Hardness is a specification, not a description
A hardness figure is a claim about a measurable property, and like any measurement it only means something when the scale, the method and the location are stated. A drawing that says 'harden' has said nothing a heat treater or an inspector can work to, and two suppliers reading it will deliver two different parts.
The practical minimum is a scale, a value and a range, applied to a defined area. Where hardness varies across a part - a case-hardened surface over a softer core, for instance - the surface figure and the core figure are different requirements, and both belong on the drawing rather than one standing in for the other.
The scales, and why the number alone is meaningless
Hardness is measured by pressing an indenter into the surface and reading the result, and different indenters and loads give different scales. Steel is usually quoted on the Rockwell C scale, softer metals on Rockwell B or on Vickers and Brinell, and the numbers are not interchangeable between scales.
That is why a value without a scale is unusable. A '60' is a different material on each of the three common scales, and an inspector who has to guess is being set up to fail. Writing 'hardness 58-62 HRC' costs four extra characters and removes the ambiguity entirely.
Case hardening, through hardening and nitriding
Through hardening treats the whole section: the part is heated and quenched so that hardness runs all the way through, then tempered to a working level. It suits parts where the whole body carries load, and it is the simplest requirement to specify because there is one figure to hit.
Case hardening leaves a hard, wear-resistant skin over a tough core. The requirement therefore has three parts - case depth, surface hardness and core hardness - and all three matter. Carburising and induction hardening produce different case profiles, so naming the process is usually clearer than naming only the result. Nitriding is a third route that adds a very hard, thin layer at a lower temperature, which distorts the part far less, at the cost of a longer cycle.
Distortion is the reason the sequence matters
Heat treatment moves metal. Quenching from temperature introduces stress that relieves itself as distortion, and a part that was dimensionally perfect before treatment may not be afterwards. The amount of movement depends on section thickness, symmetry, material and process - and it is not zero for any of them.
The consequence for a drawing is the sequence. One route is to machine oversize, treat, then finish to size by grinding or a light machining pass on the dimensions that matter. The other is to treat first and machine after, which is only possible when the material is still machinable at that hardness. Choosing between them is what decides whether a tight tolerance is achievable at all.
Machining before or after treatment
The practical rule is that anything above roughly 45 HRC stops behaving like a free-cutting material. Tool life falls away, cutting forces rise, and the surface finish that comes off the cutter degrades. Some features - sharp internal corners, fine slots, thin walls - become impractical by milling at all.
That is the point at which the process sequence has to be designed rather than assumed. Machining soft then hardening then grinding the critical fits is the conventional answer for hardened parts. Where the geometry rules grinding out - a sharp internal corner, for example - the hardened feature is produced by EDM instead, which cuts hardened steel without the cutting forces.
Hardness testing, and what a certificate proves
Hardness is verified by measurement, and the useful specification says where and how. A single test on a convenient face proves very little about a part whose critical surface is elsewhere; a requirement of so many tests at defined locations proves more. For a case-hardened part, the case depth is normally established from a sectioned sample rather than from a surface reading.
A material certificate and a heat-treatment certificate are records, not guarantees of the finished part. Where the hardness of the delivered component matters, the useful document is a test report on the treated parts, with the instrument identified and its calibration traceable - which is what a metrology framework such as NIST's is for.
References
Hardness scales and testing methods are defined by published standards, and the measurement methods most specifications cite are issued by ASTM International. The underlying property is the resistance of a material to localised deformation, described under hardness, and the traceability of the measurement itself rests on the framework published by NIST.
Frequently asked
Should I machine before or after heat treatment?
Machine soft, then treat, then finish the critical dimensions by grinding or by a light cut if the hardness allows it. That order protects the tolerances. Where the geometry rules grinding out - a sharp internal corner, a fine slot - the hardened feature is produced by EDM after treatment instead.
Why does my part distort during hardening?
Because quenching a hot part introduces stress, and stress relief moves material. Section thickness changes, asymmetry and the choice of process all affect how much. The practical answers are to design for symmetry where you can, to leave stock for a finishing operation, and to agree the sequence with the heat treater before the first part is cut.
Is a material certificate the same as a hardness certificate?
No. A material certificate describes the stock the part was made from. A hardness certificate reports measurements on the treated part. If the hardness of the finished component is a requirement, ask for the test result on the parts themselves, with the instrument and its calibration identified.