Named terms in the Section 5.1 residual identity.
The flow-pulled gradient G_l = ∇(T ∘ X_l) ∘ X_l⁻¹.
Equations
- AVenhance.Infra.Section5.G I hΦ m T l t x = AVenhance.spaceGrad (fun (y : Homogenization.Vec 2) => T t (I.xFlow hΦ m l t y)) (I.xFlowInv hΦ m l t x)
Instances For
The spatial derivative matrix of G_l, with coordinate derivatives in rows.
Equations
- AVenhance.Infra.Section5.gradG I hΦ m T l t x = Matrix.of fun (i j : Fin 2) => AVenhance.spaceGrad (fun (y : Homogenization.Vec 2) => AVenhance.Infra.Section5.G I hΦ m T l t y j) x i
Instances For
The scalar coefficient ψTilde_m obtained by pulling each ψ_{m,k} back
along the l_k inverse flow.
Equations
- One or more equations did not get rendered due to their size.
Instances For
The diffusion matrix in the ansatz equation.
Equations
- AVenhance.Infra.Section5.diffusionMatrix I hΦ m κm t x = κm • 1 + AVenhance.Infra.Section5.psiTilde I hΦ m t x • AVenhance.sigmaMat
Instances For
The twisted corrector-gradient matrix ∇ΧTilde_{m,k}.
Equations
- AVenhance.Infra.Section5.gradChiTilde I hΦ m κm k t x = AVenhance.gradMatrix (I.chiTilde hΦ m κm k t) x
Instances For
The l_k flow-gradient matrix ∇X_{m-1,l_k} ∘ X⁻¹_{m-1,l_k}.
Equations
- AVenhance.Infra.Section5.flowGradK I hΦ m k t x = I.flowGrad hΦ m (AVenhance.lIdx β I.Λ m k) t x
Instances For
The spatial gradient of the divergence of a vector field.
Equations
- AVenhance.Infra.Section5.gradDiv V t x = AVenhance.spaceGrad (fun (y : Homogenization.Vec 2) => AVenhance.vecDiv (V t) y) x
Instances For
Column-wise divergence of a matrix field: the flux coordinate is differentiated and the column/corrector index is returned. This is the convention in the per-corrector flux equation.
Equations
- AVenhance.Infra.Section5.matDiv A x j = ∑ i : Fin 2, AVenhance.spaceGrad (fun (y : Homogenization.Vec 2) => A y i j) x i
Instances For
The centered flow average used in the correction.
Equations
- AVenhance.Infra.Section5.Gbar I hΦ m T t x = ∑' (l : ℤ), I.hatXiML m l t • AVenhance.Infra.Section5.G I hΦ m T l t x
Instances For
The flux matrix attached to the k-th corrector in its cutoff window.
Equations
- One or more equations did not get rendered due to their size.
Instances For
cutoff1, the time-cutoff derivative term.
Equations
- One or more equations did not get rendered due to their size.
Instances For
twistie1, the transport of T_{m-1} through the l_k flow.
Equations
- One or more equations did not get rendered due to their size.
Instances For
twistie3 in the Σ_k ξ_{m,k} form for the line-2 ansatz.
The source's printed l-average is the line-1 hybrid; with the line 2,
the gradient defect is ∇T - G_{l_k}. Under this slot keeps its body: it is
the repaired twistie3⁺ = -div Σ_k ξ_k p_k σ (∇T - G_k) plus the molecular divergence
𝓜 = div(κ Σ_k ξ_k (F_k - I)∇T) of the corrected formulation (37) (LeftJacobian.twistie3_eq),
so the ten slot
names are unchanged.
Equations
- One or more equations did not get rendered due to their size.
Instances For
twistie4ND, the non-divergence inverse-flow corrector-gradient error
-Σ_k ξ_k G_k · Div D_k. In the corrected form, twistie4 = twistie4ND - Σ_k ξ_k D_k : ∇G_k
(LeftJacobian.twistie4Plus_eq).
Equations
- One or more equations did not get rendered due to their size.
Instances For
twistie5ND, the non-divergence push-forward Jacobian error
-Σ_k ξ_k G_k · Div E_k (since twistie5 = twistie5ND - Σ_k ξ_k E_k : ∇G_k). The gradMatrix
convention stores coordinate derivatives in rows, so the Piola push-forward
acts by the transpose on vector fields. This is the corrected orientation of
the source's I - ∇X factor (see
PushforwardOrientation).
Equations
- One or more equations did not get rendered due to their size.
Instances For
twistie4 (the corrected formulation §7, twistie4⁺): -div Σ_k ξ_k D_k G_k with the
inverse-flow corrector defect D_k = κ (∇Y_{l_k} - I)(∇χ_k ∘ Y_{l_k}).
Equations
- One or more equations did not get rendered due to their size.
Instances For
twistie5 (the corrected formulation §7, twistie5⁺): -div Σ_k ξ_k E_k G_k with the
push-forward Jacobian error E_k = (I - F_kᵀ)(p_k σ + κ ∇χ_k ∘ Y_{l_k}) (transpose
orientation: see PushforwardOrientation).
Equations
- One or more equations did not get rendered due to their size.
Instances For
The rank-two product ΧTilde_k ∇G_{l_k}, contracted as a vector before
applying the diffusion matrix.
Equations
- AVenhance.Infra.Section5.chiGradG I hΦ m κm T k t x i = ∑ j : Fin 2, I.chiTilde hΦ m κm k t x j * AVenhance.Infra.Section5.gradG I hΦ m T (AVenhance.lIdx β I.Λ m k) t x i j
Instances For
normie1, the corrector times the Hessian term.
Equations
- One or more equations did not get rendered due to their size.
Instances For
normie2, the diffusion of the H_m corrector.
Equations
- One or more equations did not get rendered due to their size.
Instances For
tiny, the residual generated by the two forcing fields.
Equations
- One or more equations did not get rendered due to their size.
Instances For
The selected cell coefficient p_k = ζ̂_{l_k} ζ_k ψ_k ∘ Y_{l_k} (the corrected formulation
§7).
Equations
- AVenhance.Infra.Section5.selCoeff I hΦ m k t x = I.hatZetaML m (AVenhance.lIdx β I.Λ m k) t * I.zetaMK m k t * AVenhance.psi β I.Λ m k (I.xFlowInv hΦ m (AVenhance.lIdx β I.Λ m k) t x)
Instances For
The unpulled cell flux C⁰_k = (κ I + p_k σ)(I + B_k), B_k = (∇χ_k) ∘ Y_{l_k}
(the corrected formulation §7): the cell flux integrand evaluated at Y_{l_k}.
Equations
- One or more equations did not get rendered due to their size.
Instances For
normie3 (the corrected formulation §7, normie3⁺): Σ_k ξ_k F_kᵀ (𝒥 - C⁰_k) : ∇G_k. Its fast
factor 𝒥 - C⁰_k has exact zero cell mean (LeftJacobian.cellFluxCell_fastFactor_mean_zero). This
corrects (𝒥 - C_k) : ∇G_k, whose slow part misses the left Jacobian.
Equations
- One or more equations did not get rendered due to their size.
Instances For
The transition-window remainder in divergence form.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Sum of the nine residual terms, using the line-2 ansatz.
Equations
- One or more equations did not get rendered due to their size.
Instances For
The Section 5.1 target proposition with the named residual terms. The source-specific theorem establishing it is kept separate from this statement.
Equations
- One or more equations did not get rendered due to their size.