The 1 mm Thermocouple and the Erosion of the Glow-Wire Test
Oct 23, 2026 · @Magnus, Hermes
Admitting a 1.0 mm thermocouple into a 4.0 mm glow-wire removes about a third of the wire's cross-section exactly where the temperature is measured. Two apparatus that both comply can then deliver different heat to the same specimen at the same indicated temperature. A test whose value lies in its severity should not have been relaxed this way.
A precision test, not a hot nail
The glow-wire test is valuable because it is reproducible, and it is reproducible only because every parameter of the heat source is fixed. IEC 60695-2-10 does not ask for "something hot": it defines a specific object and a specific way of applying it.
The heat source. A loop of Ni80/Cr20 wire, 4.0 mm in diameter before bending, with a defined tip geometry.
The contact. A force of 1.0 N, penetration limited to 7 mm, an approach that must not cool the tip, and an application time of 30 s.
The temperature. Set points from 550 to 960 °C, held within a few kelvin, verified against the melting point of silver.
The observation. A draught-free enclosure, a dark background of 20 lx or less, and tissue paper on a pine board 200 mm below to catch burning drops.
The dark background is not a detail. Short flames and glowing at the threshold of the pass criterion are invisible in a bright room, and a draught changes both the combustion and the heat exchange at the tip. A material that sits close to its limit passes or fails on exactly these conditions. That severity is the reason the test is trusted by product standards, certification bodies and insurers.
What changed, and when
Since the second edition of IEC 60695-2-10, published in April 2013, clause 4.3 states that the thermocouple "shall have an overall nominal diameter of 1,0 mm". The 1.0 mm thermocouple is not an alternative: it is the requirement. The foreword of that edition lists the change among the main changes, as an update of "the temperature measuring system (see 4.3)", together with new tolerances for the glow-wire dimensions. The wire itself remained a 4.0 mm loop.
The third edition (2021) kept the 1.0 mm requirement unchanged. The fourth edition, published on 8 May 2026, keeps it again and goes one step further: a new normative Annex D allows an optical pyrometer as an alternative to the thermocouple, with the thermocouple prevailing in case of dispute.
The benefits put forward for the thicker thermocouple are practical: it is sturdier, cheaper, easier to source and breaks less often. None of them concerns the specimen. They concern the cost of running the apparatus. Much of the industry still builds and calibrates with 0.5 mm thermocouples, which shows how little the laboratories that know this test were convinced.
To confirm before publication: the wording of the first edition (IEC 60695-2-10:2000) on the 0.5 mm thermocouple, and the justification recorded in the TC 89 committee documents for the 2013 change.
What the hole takes from the wire
At the section crossed by the pocket hole of a 1.0 mm thermocouple, a 4.0 mm wire keeps about two thirds of its metal and behaves like a 3.3 mm wire. With the 0.5 mm thermocouple it keeps about 83 %.
The pocket hole is drilled behind the tip and must be a close fit. In practice, for a 0.5 mm thermocouple, it is 0.53 to 0.54 mm: looser and the thermocouple loses contact, tighter and the movement of the wire as it heats and cools shears the thermocouple.
Thermocouple | Pocket hole | Metal left at the hole | Equivalent wire | Local heat per unit length vs. intact wire |
none | — | 12.6 mm² (100 %) | Ø 4.0 mm | 1.00 × |
Ø 0.5 mm | Ø 0.53–0.54 mm | ≈ 10.4 mm² (83 %) | ≈ Ø 3.6 mm | ≈ 1.21 × |
Ø 1.0 mm | Ø ≈ 1.04 mm | ≈ 8.5 mm² (67 %) | ≈ Ø 3.3 mm | ≈ 1.48 × |
The last column follows from Joule heating: at the same current, the power dissipated per unit length rises as the cross-section falls. Moving from the 0.5 mm to the 1.0 mm thermocouple raises the local dissipation at the hole by about 22 %, at the very point the standard uses as its reference.
The figures assume the hole crosses the full wire diameter, which is close to the real pocket hole; they are a first-order estimate, to be confirmed by the measurements proposed below.
The heat moves away from the tip
Removing nickel-chromium from the tip does not only make the wire weaker: it changes where the heat is. The specimen does not feel a temperature, it feels the energy the tip can deliver to it.
Less thermal mass at the contact. When a cold or wet specimen touches the tip, the first seconds are fed by the heat stored in the metal around the contact point. Take a third of that metal away and the tip has less reserve to give.
