Testing standard
Metallic materials — Tensile testing at low temperature
Written and technically reviewed by Dak System Inc. engineeringLast reviewed
ISO 15579 was the international method for tensile testing of metallic materials at sub-ambient temperature, and it is withdrawn — ISO 6892-3:2015 cancels and replaces it, and is what should be specified for new work. The method itself is the room-temperature tension test carried out on a test piece cooled and held cold, over a range reaching down to −196 °C, so that the strength and, more importantly, the ductility a metal retains in the cold are measured rather than inferred.
The tension test ISO 6892-1 runs at room temperature, carried out on a test piece cooled and held at a specified sub-ambient temperature. The successor sets the range at +10 °C down to −196 °C, the boiling point of nitrogen at atmospheric pressure. The test piece is cooled, soaked until its temperature has stabilised through the section, then pulled to fracture while that temperature is held. Everything geometric — original cross-sectional area, original gauge length, elongation after fracture, reduction of area — is measured at room temperature, so reported stresses sit on a room-temperature section rather than the contracted cold one.
How much of a metal's strength and, more importantly, its ductility survives the cold. Most steels and many alloys gain strength as temperature falls and lose toughness at the same time, and the transition can be narrow. The materials that need this test are the ones that meet the cold in service: liquefied gas tanks and their pipework, cryogenic plant, offshore and Arctic structures, and aerospace components. A yield strength with no temperature beside it answers none of those questions.
The successor notes that several properties reported at room temperature are not usually determined cold unless a specification calls for them — permanent set strength, percentage yield point extension, the extensions at maximum force and total extension at fracture among them.
The test piece is the one ISO 6892-1 uses. Everything that makes this method difficult is the temperature — holding it on the metal, proving it, and deciding what the gauge length means once the specimen has contracted.
Rp0,2 = stress at 0,2 % plastic extension, at the test temperature
The equation is unchanged from the room-temperature method. What changes is that every quantity in it belongs to the temperature it was measured at, and a proof strength quoted without that temperature answers nothing.
Rm = F_m / S₀
S₀ is a room-temperature dimension by convention. The method makes no correction for thermal contraction, so the stress reported is force divided by the warm area.
Where the test is run by immersion in liquid nitrogen, no temperature measurement is required and the report says so. That is the one case in the method where the absence of a temperature record is correct rather than a defect.
The liquid-nitrogen chamber is the sub-ambient enclosure in our range, working down to −150 °C. That covers the greater part of this method's span without reaching its −196 °C floor, which is normally met by immersing the test piece in boiling nitrogen rather than by cooling a cabinet to it.
Specifications
Grips outside the cold zone, loading the test piece through the enclosure, so the clamping is done on material at room temperature rather than on metal that has gone brittle.
Specifications
Strain has to be read on the test piece itself. The part of the instrument that projects beyond the cooling device is what the method asks to be shielded from draughts, because a room-temperature fluctuation there reads as extension.
SpecificationsAn ordinary tension frame with force verified to ISO 7500-1 class 1 or better, an extensometer to ISO 9513 class 1 or better where proof strength is determined and class 2 for higher extensions, and a cooling arrangement that holds the test piece rather than the air around it. ISO 6892-3 permits a refrigeration unit, expansion of a compressed gas such as carbon dioxide or nitrogen, or immersion in a liquid at its boiling point or in a refrigerated liquid such as alcohol, and notes that the medium affects both cooling time and heat transfer during the test.
The tolerances are on the specimen. The indicated temperature must be within ±3 °C of the specified temperature and the gradient along the surface must not exceed 3 °C. Below a 50 mm gauge length, one sensor measures at each end of the parallel length; at 50 mm and above, a third sits near the centre. Temperature equipment resolves to 1 °C or better and is accurate to ±2 °C from +10 °C to −40 °C and ±3 °C from −41 °C to −196 °C, verified at intervals not exceeding a year. Soaking is at least 10 minutes, longer for heavier sections, and loading only begins once the extensometer output has settled. Testing in liquid nitrogen is the exception: no temperature measurement is required, and the report records that none was taken.
