JOVANA
Explore Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
All guides

Metallography: Seeing Grains with Light

For the whole ladder we argued that structure sets properties. This opener finally lets us SEE that structure: how sectioning, grinding, polishing, and etching turn a dull chunk of metal into a landscape of grains under an ordinary light microscope — and how to put an ASTM number on what we find.

Why look at all

Every rung so far has rested on one promise: structure sets properties. We said grains are small when a metal is strong, that pearlite is fine when a steel is quenched fast, that a crack starts at the worst flaw. But how does anyone actually KNOW a grain is 20 microns across, or that a casting has porosity, or that a weld cooled too fast? They looked. This whole rung is about the instruments that turn our confident stories into measured fact, and it opens where the field itself began — with light. Metallography is simply the craft of preparing a piece of metal so its internal microstructure can be seen and measured under a microscope. It is the oldest tool in the box and still the first one you reach for.

There is a real payoff waiting, not just a pretty picture. Recall the grain boundary from the defects rung — the mismatched seam where two crystal patches meet like floor tiles laid at different angles. Recall too that packing those seams closer together, by making grains smaller, makes a metal stronger: that is the Hall-Petch relationship, strength rising roughly as one over the square root of grain size. So if we can see and count the grains, we can predict the strength before ever pulling a tensile bar. Metallography is not decoration; it is the bridge from a picture to a number to a property.

The light microscope, and where its eyesight ends

A biology microscope shines light THROUGH a thin, transparent slice. A metal is opaque, so metallography uses a reflected-light (metallurgical) microscope instead: light comes down through the objective lens, bounces off the polished surface, and returns through the same lens to your eye. Everything we see is therefore a map of how the flat surface reflects — which spots send light straight back and look bright, and which tilt or scatter it and look dark. That single fact is the key to the whole method, and we will use it in a moment to make grains appear.

But light has a hard ceiling on how fine a detail it can resolve, and being honest about it is what motivates the next four guides. Two features closer together than roughly half the wavelength of the light blur into one; visible light runs about 400 to 700 nm, so the finest thing a perfect optical microscope can separate is around 0.2 microns, that is 200 nm. In practice useful magnification tops out near 1000 to 1500 times — beyond that you get a bigger, blurrier image, not a sharper one, so-called empty magnification. That is plenty to see grains, phases, cracks, and porosity, all of which are microns to tens of microns in size. It is nowhere near enough to see a single dislocation or the nanometre-scale precipitates that do so much of the strengthening.

Getting to a mirror: sectioning, mounting, grinding, polishing

Here is the counter-intuitive heart of the craft: before you can see the structure, you must destroy all evidence of your own handling. Any saw cut, scratch, or rough grind smears a thin, badly deformed layer over the true surface — a cold-worked skin that hides the very grains you came to see. So specimen preparation is a patient march from coarse to fine, each step removing the damage left by the one before, until the surface is a flat, scratch-free mirror that reflects the microstructure honestly rather than your tool marks.

  1. Section: cut a small representative piece from the part, using a water-cooled abrasive wheel so friction heat does not alter the very microstructure you want to inspect.
  2. Mount: embed the little specimen in a puck of plastic (hot-pressed or cold-cured resin) so it is easy to hold and grip, and so the edges stay flat instead of rounding off.
  3. Grind: rub the face on progressively finer silicon-carbide papers (say 240, then 400, 600, 1200 grit), rotating the specimen 90 degrees each step so you can see the previous set of scratches vanish.
  4. Polish: finish on a soft cloth charged with a fine slurry — diamond paste down to 1 micron, then perhaps 0.05-micron alumina — until, under the microscope, the unetched surface is a featureless silver mirror.

A well-polished but unetched surface looks almost blank on purpose — a clean mirror shows no grains. That blankness is actually good news: it means the deformed layer is gone and the surface is flat enough to reflect a true image. It is also the right state for spotting things that are already contrast without any chemistry: cracks, non-metallic inclusions, graphite flakes in cast iron, and porosity all show up on the as-polished mirror because they are holes or foreign particles, not crystal orientation. Everything else — the grains themselves — is still invisible, and revealing it takes one more deliberate step.

Etching: how a chemical makes grains appear

The trick that finally reveals grains is etching: wiping or dipping the mirror in a mild chemical reagent (for plain steel, the classic is nital — a few percent nitric acid in alcohol) for a few seconds. The etchant attacks the surface unevenly, and that unevenness is what our reflected-light microscope can read as contrast. Two different mechanisms do the work, and both trace straight back to structure we already understand.

