Microindentation Hardness Testing of Tool Steel

What microindentation testing is

In microindentation hardness testing a diamond indenter of a defined geometry is pressed into a prepared surface with a known force of 1 to 1000 gf, which is what ASTM E384 sets as the range. The same range is sometimes called microhardness, although the term describes small indents rather than low hardness values. European practice splits the field differently and calls 200 to 3000 gf the low load range, for a reason that matters here: forces below about 200 gf generally produce hardness numbers that differ from those produced at 200 gf and above.

The Vickers test uses a square based diamond pyramid with an angle of 136° between opposite faces. The angle was chosen in 1925 to bring the numbers as close as possible to Brinell hardness on the same specimen, which is why a Vickers value sits near the Brinell value for the same steel. Both diagonals of the indent are measured after the force is removed and the mean is used, with the hardness calculated from the surface area of the indent divided by the applied force. Unlike Rockwell, one scale covers every material.

The Knoop test uses a rhombohedral diamond whose long diagonal is about seven times the short one. Only the long diagonal is measured, and the hardness is calculated from the projected area of the indent. The narrow shape is the reason the Knoop indenter is used on case hardened parts and on thin sections, because on a steep hardness gradient the long diagonal can be laid perpendicular to the gradient and kept in material of one hardness, while a Vickers indent across the same gradient can end up with its two halves of different length and no valid measurement at all.

Both tests are computed on the assumption that the material does not recover when the force is removed and the indenter lifted. That assumption rarely holds, and although the Knoop test is sometimes said to eliminate recovery, it does not do so on metallic materials. The measured indent is therefore never exactly the indent that was formed.

Vickers and Knoop microindentation side by side, compiled from the article text.

Choosing the test force: keep the indent above 20 micrometres

The force is not a free choice, and this is the point that decides whether a microhardness figure is worth anything. ASTM E384 recommends that the operator keep the indents larger than 20 µm in diagonal, because the smaller the indent, the larger the penalty for every micrometre of measurement error. On a Vickers indent with a 10 µm mean diagonal, a measurement variation of ±0.5 µm is worth roughly a 10 per cent rise or fall in the calculated hardness. On a soft steel that may be tolerable; on a tool steel at 60 HRC a 10 per cent error is not.

Since a harder steel makes a smaller indent at the same force, the harder the specimen, the higher the force has to be to keep the indent above the limit. The practical consequence is that a test force which works in the core of a die may be far too low in its hardened case, and a traverse across a case is usually run at a single force chosen so that the smallest indent on the traverse, the one in the hardest material, still clears 20 µm. Where the case is shallow, the indents can be spread over several parallel traces at different depths so that the spacing does not distort the gradient.

An example makes the arithmetic concrete. On an induction hardened part expected to run from about 750 HV at the surface to 250 HV in the core, a 100 gf force gives an indent of under 16 µm at 750 HV, which is below the recommended limit and a poor choice for the hard end of the traverse. A 200 gf force gives about 22 µm at 750 HV and about 40 µm at 250 HV. A 300 gf force gives about 27 µm and about 47 µm, and can be the better choice still, provided the case is deep enough for a 300 gf indent to sit inside it.

Indent size in micrometres against test force, for a traverse across an induction hardened part. Figures read approximately from the source. ASTM E384 recommends keeping the diagonal above 20 µm.

Spacing follows the same logic. Indenting creates an elastic and plastic strain field around the indent, and a second indent placed too close is deformed on the side facing the first, which produces an erroneous number. Centre to centre spacing should be at least 2.5 times the diagonal for a Vickers indent, and at least twice the short diagonal for a Knoop indent. The centre of an indent should sit at least 2.5 diagonals from a specimen edge, although 1.8 diagonals has been shown to be acceptable.

Thin product has one more limit on top of the diagonal rule. For foil and wire in a quality control program the indent depth should be no more than 10 per cent of the thickness or the diameter. A hard foil of 0.002 in. (51 µm) can take a force up to about 800 gf, but the validity of the result depends on the hardness that is found: if a 500 gf test on that foil returns about 200 HK, the force was too high for the result to be trusted and the foil should be retested at no more than 300 gf.

Why the same tool steel can give two different microhardness numbers

It is widely stated that Vickers hardness is constant with test force in the macro range, and the literature that reports a change at lower forces has often put it down to the material. A controlled ASTM E4 interlaboratory round robin says otherwise, and the finding is worth knowing before a microhardness number is used to accept or reject steel.

In that program a single operator made indents in a set of ferrous and nonferrous specimens at forces from 25 to 1000 gf, five times at each force, and then two dozen people measured the very same indents. All four of the trends reported in the literature for hardness against test force appeared in the measurement data for identical indents. Some raters produced decreasing hardness with decreasing force, one produced increasing hardness, one produced no change, and others produced a rise followed by a fall. Laboratories that were outliers on other specimens did not agree with the rest. Nothing about the steel differed between the raters; the variation came only from how they measured the indents. The trend most often obtained, and the one most often reported in the literature, is hardness falling as the force falls, which corresponds to small indents being measured oversized at the magnifications in use.

