stereology

A microscope shows a flat, polished slice through a solid — a 2D window into a 3D world. Stereology is the mathematics that lets you infer the real three-dimensional structure from those two-dimensional sections, honestly and quantitatively. It answers questions like: what volume fraction is this phase, how much grain-boundary area is there per unit volume, how big are the grains — all from measurements made on a flat cut.

The power of stereology is a handful of exact, assumption-light relations between what you count on the plane and what is true in the volume. Volume fraction equals area fraction equals point fraction (Vv = Aa = Pp) — count points on a phase to get its volume fraction. Surface area per unit volume equals twice the number of interface intersections per unit length of test line (Sv = 2 PL) — count how often random lines cross grain boundaries to get grain-boundary area per volume. Mean grain size follows from the mean intercept. These hold for any shape, needing only that the sampling be random and unbiased.

Stereology is the backbone of quantitative metallography: it turns qualitative pictures into numbers you can put into property models and quality specs, and it is why grain size, phase fraction and interface density can be measured at all from ordinary micrographs. Its central honest lesson is a warning about naive reading: a 2D section systematically distorts apparent sizes — a random plane rarely cuts a grain across its widest part, so grains look smaller and size distributions look wrong unless you apply the stereological correction. Seeing is not measuring.