Portable Hardness Testing of Tool Steel

A press tool, a forging die or a mill roll cannot be carried to a bench hardness tester, so the hardness reading on a large tool is usually taken with a portable instrument. The three methods in common use do not all measure the same thing, and that difference decides what a number taken in the press shop can and cannot be used for. This page covers the Scleroscope, the Leeb or Equotip tester and the ultrasonic tester, the limits of each, and a documented case where an on-site reading found a problem that a bench Rockwell test had passed.

Rebound tests and indentation tests measure different properties

Hardness can be related empirically either to the elastic response of a material or to its plastic deformation, and the two families of test are built on those two things. The bench tests, Rockwell, Brinell, Vickers and Knoop, determine hardness in terms of plastic behaviour by measuring the size or the depth of an indentation. The portable rebound tests relate hardness to the elastic response of the material. That is the first thing to hold on to when a number taken with a handheld instrument is compared against a hardness figure on a mill certificate or a drawing, because the two numbers were not obtained in the same way.

Three methods cover almost all on-site work on tooling. Two of them are rebound tests, the Scleroscope and the Leeb tester, and the third, the ultrasonic tester, is an indentation test made portable.

Scleroscope hardness testing

The Scleroscope is a diamond tipped hammer dropped from a fixed height onto the surface of the part. The height of rebound is the measure of hardness. It was invented by Albert F. Shore in 1907 and was the first commercially available metallurgical hardness tester made in the United States. It is no longer manufactured in the United States, but it is still used for testing very large specimens such as forged steel or wrought alloy steel rolls, which is exactly the class of part a tool steel supplier and a die shop deal with.

The scale is defined on tool steel, which is worth knowing when a Scleroscope number is quoted back at you. One hundred units is the average rebound of the hammer from a water-hardened tool steel quenched to maximum hardness and left untempered. The scale continues above 100 so that materials harder than fully hardened tool steel can be measured, and the model C instrument is graduated from 0 to 140. As a rough guide, hard steel tests around 100, medium-hard steel around 50, and soft metals between 10 and 15.

Two models were made. The model C reads the rebound directly from a vertical scale through a glass tube, and can be used unmounted on workpieces of unlimited size, provided the part weighs at least 2.3 kg (5 lb) so that it absorbs the inertia of the hammer. The model D arrests the hammer at the top of its rebound and reads it on a dial or a digital display; because its vertical alignment is critical, it should not be used unmounted. Both are available in a standard calibration that conforms to ASTM E448 and correlates directly with Rockwell C, Brinell and Vickers values, and in a roll calibration to ASTM A427 that is used for homogeneous wrought hardened alloy steel rolls for flat rolled products, symbolised HFRSc or HFRSd.

The method’s advantages on large tooling are that it is fast, over 1000 tests an hour are possible, it needs no highly skilled operator, it leaves no obvious crater on a finished surface, and one scale covers the whole range from the softest to the hardest metal. Its limitations are just as practical. The instrument has to be kept vertical so the hammer falls freely. It is more sensitive to surface condition than other hardness tests, so an excessively coarse surface gives erratic readings and the surface may have to be filed, machined, ground or polished, without overheating or cold working it. The mass or the configuration of the part affects the reading, and thin or light parts have to be clamped solidly; a hollow ringing sound instead of a dull thud on impact means the part is not properly supported. Five determinations are made and averaged. Indentations should be spaced at least 0.50 mm (0.020 in.) apart, and a flat part with parallel faces can be tested within 6 mm (0.25 in.) of the edge if it is properly clamped.

The minimum sheet thicknesses the method will take are set out below.

Minimum sheet thickness that can be hardness tested with a Scleroscope. The sheet should be flat and free of undulation; a bowed sheet is placed concave side up.

Readings are indicated by the maximum rebound of the hammer on the first bounce, which is why even the most experienced operators may disagree by one or two points on the same part.

Leeb (Equotip) hardness testing

A Leeb tester is a portable instrument that works on the same rebound principle. An impact body is propelled into the sample by a spring, forms a small indent, and rebounds away from the surface according to the hardness and the elasticity of the material. An electronic induction coil measures the velocity of the impact body before and after contact, and the Leeb hardness number is calculated from the ratio of the two velocities. The number is written HL followed by one or more suffix letters that identify the impact device. The method is also known as Equotip hardness, after the trademark, and is standardised as ASTM A956.

All Leeb testers are electronic and display the result digitally, and the choice of impact device is the first decision on a job. The devices differ in maximum hardness and in the minimum sample mass they need, because the test rests on a rebound and the mass of the sample is therefore critical to the result. A part that is too light can still be tested by coupling it carefully to a larger mass. The table below is the selection guide.

Leeb impact devices and their applications.

Because the Leeb number is not universally recognised, most instruments carry a built in conversion to Rockwell, Brinell, Vickers and Scleroscope. The problem with those conversions is that they are strongly material dependent, usually through the elastic modulus of the material, so a tester holds separate conversion data for several classes of material: steels, cast iron, aluminium and brass among them. Converting a Leeb reading on tool steel with the wrong material class loaded is a common way to produce a number that will not match a bench test.