The Knoop test behaves in the opposite direction for the same underlying reason. Contrast at the tips of the long diagonal is weaker than at the corners of a Vickers indent, so the long diagonal is more likely to be measured short, which returns a hardness number that is too high. Knoop hardness therefore rises as the test force falls, and the rise is larger on harder steel; results are statistically identical from 200 to 1000 gf.

Two conclusions follow for tool steel. Above about 100 gf, and in particular between 200 and 1000 gf, Vickers numbers are stable enough to compare between laboratories. Below 100 gf the number carries a measurement uncertainty that grows with the hardness of the steel, and a disagreement between a supplier and a customer at those loads is more likely to be a measurement disagreement than a metallurgical one. The load and the measuring procedure belong in the purchase order whenever a microhardness figure is being used as an acceptance criterion.

Converting Vickers and Knoop values to Rockwell C

Product specifications sometimes define a case depth in the Rockwell C scale, which is a bulk test scale and unsuited to the purpose, but the practice is common enough that the conversion has to be dealt with. Hardness conversions are developed empirically, and every one of them carries an error. The primary source is ASTM E140, which lists conversions as tables and also as equations, and most modern microindentation testers carry those tables so that an equivalent hardness can be printed next to each measurement.

Converting Vickers data to another scale is more straightforward than converting Knoop data, because Knoop hardness varies more with load. The ASTM E140 conversions between Vickers and other scales can be used at any test force above 100 gf. A Knoop conversion is a load specific instrument: one published for a 500 gf load is best at that load, reasonably accurate at slightly lower loads, and generally adequate at higher ones, because Knoop hardness becomes reasonably constant at 500 gf and above. The safe practice on a material you actually buy is to test it in both scales and see how well the chart agrees with your own specimens before the conversion is written into a specification. Our page on Vickers to Rockwell C conversion for tool steel sets out the hardened range, and the wider calculator on tool steel hardness conversion covers the HRC and HRB bands.

Specimen preparation decides the result

Preparation for microindentation is not a minor matter, and it becomes more critical as the force decreases. At 300 to 1000 gf a perfect finish is not required, but that does not excuse leaving sectioning and grinding damage in place; a normal preparation can be stopped after grinding and polishing to a 6, 3 or 1 µm diamond finish. At lower forces the specimen should be prepared right through to a damage free condition, following ASTM E3. Residual damage from cutting and grinding influences the result, and depending on its nature it can push the apparent hardness either up or down relative to the true value.

Testing near an edge is where preparation and technique meet, and it is exactly what a tool steel page needs, because it is how the hardness of a coating, a carburized or nitrided case, an induction hardened surface, or a decarburized skin is measured. Edge preservation means flatness out to the edge. A good thermosetting epoxy, mounted and cooled back to ambient under pressure, automated preparation equipment and napless cloths will keep an edge well enough without plating the specimen, and unmounted bulk specimens can also be prepared with adequate edge retention on automated equipment.

Where this is used on tooling

The classic application of microindentation testing is the assessment of a change in surface hardness. Increases from carburizing, nitriding or localized surface hardening are measured, and so are decreases from decarburization or from localized heating. A hardness traverse turns what the eye can see on a polished cross section into a number: the magnitude of the change and the depth at which it happens.

Case depth on a tool is normally specified as the depth to a stated hardness, and 550 HV is the usual threshold for a carburized or nitrided part. The traverse that produces it is a line of indents from the surface inward, taken at a single force and a spacing that keeps each indent clear of the next, until the reading reaches the core hardness of the steel. Decarburization is measured the same way from the other direction: a soft skin of measurable depth under a case that may be perfectly acceptable further in. For a die that heat checks or a cutting edge that dulls early, a traverse across the failed surface is what separates a steel or heat treatment problem from a service problem, because it puts a number on the soft layer instead of leaving it as an impression.

What this means on a tool steel order

Delivery hardness on tool steel is a bulk value in HRC or HB, taken on a prepared surface, and it describes the steel as supplied. A Vickers or Knoop figure describes a smaller volume, and it is only comparable with another microindentation figure taken at the same force with the same measurement procedure. Where a microhardness value is an acceptance criterion, for a nitrided die or a hardened blade, the order should state the method and the force, and where a case depth is specified, the depth to a stated hardness rather than a single hardness at the surface.

Source: ASM Handbook, Volume 8, Mechanical Testing and Evaluation, ASM International, 2000. Article Microindentation Hardness Testing by George F. Vander Voort, Buehler Ltd. (PDF p469-494).