Hardness testing: methods, scales and drawing callouts

What hardness measures

Hardness is the resistance of a material to local penetration. That definition already tells you why the number depends on the method. A sharp diamond pyramid pushed into a surface at a light load samples a small volume. A hard ball pushed at a heavy load samples a much larger one, including grain boundaries and inclusions. The two tests are related but not interchangeable.

That is why a certificate that says 45 HRC and a drawing that says 400 HV are not automatically in agreement. They may describe the same heat treatment, but the conversion between them is an estimate. Treat hardness as a test result with a method attached, and keep the method with the number wherever it travels.

Rockwell: the production test

Rockwell presses an indenter into the surface under a minor preload and then a major load, and reads the difference in depth. The result appears directly on the dial or the screen, which is why it dominates production checking. The common scales are C for hardened steel, B for softer steel and brass, and A for thin or case hardened parts.

Scale selection matters. The C scale with a diamond cone is for hard materials, roughly above 20 HRC. The B scale with a 1.6 mm ball is for softer metals and gives a different number for the same material. A callout of 30 HRC on a soft brass part is a mistake in the drawing, not a challenge for the lab.

• Check the section thickness against the depth of the indentation; a thin part can show the anvil underneath.
• Keep the indentation away from edges and from other indentations by the distance in the standard.
• Use a calibrated test block at the start and end of a session.
• State the scale, not just the number, on the drawing and the certificate.

Vickers: one scale, many loads

Vickers uses a square based diamond pyramid and measures the diagonals of the impression under a microscope. The result is a Vickers hardness number with the test load attached, such as 400 HV10 for a 10 kgf load or 350 HV0.3 for 300 gf. The geometry gives a continuous scale from soft to very hard, which is why it is used for research, for case depth traverses and for small features.

Because the load can be reduced, Vickers can measure a weld zone, a heat affected zone or a thin coating. The cost is time and surface preparation. A light load indentation is small, so the surface must be polished and the measurement of the diagonals is operator dependent. At very light loads the result is sensitive to grain orientation and to the polishing damage itself.

Brinell: coarse material, big impression

Brinell presses a hard ball, usually 10 mm tungsten carbide, into the surface and measures the diameter of the impression. The large impression averages over many grains, which suits castings, forgings and materials with a coarse structure. The result is written HBW, with the ball diameter, the load and the dwell time often implied by the standard.

The large impression is also the limitation. Brinell is not suitable for a thin sheet, a small feature or a finished surface that must stay undamaged. It is a bulk check for a bulk material, and it should be placed where the mark will not matter or will be removed later.

Choosing the method

Start with the thickness and the feature size. A thin strip needs a light load Vickers test or a superficial Rockwell scale. A heavy casting suits Brinell. A production line checking a hardened shaft wants Rockwell C because it is fast and the scale is familiar to the shop.

Then consider the question. If the question is whether the bulk heat treatment is correct, a Rockwell or Brinell check is enough. If the question is whether the case depth or the weld zone meets a specification, a Vickers traverse is the right tool. If the question is whether a small stamped contact has the right temper, a microhardness test with a stated load is the only sensible option.

Conversions and their limits

Conversion tables exist because industry needs a common language, and they are useful when a customer asks for an equivalent. They are not exact. The relationship between scales depends on the material, the heat treatment and the work hardening history. A conversion table built for carbon steel does not describe a stainless steel or a copper alloy.

The safe rule is to test in the scale the specification names. If the drawing says 40 HRC, test Rockwell C. If a supplier only has Vickers equipment, agree the conversion basis in writing before the first article, and record both the measured value and the converted one. Do not convert a result and then treat the converted number as if it had been measured to that scale.

Writing the callout

A complete callout names the method and scale, the load where the method allows a choice, the acceptance range, the location and the number of readings. It also states the condition of the surface, because a decarburised or ground surface reads differently from a mill finish. For a small part, add a note about the minimum section and the distance from edges.

For a heat treated part, separate the core hardness from the surface hardness and, where relevant, the case depth. A single number cannot describe a carburised case and a tough core. State the depth at which the limit applies, for example a minimum of 550 HV at 0.4 mm below the surface, and name the traverse method.

• Method and scale, such as HRC, HV10 or HBW 10/3000.
• Acceptance range with units, not a single target.
• Location: which face, how far from an edge, and whether the mark is allowed.
• Number of readings and the rule for averaging.
• Surface condition and any preparation allowed.
• Case depth definition where the part is surface hardened.

Records that support the number

The record should show the method, the load, the indent locations, the individual readings and the average. A certificate that reports a single value with no location is difficult to defend when a part fails. A photograph or sketch of the indents is cheap and settles most arguments.

Keep the test block certificate and the calibration date with the results. A hardness tester that has drifted will produce a consistent set of wrong numbers, and the only warning is the test block. For a dispute, a third party laboratory with a traceable block and a stated method is the practical route.