EFFICIENT MEASUREMENT OF MICROSTRUCTURAL SURFACE-AREA USING TRISECTOR

It is shown that just three appropriately chosen metallographic sectio ns can yield an efficient and reliable estimate of the total area of t he internal microstructural boundaries per unit volume S, in a microst ructure having an arbitrary and unknown anisotropy and geometry. The f ollowing sampling scheme is required. (1) Choose a reference direction (called ''vertical axis'') such that most of the surface elements of interest are not parallel to the selected reference direction (vertica l axis). (2) Observe the microstructure on three metallographic sectio ns that are mutually at an angle of 120 deg and contain the vertical a xis. (3) Superimpose a set of cycloid shape test lines on these metall ographic sections (called vertical sections) such that the cycloid min or axis is parallel to the vertical axis, and estimate the average num ber of intersections between these test lines and the microstructural boundaries per unit test line length, [P(L)c]3BAR. The surface area pe r unit volume S(v) is given by S(v) = 2[P(L)c]3BAR To emphasize the pr eceding sampling requirements, the composite test probe is called ''tr i-sector.'' It is also shown that independent random sections are not useful for estimation of S(v) of anisotropic microstructural surfaces, in general. Further, two perpendicular vertical sections are also not generally useful for a reliable estimation of S(v) in anisotropic mic rostructures.