How the Chemistry of a Tool Steel Heat Is Tested

Four methods, not one

The elements that define a tool steel grade are split across techniques. The alloying elements and the common impurities are reported by spark emission spectroscopy, which is the workhorse of a steelworks laboratory. Carbon and sulphur are measured by high-temperature combustion, which is the reference method for those two elements because emission sources struggle with them. Oxygen, nitrogen and hydrogen are measured by inert gas fusion, and no emission method reaches them at all. X-ray fluorescence sits alongside as an independent technique that needs no standards for a semiquantitative answer.

The four methods that between them cover the chemistry of a steel heat. Compiled from the three source articles in ASM Handbook, Volume 10.

The consequence for a certificate is that its rows came from different methods on different sample forms, and that is normal. Where it matters is when a customer has one of the rows rechecked. Carbon, for example, can be reported by emission, by combustion or by a dedicated glow discharge set-up, and those three routes do not have the same uncertainty.

How the heat is sampled and analysed

Spark emission is the fastest route to an elemental analysis of a metal, and the steel industry’s use of it explains why. During production of a heat, the metal is sampled and the molten sample is allowed to cool. A flat surface is ground or sanded onto the cooled sample, which then goes into the spark source with no further preparation. The analysis runs, and the result goes straight back to the furnace so the alloying additions can be trimmed into range. Sampling and analysis together take a few minutes at most.

That speed is the reason the method is trusted at the furnace, and it is also the reason its limits matter. The sample forms one electrode of the spark and a tungsten pin forms the other, so the material has to conduct. The spark is normally run in a flowing argon sheath, which makes the discharge strike much more reproducibly and reduces the burn area by a factor of ten compared with an unstabilised spark. The light is not recorded immediately. A preburn period of about a minute conditions a fresh electrode surface, after which the emission for most elements stays reasonably constant for the 30 seconds needed to record a spectrum.

The reason for that preburn is worth knowing, because it is the mechanism behind a discrepancy a customer will occasionally see. Emission from a spark train changes within minutes as the spark conditions the surface, a behaviour called sparking-off. Its exact shape depends on the spark parameters, the sample composition, the sample phase structure, the surface condition, the sparking atmosphere and the burn area. Because the composition and phase structure of the electrode are among the inputs, the emission result for an element is matrix dependent, and the correction used in practice is to ratio the line intensities of the minor elements against a line from iron, the major constituent. That compensates for variations in sampling and excitation between samples, and it carries an assumption with it. The assumption is that the reference element is sampled and excited the same way as the minor constituents, which is not always safe in a steel carrying inclusions whose composition differs sharply from the bulk.

Calibration is where an emission result is won or lost

None of that produces a usable number unless the instrument has been calibrated against standards that match the unknown closely, in both chemical composition and physical form. A laboratory running spark analysis has to hold a set of standards for every type of material it analyses, and the article is blunt that spark standards are not easily produced and generally have to be bought, from the national reference material supplier or from a private company. A tool steel heat checked against a low alloy steel standard is being measured on a curve that does not describe it.

Emission spectroscopy also has a hard limit on which elements it can see. Nitrogen, oxygen, hydrogen, the halogens and the noble gases are listed as difficult or impossible to determine by optical emission. Every emission method also carries the same general caveat, which is that its response depends on the matrix. That is why the certificate row for a tool steel’s nitrogen or oxygen content cannot have come from the same instrument as the chromium row.

Carbon and sulphur by combustion

Combustion analysis reverses the logic of the other methods. Instead of exciting the sample, it burns it. A weighed sample of 1 g or less goes into a ceramic crucible with an accelerator and is brought to between 1370 and 1425 °C in a stream of oxygen. The carbon leaves as carbon monoxide and carbon dioxide, the sulphur as sulphur dioxide, and both are measured by infrared absorption or thermal conductivity and reported as a percentage of specimen weight. Analysis takes 40 seconds to two minutes once the sample is prepared.

The accelerators are not optional, and their selection rules are specific enough to be a useful check on whether a laboratory knows what it is doing.

Combustion accelerators and what each contributes. Source, Table 1 of High-Temperature Combustion, in ASM Handbook, Volume 10.

