ABSTRACT
Here we present an experimental and numerical investigation on the grain-scale geometrical and mechanical properties of partially crystallized structures made of macroscopic frictional grains. Crystallization is inevitable in arrangements of monosized hard spheres with packing densities exceeding Bernal's limiting density Ï(Bernal)≈0.64. We study packings of monosized hard spheres whose density spans over a wide range (0.59<Ï<0.72). These experiments harness x-ray computed tomography, three-dimensional image analysis, and numerical simulations to access precisely the geometry and the 3D structure of internal forces within the sphere packings. We show that clear geometrical transitions coincide with modifications of the mechanical backbone of the packing both at the grain and global scale. Notably, two transitions are identified at Ï(Bernal)≈0.64 and Ï(c)≈0.68. These results provide insights on how geometrical and mechanical features at the grain scale conspire to yield partially crystallized structures that are mechanically stable.
ABSTRACT
In this paper, we study many geometrical properties of contour loops to characterize the morphology of synthetic multifractal rough surfaces, which are generated by multiplicative hierarchical cascading processes. To this end, two different classes of multifractal rough surfaces are numerically simulated. As the first group, singular measure multifractal rough surfaces are generated by using the p model. The smoothened multifractal rough surface then is simulated by convolving the first group with a so-called Hurst exponent, H*. The generalized multifractal dimension of isoheight lines (contours), D(q), correlation exponent of contours, x(l), cumulative distributions of areas, ξ, and perimeters, η, are calculated for both synthetic multifractal rough surfaces. Our results show that for both mentioned classes, hyperscaling relations for contour loops are the same as that of monofractal systems. In contrast to singular measure multifractal rough surfaces, H* plays a leading role in smoothened multifractal rough surfaces. All computed geometrical exponents for the first class depend not only on its Hurst exponent but also on the set of p values. But in spite of multifractal nature of smoothened surfaces (second class), the corresponding geometrical exponents are controlled by H*, the same as what happens for monofractal rough surfaces.
Subject(s)
Fractals , Models, Chemical , Models, Molecular , Nanoparticles/chemistry , Surface Properties , Computer Simulation , Hardness , Molecular ConformationABSTRACT
We investigate the Markov property of rough surfaces. Using stochastic analysis, we characterize the complexity of the surface roughness by means of a Fokker-Planck or Langevin equation. The obtained Langevin equation enables us to regenerate surfaces with similar statistical properties compared with the observed morphology by atomic force microscopy.