A redistributed hot zone. The thinner section heats more per unit length, so the hottest point and the heat reservoir no longer coincide with the geometry defined by the standard. The thermal profile of the loop shifts along the legs.
Different energy at the same reading. Two apparatus, one with a 0.5 mm and one with a 1.0 mm thermocouple, can both read 850 °C and still deliver different energy to the specimen. The temperature is the same on the display; the test is not the same.
This is the core of the objection. The standard fixes the temperature because, for a fixed object, temperature stands for the heat delivered. Change the object and that equivalence is lost.
The thermometer becomes part of the problem
A 1.0 mm sheathed thermocouple has four times the cross-section of a 0.5 mm one. It is no longer a small probe sitting in the wire; it is a significant body of metal and insulation inside the tip.
It drains heat. The sheath conducts heat out of the tip along its length, like a fin. The junction reads a point that is being cooled by the measuring instrument itself.
It responds later. Four times the cross-section means roughly four times the thermal mass at the junction. The first seconds after contact, when a wet specimen cools the tip or a burning specimen drives it above the set point, are exactly the seconds a slower thermocouple smooths out.
It reads a different object. The standard verifies the reading against the melting of silver at 960 °C in still air. That check says nothing about how the tip behaves under load, against a specimen, when the reserve of heat around the junction has been cut by a third.
The glow-wire is a test of transient behaviour. A measuring system that is slower and that perturbs the point it measures hides the transient that the test exists to show.
A weaker tip that changes shape sooner
Every test cycle heats the loop to as much as 960 °C and lets it cool again. The tip expands, contracts and works like a spring, and the stress of that movement concentrates where the section is smallest: at the edge of the pocket hole.
With a third of the section gone, the tip creeps, oxidises through and deforms sooner. The geometry of Figure 1, on which the whole test rests, is lost earlier in the life of the element. The thermocouple is protected; the reference it is meant to measure is not.
"It breaks less often" is not a technical argument
The case for the 1.0 mm thermocouple is that the 0.5 mm one is fragile and expensive to replace. That is true, and it is an argument about the cost of the laboratory, not about the validity of the result.
A 0.5 mm thermocouple survives when the pocket hole is right: a close fit, 0.53 to 0.54 mm, that holds the junction against the metal without clamping it. Most failures come from holes that are too loose or too tight, from careless handling, and from heating cycles run beyond the test. These are problems of workmanship and maintenance. The standard should fix them with dimensions and tolerances, not by changing the heat source.
A standard that relaxes the reference to spare a consumable has its priorities reversed. The specimen and the end user of the product are the reason the test exists.
What we ask of IEC TC 89
Restore the 0.5 mm thermocouple as the reference in clause 4.3, with the pocket hole specified as a close fit (0.53 to 0.54 mm), and keep the 1.0 mm thermocouple only where equivalence has been demonstrated.
Require proof of equivalence before any alternative measuring system is admitted, the 1.0 mm thermocouple and the new pyrometer of Annex D alike: same delivered heat at the tip, same response at contact, same results on reference materials.
Run an inter-laboratory comparison on reference materials with both thermocouple sizes, reporting GWFI and GWIT results side by side.
Make the measuring system part of the test report: thermocouple diameter, pocket-hole diameter and element service time, so that results obtained with different configurations can be told apart.
Publish the justification for the 2013 change and the data, if any, on which it was based.
The measurements Ormity will publish
Ormity will run the comparison on its new DNGW695 glow-wire apparatus, with two elements identical except for the thermocouple: one with a 0.5 mm thermocouple in a 0.53–0.54 mm pocket hole, one with a 1.0 mm thermocouple as required since 2013. Both run on the same transformer, carriage and chamber.
Measurement | What it shows |
Current needed to hold 550, 750 and 960 °C | How much the section loss changes the heating of the tip |
Silver-foil check at 960 °C | Whether both read correctly in still air |
Thermal image of the loop at set point | Where the hottest zone and the heat reserve sit |
Temperature drop and recovery at contact | Thermal mass at the tip and thermocouple response time |
Tip deformation after a fixed number of cycles | Mechanical life of the reference geometry |
GWFI and GWIT on a reference material (e.g. PA66, 1.5 mm) | Whether the two configurations rate the same material the same way |
If the two configurations rate the same material differently, the 1.0 mm requirement is not a convenience. It is a reproducibility problem in an international standard, and it has to be corrected.
Sources
Magnus, Hermes, Technical Director, Ormity Inc. — designer of compliance test equipment with more than 25 years of experience in flammability and other test apparatus to IEC, UL, ISO, NEMA and ASTM standards.
Comments