²)^{3/2} = ⅓ |v|³.

theorem ESS.fderiv_scalar_comp_slice_apply {φ : CKN.Foundation.Parabolic.Vec3 → ℝ} (hφ : Differentiable ℝ φ) {U : Fin 3 → CKN.Foundation.Parabolic.Vec3 × ℝ → ℝ} (hU : ∀ (k : Fin 3), Differentiable ℝ (U k)) (x : CKN.Foundation.Parabolic.Vec3) (t : ℝ) (v : CKN.Foundation.Parabolic.Vec3) :
(fderiv ℝ (fun (y : CKN.Foundation.Parabolic.Vec3) => φ fun (k : Fin 3) => U k (y, t)) x) v = (fderiv ℝ φ fun (k : Fin 3) => U k (x, t)) fun (k : Fin 3) => (fderiv ℝ (U k) (x, t)) (v, 0)

Chain rule for a scalar map applied to a time slice of a smooth vector field.

A time window as a restriction of the space-time product measure.