For perfect surface roughness description is not enough to know characteristics of surface profile. It is necessary to use topography methods, so called microtopography.
Thereby, surface roughness in microtopographycal understanding must be described with three coordinates, whose in Cartesian coordinates system compose point under consideration height h, abscissa and ordinate, determines point position in the plane. Most efficient methods in irregular surface roughness research are random function theory methods. Therefore, microtopography, analogically to profile, may consider as random function, but two dimensional function, i.e. two variable x and y random field h(x,y).
From analogy with random process, random field can be normal – ordinates are distributed by normal (Gaussian) distribution. Moreover, random field can be homogeneous and heterogeneous. Random field is deemed homogeneous if its mean value is discretionary and correlation function depends only from distance between surface points.
Important characteristic of random field is correlation function, whose depends of two variables t1 and t2 – orthogonal Cartesian coordinates of vector t.
Random field is homogeneous and isotropic when its characteristics are equivalent in any direction. There are three types of surface anisotropy:
• General event of surface anisotropy. Characteristics of this event roughness parameters are depend of surface split direction.
• Surface roughness with direct anisotropy. Those surfaces are with typical traces of tool and they proper two mutually perpendicular surface roughness directions.
• Extended anisotropy area – special event of anisotropy roughness. Of analytical opinion, gainfully anisotropy roughness see as extended occasional isotropy area. This let easy cross from anisotropy surface to isotropy and contrariwise, thereby embrace amount class of surface roughness.
Let’s formulate microtopography model of rough surface [1]. Surface roughness is described with homogeneous normal random field h(x,y) that has uninterrupted correlation function and uninterrupted deriviates. We may consider that E{h(x,y)}=0. The mean random field value is plane called mean plane.
For describing random field we must know mathematical expectation and field correlation function, what in fact reduces on determining dispersion and rationed correlation function r(t1, t2).
Homogeneous random field dispersion D{h} doesn’t depends of direction and can be founded in any surface split.
Given model of rough surface let inspect surfaces produced by abrasive instruments and friction surfaces after wear-in period.
STATISTICAL SIMULATION OF 2D RANDOM FIELD WITH CAUCHY CORRELATION FUNCTION IN THE GEOPHYSICS PROBLEM OF ENVIRONMENT MONITORING