Cemented Carbide Substrates

Cemented carbide is tungsten-carbide particles held together by cobalt. Grain size and cobalt content are the two dials: fine grain and low cobalt resist wear but are brittle; coarse grain and higher cobalt absorb shock but crater sooner. This page reads those dials and matches a grade family to a job.

Concept

Cemented carbide is a cermet-like composite: hard tungsten carbide (WC) grains are sintered with a cobalt (Co) metal binder. The WC grains provide hardness and wear resistance; the cobalt binder provides toughness by holding the grains together. Manufacturers vary two main levers — grain size (fine to coarse) and cobalt content — and add titanium/tantalum carbides for steel-machining grades.

Why It Matters

Not all carbide is the same. A “wear” grade for finishing aluminum and cast iron is brittle and will chip on a steel interruption; a “tough” grade for roughing steel will crater too fast in a long finishing pass. The grade designation in the catalog is the manufacturer's position on these two dials. Reading that position tells you whether the insert can survive your interruption and hold up to your speed.

How It Works

Grain size

Cobalt content

More cobalt = tougher and shock-absorbing, but softer at high temperature and more prone to crater/wear. Less cobalt = harder and more wear-resistant at temperature, but more brittle. Steel grades also add cubic carbides (TiC/TaC/NbC) to resist cratering from the long chip contact in steel turning.

The two broad grade families

ISO groups cutting materials by application: P = steel, M = stainless steel, K = cast iron, N = non-ferrous metals (aluminum, copper), S = heat-resistant superalloys and titanium, H = hardened material. A single grade number may cover more than one of these groups, and the letter is not a chemical recipe or a ban on the others — it marks the application the grade is tuned for, so confirm against the specific product range.

Worked Example

A shop runs two operations on the same part. (1) Finish boring a steel bore at high speed to a tight tolerance — continuous cut, needs edge retention: a fine-grain, lower-cobalt, steel grade with coating resists crater wear. (2) Rough facing a forging with heavy scale and interrupted entry — shocks the edge: a tougher, higher-cobalt grade survives the impacts even though it wears a little faster. Choosing one grade for both either chips the finishing edge or wears out the roughing edge.

What grain size buys you

Cemented carbide is hard tungsten-carbide particles held in a cobalt binder. Fine-grain grades wear slower and hold a sharper edge, but are more brittle; coarse-grain grades are tougher for interrupted cuts and roughing. More cobalt binder adds toughness but reduces hardness and hot strength. The substrate is the "body" onto which a coating is applied (page 12), so choosing it is a toughness-vs-wear trade: roughing and interrupted cuts → tougher, coarser; finishing hard material → finer, harder. Most shop-floor carbide is graded by application (steel / stainless / cast iron) precisely so you do not have to derive the grain-binder balance yourself — the catalog already encodes it.

Reading the ISO P/M/K label

Catalogs group carbide by the ISO application class rather than by grain size directly, because you rarely set cobalt or grain yourself. The P class is for steel (long chip, abrasive cratering), M for stainless and heat-resistant alloys (work-hardening, gummy), and K for cast iron and non-ferrous (short chip, abrasive). Within each band the catalog number moves from wear-resistant toward shock-resistant as the digits climb. The practical read: a P-class insert is built with cubic carbides to fight cratering from the long steel chip, which is why it should not be used on cast iron and vice versa — the K-grade's wear-toughness balance is tuned for a short, abrasive chip. When you are choosing a grade, first pick the letter that matches the workpiece family (steel, stainless, or cast iron/non-ferrous), then choose wear vs toughness within that letter by operation (roughing interruption vs continuous finishing). This two-step read is how a catalog page becomes a defensible choice instead of a guess, and it connects directly to the substrate dials on this page.

Why the substrate sits under the coating

A coating (page 12) is only a few micrometres thick, so it cannot change the bulk toughness of the insert — it adds a hard, low-friction, heat-shielding skin on top of whatever body lies beneath. That is why pairing matters: a hard coating on a brittle, low-cobalt substrate helps a continuous finishing cut where the edge never sees a shock, but if the same coated insert meets an interrupted forging entry the coating and substrate chip together, because there was no toughness underneath to absorb the impact. Conversely, a thick CVD coating on a very tough high-cobalt roughing grade gives both the shock absorption and the heat shield needed for heavy steel roughing. The practical rule is to choose the substrate for the operation's toughness demand first, then add the coating for the speed/heat demand on top — never expect a coating to rescue a grade that is the wrong side of the wear/toughness balance. This is also why "more coating" is not the universal upgrade: it is a skin, and the body beneath it has to survive the interruption the cut actually delivers.

Why more cobalt helps toughness but limits speed

The cobalt binder is what makes carbide tough, but it is also the softest component in the insert: cobalt is a metal, and unlike the hard WC grains it softens and flows at cutting temperature. That is the whole reason the two dials fight each other. Raise the cobalt content and the insert absorbs shocks and interruptions beautifully, but at the high temperatures of a fast cut the binder softens, the WC grains pull out earlier, and the edge deforms or craters — so a tough roughing grade cannot be run at finishing speed. Lower the cobalt and the body keeps its hot hardness, holds an edge at speed, but becomes brittle enough to chip on an interruption. This is the physical reason behind the catalog's split into wear-resistant (low cobalt, fine grain, fast, brittle) and shock-resistant (higher cobalt, tougher, slower) grades, and why you cannot have both extremes in one insert. It also explains why a coating on top (page 12) does not remove this trade: it shields the edge from heat and friction, but the binder underneath still decides whether the body can survive the temperature you run. Choose cobalt for the shock your cut delivers, then rely on coating and speed to protect the edge between shocks.

