Power Laws Variance Scaling of Boolean Random Varieties

Dominique Jeulin (1)

1 CMM - Centre de Morphologie Mathématique

Long fibers or stratifed media show very long range correlations. These media can be simulated by models of Boolean random varieties and their iteration. They show non standard scaling laws with respect to the volume of domains K for the variance of the local volume fraction: on a large scale, the variance of the local volume fraction decreases according to power laws of the volume of K. The exponent is equal to (n-k)/ n for Boolean varieties with dimension k in the space Rn : 2/3 for Boolean fibers in 3D, and 1/3 for Boolean strata in 3D. When working in 2D, the scaling exponent of Boolean fibers is equal to 1/2. From the results of numerical simulations, these scaling laws are expected to hold for the prediction of the effective properties of such random media.

Mots clés: Integral range; Long range correlations; Scaling law; Numerical homogenization; Boolean model; Random fiber networks; Random strata; Poisson varieties; RVE

Domaines: Traitement du signal et de l'image [eess.SP]; Mathématiques [math]

Dominique Jeulin. Power Laws Variance Scaling of Boolean Random Varieties. Methodology and Computing in Applied Probability, 2015, ⟨10.1007/s11009-015-9464-5⟩. ⟨hal-01207456⟩