How to Select the Right CBN Grinding Wheel for Hardened Steel

Diamond is harder, but diamond is carbon. Above roughly 700 °C at the grinding interface, carbon diffuses into iron and the diamond graphitizes, so it chemically degrades against ferrous material. CBN has no equivalent reaction with iron and remains stable to much higher interface temperatures. CBN also has high thermal conductivity, which moves grinding heat away from the part. That combination is why CBN is the standard superabrasive for hardened alloy steel, tool steel, die steel, high speed steel, and bearing steel, and why diamond is used for carbide, ceramics, glass, and composites.

As a working rule, above roughly 45–50 HRC. Below that, conventional aluminum oxide removes material economically enough that the CBN premium is hard to recover. Above 55 HRC the case for CBN is usually straightforward, and above 60 HRC it is difficult to justify anything else in production volume.

Work backwards from the required surface finish, then check that it supports your stock removal. B107–B126 (140/170 to 120/140 mesh) is a reasonable general-purpose starting point, giving roughly 0.3–0.8 µm Ra. Go to B64–B76 for fine finishing around 0.15–0.4 µm Ra, and to B151–B181 for roughing. Achievable finish also depends on dressing parameters, speed ratio, and spark-out — expect to move a factor of two at constant grit by changing those.

C100 is the general-purpose default. Move to C125–C150 for vitrified production grinding, higher removal rates, and form retention. Move to C75–C100 for resin bond finishing and tool grinding. Higher concentration lowers force per grit and extends life but reduces chip clearance and coolant access, so it is not a heat remedy.

There is no universal answer, but there is a usable default: vitrified for production grinding of hardened steel on a machine with rotary dressing. It gives the best combination of form holding, dressability, wheel life, and coolant access.

Resin bond is the choice for tool grinding, fine finishing, less rigid machines, and shops whose only conditioning capability is a brake-controlled truer and dressing sticks. Sintered metal bond is for deep forms and heavy removal on rigid machines with proper dressing equipment. Hybrid sits between resin and metal. Electroplated single-layer is for complex forms and low-force applications where no dressing is possible or wanted.

The constraint that most often decides it is not performance — it is what your shop can true and dress.

Within the wheel’s rated maximum operating speed, your spindle rating, and your guard rating — all three. Typical operating ranges: resin bond 25–45 m/s, vitrified 45–80 m/s in standard construction, sintered metal 20–35 m/s, electroplated 60–160 m/s. Higher wheel speed reduces chip thickness per grit, which improves finish and wheel life but increases the energy going into the arc. Raise wheel speed and coolant delivery together, never separately.

Burn comes from thermal load the coolant cannot remove. The usual causes, in order of frequency: a dull or loaded wheel, coolant not reaching the arc, coolant concentration or filtration drift, Q’w above what the system supports, speed ratio too high, dressing parameters producing a closed wheel, and — last — an incorrect wheel specification.

Detect it with a nital temper etch procedure (ISO 14104 or SAE AMS2649), Barkhausen noise for non-destructive 100% inspection, or X-ray diffraction residual stress measurement for definitive results. Visual inspection is not detection; rehardening burn produces a harder surface than specification and will pass a hardness check.

For production CBN grinding of hardened steel, straight oil is technically superior — significantly better lubricity, lower grinding forces, lower burn risk, and typically 1.5–3× the wheel life. It is standard in high-volume bearing, gear, and automotive grinding for that reason.

Water-soluble fluid is the correct choice when fire risk, misting, ventilation, machine compatibility, or plant policy rules oil out. If you run water-soluble, hold concentration at 5–10% and verify it with a refractometer on a schedule.

Whichever you run, jet velocity matters more than flow volume. Aim for 0.8–1.0 × wheel speed through a coherent-jet nozzle, aimed into the arc.

There is no universal value, and any supplier who gives you one without asking about your process is guessing. Wheel life depends on material and hardness, stock removal, wheel specification, machine condition, dressing parameters, coolant, and operating conditions.

Measure it as acceptable parts produced inside the process window, not as hours or as abrasive layer consumed. Record G-ratio alongside it for wheel-to-wheel comparison, and record dressing frequency and cycle time, because those often drive cost more than wheel wear does.

Same workpiece material and hardness, from the same heat treat lot if possible. Same machine, same coolant condition, same fixturing. Controlled speed, feed, depth of cut, and dressing parameters. Same inspection method and measurement points.

Run enough parts to cross several dressing cycles. Record parts per dress, parts per wheel, radial wear, cycle time, finish trend across the dress cycle, dimensional and form accuracy, spindle power, temper etch results, and scrap. Then calculate G-ratio and cost per acceptable part including machine time. Never compare on purchase price.

CBN is suited to hardened ferrous materials — hardened alloy steels, tool steels, die steels, high speed steels, bearing steels, and hardened stainless grades. On soft or annealed steel it works but is rarely economical; conventional abrasive is usually the better choice below about 45 HRC. Confirm actual hardness and heat treatment before specifying, and give your supplier the grade rather than a category.

Yes. We produce custom diamond and CBN wheel specifications to application requirements and to customer drawings — non-standard diameters, rim widths, abrasive layer depths, bond grades, grit and concentration combinations, core materials, mounting configurations, and multi-radius or dedicated profiles.

Custom specification is worth pursuing when a standard wheel cannot meet the profile, the machine interface, the speed requirement, or the process window. Send the application data and, where a form is involved, the wheel drawing.