Three operating points decide whether a Leeb reading means anything. The first is angle: testers can work at various angles but need a correction for gravity acting on the impact body when the device is not vertical, and the correction is frequently built into the instrument or supplied as a table to subtract. The second is surface finish, where a rougher surface normally gives a softer result. The third is curvature: concave or convex surfaces can be tested if the radius of curvature is greater than 30 mm (1.18 in.) for every device except the G type, which needs 50 mm (2 in.). Adapters are available for round and awkward surfaces. Calibration can only be verified against standardised test blocks, ten readings averaged on the block surface, with the unit considered calibrated if the average is within 13 numbers of the certificate value.

Ultrasonic hardness testing

The ultrasonic method is not a rebound test. A Vickers diamond is attached to one end of a magnetostrictive metal rod that is excited to its natural frequency by a piezoelectric converter. When the free end is brought into contact with a solid, the resonant frequency of the rod changes, and once the instrument is calibrated for the known elastic modulus of the material being tested, the area of contact between the diamond tip and the surface can be derived from the frequency shift. At constant force, the contact area is inversely proportional to hardness, so the measured frequency converts into a hardness number, displayed directly in Vickers or Rockwell C.

The maximum indentation load is around 800 gf, which puts the indentation depth between 4 and 18 µm. In the vast majority of cases the workpiece surface is unharmed and the test can be treated as nondestructive. A reading takes under 15 seconds, up to 1200 parts an hour can be tested, and the instrument is small enough to run on batteries, permits measurements in any orientation through a motor driven probe, and needs only a one point calibration. It is the most portable of the microhardness testers and, unlike the Scleroscope, it does not care how heavy the part is.

There is one technical advantage over conventional microhardness testing that is worth understanding. The ultrasonic instrument measures the contact area while the load is still applied, whereas a conventional Vickers or Knoop test measures the indent after the load has been removed, and elastic recovery on unloading can bias that measurement. Ultrasonic testing also uses a single loading force rather than a minor and a major load, so it does not need the sensitive displacement measuring system or the rigid frame that Rockwell testing does, and in many cases it can be used without clamping or rigidly supporting the part.

The limits come from the same design. There is no optical system, which is what makes the probe small and fast, but it also means the indenter cannot be precisely positioned on a microconstituent, so hardness of individual carbides is out of reach for this instrument. It also makes a hardness traverse across a case hardened layer awkward, because positioning the indenter 0.05 or 0.10 mm (0.002 or 0.004 in.) from an edge is difficult. The workaround is to reverse the traverse: take the first reading well inside the part, beyond the depth of the case, and work outward at prescribed intervals until a soft reading shows the indenter has reached the mounting material. Finally, ultrasonic testing is microhardness testing, so the surface finish requirements are stringent and become more so as the load falls; a metallographic finish prepared to ASTM E3 is required, and grinding or polishing has to avoid heating and distorting the surface.

A die that passed the bench test and still heat checked

The practical value of an on-site hardness method shows up in a documented case in the source. A die casting plant was having trouble with heat checking on dies made from H13 tool steel, quenched and tempered. On-site hardness tests with an ultrasonic instrument proved that the superficial surface was quite soft as a result of decarburization, even though the actual Rockwell C readings on the same dies were acceptable. The decarburized layer at the surface was the cause of the heat checking, and the heat treating procedure was corrected.

The order of the two readings is the point. A bench Rockwell test on a die is normally taken on a prepared surface, and the decarburized skin can be removed or left outside the measurement volume; the ultrasonic reading reaches only the outermost few micrometres, which is precisely where decarburization sits. For a tool that fails at the surface, in heat checking, in cracking from a working edge or in wear, a bulk hardness that meets the drawing is not evidence that the surface the tool works with is sound.

Comparing the three methods

The table below sets the three methods side by side on the points that decide which one to use on a tool or a die. Note that the rebound methods relate hardness to elastic response while the ultrasonic method relates it to plastic behaviour, which is why a Scleroscope or Leeb number should not be presented as a Rockwell C value without saying how it was obtained.

The three portable hardness methods used on tooling.

What this means on a tool steel order

Delivery hardness on tool steel is a bench value, taken on a prepared surface of a test piece or of the bar itself, and it is quoted in HRC or HB. A reading taken on a finished die with a handheld instrument is a different measurement with a different uncertainty, and the two should not be compared directly without saying which method produced which number. When a portable reading comes back low on a large tool, the sequence that settles it is the one from the case above: check whether the low reading is a surface effect from decarburization or grinding, check the instrument against its reference block, and only then treat it as a hardness problem in the steel. On round bar the geometry of the part changes the bench reading too, which is covered on our page on hardness test correction factors for round bar, and the smallest section that can carry a bench test on minimum thickness for hardness testing.

Before quoting a portable hardness reading

A reference table only, it is not an Aobo Steel specification. A rebound reading and a bench Rockwell reading are not the same measurement, and a portable reading carries the surface condition, the mass of the part and the operator into the number. Confirm the method, the instrument and its calibration before a value goes into a drawing, an order or an acceptance decision.

Source: ASM Handbook, Volume 8, Mechanical Testing and Evaluation, ASM International, 2000. Scleroscope, Leeb and ultrasonic sections from Miscellaneous Hardness Tests by Edward L. Tobolski, Wilson Instrument Division, Instron Corporation (PDF p547-563).