Why steel grades carry cubic carbides and cast-iron grades do not

The reason a P-grade (steel) and a K-grade (cast iron) are built differently is the shape of the chip. In steel turning the chip slides a long way across the rake face at high temperature, and the iron in the chip reacts with plain WC, causing crater wear and diffusion. To resist that, steel substrates add cubic carbides — TiC, TaC, NbC — which are chemically stable against the long steel chip and slow cratering. Cast iron produces a short, broken chip with little sliding contact, so that diffusion threat barely exists; a K-grade instead optimises WC grain and cobalt for abrasion resistance against the graphite in cast iron and for the thermal shock of its interrupted-style cut. This is why you should not substitute a K insert into a long steel finishing pass and expect crater life: it lacks the cubic-carbide defence against the sliding steel chip, and it will crater far faster than a P-grade even though both look like identical carbide inserts. The ISO letter is therefore not a marketing label — it encodes which chemical wear mode the substrate is built to fight, and choosing it right is the cheapest tool-life decision on the page.

Why ultra-fine grain enables small tools

When carbide grain size drops below about a micrometre, a useful thing happens: the material can be made both harder and sharper without the usual brittleness penalty, because the fine grains let an edge be ground to a smaller radius while still holding together. That is why tiny drills, small end mills and precision boring bars are almost always ultra-fine-grain solid carbide — the edge has to be sharp to cut a small feature, and a coarse grade would either roll the edge over or break it. This is a different lever from the roughing-versus-finishing dial above: it is about geometric size, not chip load. When you move down to small-diameter tools, fine-grain substrates are commonly specified because the edge must be sharp and held precisely; confirm the exact grain and grade in the manufacturer's product data rather than assuming a coarse high-cobalt roughing grade can hold a usable edge on a Ø3 mm tool. Read the grain spec on a small tool as part of why it is sharp, and handle it accordingly.

Why a "universal grade" is a compromise, not the best

Shops often reach for one grade that "works on everything" to simplify inventory, and it is true that a mid-grade P insert will turn most steels acceptably. But that middle position on the wear-vs-toughness graph is a compromise on both axes: it is not quite tough enough for the heaviest interrupted roughing, and not quite hard enough for the longest, hottest finishing pass. The cost is hidden in tool life — the universal grade wears faster than a purpose-built finishing grade in a long run, and chips sooner than a purpose-built roughing grade on a forging scale. This is why page 29 counts grade choice as a cost-per-part decision, not a convenience decision: a slightly larger inventory of two matched grades usually beats one universal grade once the parts-per-edge difference is costed. The point is not that you must stock dozens of grades — it is that when an operation runs at high volume, moving from the universal middle to the purpose-built end of the scale often pays for itself in edge life, and the substrate dials on this page are exactly where that choice lives.

Why the same carbide can wear out in the corner first

When a tough, higher-cobalt grade softens at the edge, the failure often shows as plastic deformation rather than clean flank wear: the cutting edge itself deforms and rounds under heat and force, because the cobalt binder has flowed and let the WC grains displace. This is the signature that the substrate is running above its hot-hardness limit — not that the coating has worn. The response is not a new coating but a lower-cobalt, finer-grain or more heat-resistant grade, or a lower speed, because the body itself is deforming. Reading whether a worn insert shows uniform flank wear (abrasion, fix by speed/coating) or a squashed, deformed edge (thermal softening, fix by substrate/speed) is exactly why the substrate dials on this page matter: they decide how the edge fails, not just how long it lasts.

Grain size and binder content are why two "carbide" grades behave so differently. A fine grade wears well at speed but is more prone to chipping when interrupted; a coarser, tougher grade survives roughing but builds up edge wear sooner. Reading the substrate description on the datasheet explains which side of the trade-off you are buying.

When two carbide grades look similar on paper, the datasheet's grain-size and binder description is what separates them. A grade that is too fine for a roughing job will chip even though its wear number is attractive; read the recommended operation on the catalog rather than inferring it from hardness.

Common Mistakes

• Treating all carbide as the same. A finishing grade chips on roughing. Fix: pick wear vs toughness by operation, not just material.
• Running a tough roughing grade at high finishing speed. It craters. Fix: use a finer, more wear-resistant grade for continuous finishing.
• Ignoring the ISO application class (P/M/K). A K-grade for cast iron is not ideal for steel. Fix: match the grade class to the workpiece family.

On a small-diameter tool, confirm the actual grain size and binder in the product data rather than assuming a roughing grade can hold a sharp edge. If a new edge deforms plastically rather than wearing evenly, the substrate is running above its hot-hardness limit — the fix is grade or speed, not another coating layer.

Practice