Grinding Cracks and Grinding Burn in Hardened Tool Steel

How grinding changes the surface

Almost all of the energy used in grinding is converted into heat, part of it through friction and part of it through deformation of the material. If a correct grinding wheel has been chosen, most of the heat is carried away with the chips and only a smaller part heats the workpiece. Incorrect grinding of a hardened tool steel can raise the temperature of the ground surface past the tempering temperature of the material, and the hardness of the surface falls. The tool then performs poorly in service even though the rest of it is sound. Allow the temperature to rise further and it reaches the hardening temperature of the steel, and the surface is rehardened.

Rehardening is the more serious of the two outcomes. The steel in the affected layer was austenitized and then quenched by the cold mass of the tool behind it, so the layer ends up as a mixture of untempered martensite and tempered martensite together with retained austenite. Untempered martensite is hard and brittle, and in the high carbon cold work and high speed grades it is the most brittle product the steel can form. The layer also carries microcracks of its own, and those act as stress raisers the moment the tool goes to work.

The white layer and the zone under it

After metallographic preparation and etching, the rehardened layer shows white under the optical microscope, because untempered martensite resists the etchant more than tempered martensite does. The same appearance is produced by the recast layer left by electrical discharge machining, and the two are different damage with different fixes. The hardness profile through the surface is what separates them, and it also shows why the ground surface behaves badly. The surface itself is hard, because it is untempered martensite. Immediately below it sits an overtempered zone, where the grinding heat was enough to temper the steel further but not enough to reharden it, and there the hardness is below the hardness of the body of the tool.

The three zones produced when a hardened tool is ground too hot. The white layer and the overtempered band under it are often present in the same tool, which is why a hardness reading taken on the surface can look acceptable while the tool fails from the zone below it.

Why grinding cracks form and why they are hard to see

The majority of grinding operations leave residual stresses in the ground surface, and the stresses are usually at a maximum close to the surface. The first and most common effect of those stresses is cracking. On thin sections the same stresses can deform the part permanently, and retained austenite in the rehardened layer adds to the movement. Grinding cracks are usually not as obvious as the examples printed in textbooks. In most cases the part has to be examined under a microscope, or with magnetic particle inspection, before the cracks can be seen at all, and a tool can pass a visual check and still be cracked.

The visible warning is the burn mark. Incorrect grinding that has modified the surface layer usually shows itself as a discoloration of the ground surface. A burn mark means the surface temperature passed the tempering range, and it is the point at which the tool should be examined before it goes into service.

The three conditions a ground surface can be in. A clean appearance does not clear the part, because the cracks that end tool life are usually below the resolution of the unaided eye.

Which grades carry the most risk

Hardenable steels are more prone to grinding cracks than low carbon, low alloy steels, and inside the tool steels the risk is concentrated in the grades that are hardest as quenched. Cold work tool steels and high speed steels are the most sensitive of all, because the as-quenched martensite is very hard and its toughness is low. Carbon content is the property that decides how brittle the rehardened layer becomes. Martensite hardness rises quickly with carbon, so the high carbon cold work grades form a layer that is both harder and more brittle than anything a lower carbon tool steel can produce.

Two failures in the source show the pattern. An S1 cutter die cracked and spalled after grinding, and on a D2 die the crack was assisted by incorrect grinding together with a row of closely spaced holes. The same class of brittle surface layer, formed by electrical discharge machining instead of by the wheel, cracked an A2 blanking die in service after it had produced twenty thousand parts.

Grinding damage risk across the tool steel groups. Carbon content is taken from the typical compositions listed by the source, and the sensitivity it reports for the cold work and high speed families sets the upper band. The lower carbon groups are the ones the source does not single out as sensitive.

Three ways to reduce grinding damage

The first is to control the heat. Good cooling and properly dressed grinding wheels that cut the material with sharp cutting edges, instead of rubbing and generating heat through friction, keep the temperature of the ground surface down. The second is stress relief after grinding. A tempering treatment at about 30 °C below the previous tempering temperature relieves the grinding stresses and also tempers any untempered martensite that the grinding formed, and because it stays below the original tempering temperature it does not reduce the hardness of the workpiece. The third is to tumble or blast the ground parts, which lowers the surface stress mechanically.

All three have a limit. If the heat damage is high enough that cracks have already formed, stress relief does not help, and the damaged layer has to be removed.

The three remedies given by the source, with the condition under which each one stops being effective.

The link back to tempering

Grinding cracks are far more likely when the tool was not tempered properly in the first place. A study of a worn camshaft in the source measured a surface layer whose hardness was close to the as-quenched condition, and it records that condition as deleterious because worked surfaces with that structure are more prone to grinding cracks. The surface of a tool that was never tempered properly is already brittle before the grinding wheel touches it, and the grinding stresses are enough to finish the job.

The rule that follows is to temper fully, and to temper twice. No operation should be carried out with the steel in the as-quenched condition, and the time at the tempering temperature matters as much as the temperature itself.

Where this leaves a tool steel enquiry

How hard a grade is to grind, and the wheel that suits each group, is on the tool steel grindability and wheel selection page. The same brittle layer produced by a spark is worked through on the EDM white layer page. Cracks that opened during the quench rather than at the wheel are on the quench cracking page, and the time a section needs to reach the tempering temperature is on the tempering time page. The inspection method that finds grinding cracks is on the magnetic particle inspection page, and the wider set of failure causes is on the tool and die failure analysis page.

Source, Failure Analysis of Heat Treated Steel Components (L. C. F. Canale, R. A. Mesquita and G. E. Totten), ASM International, 2008.