First, grain boundaries etch fastest. A boundary is a region of higher energy — atoms there are mismatched and loosely held, exactly the disordered seam between mismatched tiles — so the acid eats it preferentially and cuts a tiny V-shaped groove along every boundary. Those grooves tilt light away from the lens, so they come back dark: the grains show up as bright patches outlined by a dark net of boundaries. Second, the grain interiors etch at different rates depending on their crystal orientation. Because a polycrystal is a mosaic of grains each pointing a different way (that is anisotropy at work), each grain's exposed face reflects a slightly different brightness, so neighbouring grains read as different shades of grey. Between the two effects, a blank mirror blooms into a visible landscape of grains.

  WHY AN ETCHED GRAIN BOUNDARY LOOKS DARK   (polished + etched cross-section)

    light in           light in                light in
       |                  |                        |
       v                  v                       V-groove the etchant
   ====flat====       ====flat====             \   cut along the
    grain A            grain B                   \ / high-energy boundary
   (reflects           (reflects                  V  scatters light aside
    straight up        straight up,                   -> reads DARK
    -> BRIGHT)          slightly tilted -> GREY)

   as-polished, unetched:  clean mirror, almost featureless
   after etching:          bright grains + dark boundary net + grey shading

  ASTM grain-size number n:   N = 2^(n-1)   (N = grains per sq. inch at 100x)

     n = 1  ->    1 grain / in^2     COARSE   -> fewer boundaries, weaker
     n = 4  ->    8 grains / in^2
     n = 8  ->  128 grains / in^2    FINE     -> more boundaries, stronger
Top: the etchant grooves the high-energy boundaries so they scatter light and look dark, while differently oriented grains reflect slightly differently and read as shades of grey. Bottom: the ASTM grain-size number n packs all of that into a single figure — bigger n means finer grains and, via Hall-Petch, higher strength.

The same etch also separates phases, not just grains — which is how you tell a steel's structure at a glance. In a slowly cooled 0.8 percent carbon (eutectoid) steel, the microstructure is all pearlite: alternating lamellae of soft ferrite and hard cementite. Nital attacks the two at different rates, so under the microscope pearlite shows the fingerprint of fine dark-and-light stripes, and a lower-carbon steel shows white ferrite grains sitting beside those pearlite colonies. Reading those areas back through the lever rule you met in the phase-diagram rung turns a picture straight into a carbon content — a genuinely useful piece of forensic metallurgy.

Putting a number on it: the ASTM grain-size number

Once boundaries are visible, we quantify them. The standard shorthand is the ASTM grain-size number n, defined so that N, the number of grains counted in one square inch of image at 100 times magnification, equals 2^(n-1). Turn that around: n = 1 gives 1 grain per square inch (coarse), n = 4 gives 2^3 = 8, and n = 8 gives 2^7 = 128 grains crammed into the same area (fine). Every step up in n doubles the grain count and shrinks the grains. Because it is a single tidy integer, it travels well: a spec can simply demand 'ASTM 7 or finer' and any lab in the world can check it.

Now the loop closes. A higher n means finer grains, which by Hall-Petch means a higher yield strength — so this one measured number feeds a real property prediction, exactly the bridge promised at the start. But be honest about what the number is not. It is a two-dimensional slice through a three-dimensional foam of grains, so a flat section systematically under-measures the true grain size (a random cut rarely passes through a grain's widest point), and careful work uses stereology or the line-intercept method to correct for it. And grain-size strengthening carries the caveat from the deformation rung: fine grains help at room temperature but HURT at high temperature, because grain boundaries slide and become the weak, creeping paths — which is precisely why turbine blades are grown as single crystals with no boundaries at all. A grain-size number is powerful, but it is a room-temperature virtue, not a universal one.

What light cannot see: the rest of this rung's toolbox

Light has taken us far — grains, phases, cracks, porosity, a strength prediction — but its 200 nm ceiling and shallow depth of field draw a firm boundary around what it can do, and the other four guides are each a way past that boundary. When you need to see the rugged terrain of a fracture surface, or features finer than light allows, you go to electrons: the scanning electron microscope gives deep-focus, three-dimensional-looking images of surfaces and fractures (and, with an EDS detector bolted on, tells you the local composition), while the transmission electron microscope fires electrons THROUGH a foil thin enough to reveal individual dislocations and nanometre precipitates. That fracture-reading skill even has its own name, fractography, and it is the heart of guide 2.

The remaining tools answer questions a picture cannot. To learn which crystal structure or phase is actually present, you diffract X-rays off the atomic planes and read the angles through Bragg's law — that is guide 3, alongside the spectroscopies that fingerprint chemistry. To find the temperatures where a material transforms — a polymer's glass transition, a metal's melting or a phase change — you heat a tiny sample and watch the heat flow with differential scanning calorimetry, the subject of guide 4. And to inspect a real finished part without cutting it apart, guide 5 covers nondestructive testing — ultrasound, radiography, dye penetrant, eddy current — closing with the honest engineer's question that ties the rung together: given a specimen and a question, which tool is the right one?