Two further details decide whether a low figure is real. The blank, defined as carbon or sulphur arriving from anywhere other than the sample, can come from the oxygen supply, the crucible or boat, and the accelerators, and it runs between 0.010 and 0.0005 % depending on the system and the grade of consumables. Automatic analysers compensate for it, and manual or semiautomatic ones require it to be subtracted. And the specimen has to be homogeneous, because the method is destructive and reports a total for the piece that was burned.

Oxygen, nitrogen and hydrogen by inert gas fusion

Inert gas fusion is the method for the three gases. A sample of 2 g or less is fused in a graphite crucible, heated to between 2500 and 3000 °C by an impulse current or by high-frequency induction. At that temperature the bonds between the gas and the metal break down, and the crucible itself supplies the carbon that takes up the oxygen. Hydrogen and nitrogen leave as H₂ and N₂, oxygen leaves as carbon monoxide, and an inert carrier gas sweeps them to a thermal conductivity or infrared detector. Helium is the usual carrier for nitrogen or nitrogen and oxygen, and nitrogen or argon for oxygen or hydrogen on their own.

The three gases measured by inert gas fusion, why they get into a tool steel and what they do there. Compiled from Inert Gas Fusion, ASM Handbook, Volume 10.

Sample handling for this test has a rule that is easy to break. Material has to be cut to size, and the cut has to be kept cool for oxygen and hydrogen work specifically, because heating during sectioning drives hydrogen out of the sample and oxidises the surface, and either one moves the number. The specimen must also not be contaminated with nitrides, oxides or hydrides on the way in. For a tool steel the hydrogen figure is the one with a service consequence, since the internal cracking that hydrogen causes appears during cooling after drawing, rolling or forging, which is to say during the processes that made the bar.

X-ray fluorescence as the independent check

X-ray fluorescence works on a different physical principle from emission, which makes it useful as a second opinion. It measures the characteristic x-rays a sample emits when it is irradiated, it works on bulk solids, powders, pressed pellets, glasses, fused disks and liquids, results are semiquantitative on many samples with no standards at all, and its detection limits for bulk work run from a few parts per million to a few tens of parts per million depending on the x-ray energy and the matrix.

Its limits are just as well documented. It is not suitable for elements below sodium in atomic number unless special equipment is fitted, and then only down to carbon at best, so it cannot be the method behind a carbon figure on a tool steel certificate. Sample preparation is described as the single most important step in the analysis and the one most often neglected. Surface roughness is usually allowed to reach 100 μm for higher energy radiation but has to come down to 20 to 40 μm below about 2 keV, and soft metals smear across the surface during polishing, which coats the sample with the element being measured and returns a high result. A laboratory that polishes a leaded or free machining steel the same way it polishes a tool steel is measuring its own preparation.

A glow discharge source is the third option for the same job, and it does better than the spark on two counts that matter for routine control. Its working curves are more linear and less matrix dependent, and it produces cleaner spectra with narrower lines on a source that is simpler electronically. Against that it needs a clean sample ground flat enough to hold a vacuum at the O ring, while the spark runs at atmospheric pressure on a sample that has only been coarse sanded.

What this means for a tool steel order

The chemistry on a certificate is a set of results, and the elements that decide a grade are not all measured the same way. Carbon, the element that separates one tool steel from another more than any other, is a combustion measurement or a dedicated glow discharge measurement, not a routine spark reading. Chromium, molybdenum, vanadium and tungsten come from emission. Oxygen, nitrogen and hydrogen, which govern cleanliness and the risk of internal cracking, come from inert gas fusion and will be reported separately or not at all.

So when a composition has to be checked, the two questions worth asking are which method produced each row and which standard was used to calibrate it. The composition ranges for the common tool steel families are tabulated in the tool steel composition chart, with the individual grade sheets at D2 steel composition and H13 steel composition, and the melting route that decides how low the gas and inclusion levels can go is covered in ESR tool steel. The powder metallurgy grades, where a carbon figure is central to the specification, are collected in PM high speed steel composition. How the numbers are handled on the way out of our own warehouse is set out in how we control tool steel quality, the grades with their composition ranges sit under cold work and high speed tool steels, the catalogue is at tool steels, and a certificate question can be put to us directly through the contact page.

Source: ASM Handbook, Volume 10, Materials Characterization, ASM International, 1986.