GEOMETRICRIGIDITYOFCONFORMALMATRICES1.PDFVIP

GEOMETRIC RIGIDITY OF CONFORMAL MATRICES DANIEL FARACO AND XIAO ZHONG Abstract. We provide a geometric rigidity estimate `a la Friesecke-James-M¨uller for conformal matrices. Namely, we re- place SO(n) by a arbitrary compact subset of conformal matrices, bounded away from 0 and invariant under SO(n), and rigid mo- tions by M¨obius transformations. 1. Introduction This paper is concerned with the so-called geometric rigidity esti- mates for conformal matrices. Recently, Friesecke, James and M¨uller developed a successful new approach to the classical problem of dimen- sion reduction in nonlinear elasticity [8], [7]. A fundamental ingredient was the following rigidity estimate for the group of special orthogonal n n ×n t matrices of R , SO(n) = {A ∈ M : A A = I, detA = 1}. Theorem 1.1. Let Ω ⊂ Rn be a bounded Lipschitz domain and n ≥ 2. There exists a constant C = C (Ω) with the property that for each 1 1 v ∈ W 1,2 (Ω, Rn ), there exists R ∈ SO(n) such that (1.1) Dv − R 2 ≤ C dist (Dv) 2 . L (Ω) 1 SO(n) L (Ω) Theorem 1.1 has been used after in a number of related problems concerning dimension reduction, e.g [3], [6], [18] and [17]. In all the ap- plications it is crucial that the dependence between the left and right hand side is linear and that the v is any general Sobolev mapping (the classical result of John [15] gives an L2 − L∞ estimate valid for locally Bi-Lipschitz maps). Theorem 1.1 makes quantitative the following clas- sical result of Reshetnyak [19] for sequences. Theorem 1.2. Let {v } ∈ W 1,2 (Ω) be a weakly convergent sequenc