Citing the withdrawn designation on a new order, the most common problem this standard now has. Measuring the medium instead of the metal. Loading before the interior of a heavy section has reached temperature, so the result belongs to no single temperature. Letting the test piece overshoot below the specified temperature while cooling, which the successor permits only by agreement. Extensometry that drifts because the part of the instrument outside the cooling device sits in a draught. And quoting a result without the temperature, the soaking time and how the gauge length was established.
| ISO 15579:2000 | ISO 6892-3:2015 | |
|---|---|---|
| Status | Withdrawn, stage 95.99 | Published, confirmed 2020 |
| Edition | First and only, June 2000 | First, April 2015 |
| Committee | ISO/TC 164/SC 1 | ISO/TC 164/SC 1 |
| Scope | Tensile testing at low temperature | +10 °C to −196 °C |
| Specify for new work | No | Yes |
A certificate citing ISO 15579 is not wrong about the physics — the method is substantially the same test — but the designation has been withdrawn since ISO 6892-3:2015 was published, and a purchase specification still calling for it should be updated rather than honoured indefinitely.
| Part 1 | Part 2 | Part 3 | |
|---|---|---|---|
| Temperature | Ambient, 10–35 °C | Elevated | +10 °C to −196 °C |
| Enclosure | None | Furnace | Cooling device or bath |
| Dominant difficulty | Rate control through yield | Proving uniformity when hot | Proving uniformity when cold |
No. ISO 15579:2000 was the first and only edition and is withdrawn, recorded on the ISO catalogue at stage 95.99. ISO 6892-3:2015, Metallic materials — Tensile testing — Part 3: Method of test at low temperature, cancels and replaces it and is what should be specified for new work. This page is kept because certificates and purchase specifications still cite the old designation.
It is the international method for tensile testing of metallic materials at temperatures between +10 °C and −196 °C. The test piece is cooled, soaked until its temperature has stabilised through the section, and pulled to fracture while that temperature is held. It reports the same properties as the room-temperature test, measured where the component will actually work.
Because most steels and many alloys gain strength as the temperature falls and lose toughness at the same time, and the transition can be narrow. Liquefied gas tanks and their pipework, cryogenic plant, offshore and Arctic structures and aerospace components all meet the cold in service. A yield strength with no temperature beside it answers none of those questions.
At room temperature, before cooling, and the method makes no correction for thermal contraction. That is a convention rather than an oversight — it keeps the original area and gauge length measurable and repeatable — and because every laboratory ignores the contraction in the same way, the results stay comparable.
The indicated temperature must be within ±3 °C of the specified value, and the gradient along the surface of the test piece must not exceed 3 °C. Below a 50 mm gauge length one sensor measures at each end of the parallel length; at 50 mm and above a third sits near the centre.
At least 10 minutes, and longer for heavier sections — ISO 6892-3 notes that more time can be needed to bring the whole cross-section to temperature. Loading only begins once the extensometer output has settled. Pulling before the interior has cooled produces a result that belongs to no single temperature.
No. ISO 6892-3 states that where testing is carried out in liquid nitrogen no temperature measurement is needed, because the boiling medium fixes the temperature. The report then records that no measurement was taken. It is the one case where an absent temperature record is correct.
Because the test piece contracts as it cools, and the gauge length can be established in four different ways — at room temperature, nominally at test temperature, extended at room temperature so the nominal value is reached when cold, or corrected afterwards for the contraction. Each gives a slightly different percentage extension, so ISO 6892-3 requires the method used to be documented in the report.
An ordinary tension frame with its force-measuring system verified to ISO 7500-1 class 1 or better, an extensometer to ISO 9513 class 1 or better where proof strength is determined and class 2 for the higher extensions, and a cooling arrangement that holds the test piece rather than the air around it. The temperature-measuring system resolves to 1 °C or better and is verified at intervals not exceeding one year.
Dak verifies against whichever standard the method names, and where a class applies our frames sit a class tighter than it asks.
| The method asks for | Dak supplies | |
|---|---|---|
| Capacity | As the alloy and section require. Most metals gain strength as temperature falls, so a cold coupon draws more force than the same coupon at room temperature. | Load cells from 1 kg to 60 ton on the Series 7200, and 0.5 to 100 kN on the Series 9000 |
| Force accuracy | ISO 7500-1 Class 1 | ISO 7500-1 Class 0.5 — a class tighter than the method asks |
| Strain measurement | An extensometer to ISO 9513 Class 1 or better for proof strength (plastic or total extension); Class 2 acceptable for other properties at higher extension. Extensometer gauge length not less than 10 mm, on the central portion of the parallel length., gauge length As ISO 6892-1, measured at room temperature before cooling. The extensometer gauge length shall be not less than 10 mm and may be established by any of four defined methods, which must be recorded. | Certified to ISO 9513 Class 1 and ASTM E83 — non-contact video, clip-on and high-elongation |
| Gripping | a cooling device or bath around the test piece with the load train passing through it, and grips kept outside the cold zone | Our self-tightening serrated wedge grips, with V-jaws for round specimens, built to the specimen |
| Environment | Sub-ambient: +10 °C down to −196 °C, held within ±3 °C of the specified temperature with a gradient along the test piece of no more than 3 °C | 3009 series chambers, −150 °C to +400 °C — temperature only |
This page describes the method as practised. The governing text is the current edition from the issuing body. Tell us what you are testing and we will answer with the machine, the fixture and a quotation.
