High-contrast homogenization: #
Provider.Regularity.CenteredCubeDirichletNormalizedEuclideanL2Residual
Imported from the Apache-2.0 HighContrastHomogenization development at commit
7a13dbcd8d6609264a713373f5c69ceeac870472.
Normalized Euclidean L2 residual stability for the identity Dirichlet problem #
The physical normalized Euclidean L² endpoint is stable under subtraction:
the difference of two identity Dirichlet solution gradients is controlled by
the difference of their data with constant one.
theorem
HCPolySupport.HighContrast.centeredCubeDirichletDivergence_normalizedEuclideanLpENorm_grad_sub_le
{d : ℕ}
{m : ℤ}
(h k : CenteredCubeEuclideanL2Field d m)
(w v : H10Function (openCubeSet (originCube d m)))
(hw : CubeDirichletDivergenceProblem (originCube d m) w h.toField)
(hv : CubeDirichletDivergenceProblem (originCube d m) v k.toField)
:
((centeredCubeDomain d m).normalizedEuclideanLpENorm 2 fun (x : Vec d) => w.grad x - v.grad x) ≤ (centeredCubeDomain d m).normalizedEuclideanLpENorm 2 fun (x : Vec d) => h.toField x - k.toField x
Two identity Dirichlet solutions are nonexpansive on differences in the
physical normalized Euclidean L² norm.