ASTM E8 vs ISO 6892-1: tensile testing of metals, differences and requirements (NBR ISO 6892-1 and A370)

ASTM E8/E8M and ISO 6892-1 are the two standards for tensile testing of metallic materials at room temperature. The current editions are ASTM E8/E8M-25, ISO 6892-1:2019 and, in Brazil, ABNT NBR ISO 6892-1:2024, which received a corrected version in July 2026. Both measure the same properties (yield strength, tensile strength, elongation and reduction of area), but they require different things: the gauge length of the test specimen, how the testing speed is controlled and its tolerance, the symbols, the rounding, and the standards that qualify the testing machine and the extensometer. In practice, the strength values measured by the two standards agree, and the elongation values do not.

In short:

• Round specimen: the gauge length for elongation is 4 diameters in E8 (inch-pound), 5 diameters in E8M (SI) and 5.65 times the square root of the cross-sectional area in ISO, which also works out to 5 diameters.
• Speed: the nominal value before yield is the same in both standards; ISO allows ±20% and E8 Method B allows about ±40% (in E8/E8M-09, the edition we consulted).
• Machine and extensometer: ISO 7500-1 and ISO 9513 in the ISO family; ASTM E4 and ASTM E83 in the ASTM family.

What each standard is and where it applies

The test itself is the same in both families: a test specimen with a known original cross-sectional area, S₀, is subjected to an increasing axial tensile force until it fractures, and the machine records force and extension. Before the test, two marks are made on the reduced section, separated by the original gauge length L₀, the base over which elongation is measured. The maximum force gives the tensile strength; the beginning of the curve gives the yield strength; the marks, after fracture, give the elongation. The curve, the step-by-step procedure, the use of the extensometer and the basic calculations are covered in the article on the tensile test. Here the subject is what each standard requires and where they differ.

• ASTM E8/E8M-25 (active since August 26, 2025): test methods for tension testing of metallic materials in any form, between 10 and 38 °C. It is a dual-designation standard: E8 uses inch-pound units and E8M uses SI units.
• ISO 6892-1:2019: third edition of the international standard for tensile testing of metals at room temperature, between 10 and 35 °C (23 ± 5 °C when the test requires controlled conditions). It replaced the 2016 edition.
• ABNT NBR ISO 6892-1:2024: second edition, dated April 21, 2024, an identical Brazilian adoption of ISO 6892-1:2019, issued by committee ABNT/CB-004, 79 pages. It cancels and replaces the 2013 edition and received Erratum 1 and the corrected version on July 13, 2026. Anyone buying the standard in Brazil today should ask for the corrected version.
• ASTM A370-26 (active since May 21, 2026): test methods and definitions for mechanical testing of steel products, covered in its own section further on.

Test specimens and gauge length

The photo shows machined metal test specimens: the round ones have threaded ends and a reduced section in the middle, and the flat one is cut from sheet. Both standards define the dimensions of these two types, and this is where the differences between them begin.

Round: 4D, 5D and 5.65√S₀

ISO 6892-1 prefers the proportional test specimen, in which L₀ grows with the cross section: L₀ = k·√S₀, with k = 5.65 and L₀ of at least 15 mm. Since the area of a round specimen is π·d₀²/4, the calculation 5.65·√S₀ gives 5.01·d₀: in practice, five diameters. When the cross section is too small for the 15 mm, k = 11.3 or a fixed L₀, such as 80 mm, is used. Proportionality makes it possible to compare the elongation of test specimens of different sizes.

In E8/E8M, the gauge length of the round specimen depends on the unit system of the standard. The standard specimen is 12.5 mm in diameter; under E8, the gauge length is 50 mm (4D); under E8M, 62.5 mm (5D). ASTM itself states that the gauge length is the most significant difference between E8 and E8M test specimens. The smaller round specimens follow the same rule: diameters of 9 mm and 6 mm give gauge lengths of 36 and 24 mm in E8 and of 45 and 30 mm in E8M. These dimensions were checked in the 2009 edition (E8/E8M-09); the 4D and 5D rule is confirmed in the scope of E8/E8M-25.

ISO also specifies what surrounds L₀ on the machined round specimen, in Annex D: a parallel length of at least L₀ + d₀/2, or L₀ + 2d₀ in case of dispute, and a transition radius of at least 0.75·d₀. The diagram below applies these rules to the 10 mm round specimen from the table in the standard (L₀ = 50 mm and L_c of at least 55 mm) and shows, at the bottom, what is measured after fracture: L_u, with the halves fitted together, and the minimum cross section S_u, which give A and Z.

Sheet: fixed width and gauge length

For sheet, both standards use a fixed L₀, with no proportionality. In E8/E8M-09, the sheet-type specimen is 12.5 mm wide with a 50 mm gauge length, for thicknesses up to 19 mm, and the plate-type specimen is 40 mm wide with a 200 mm gauge length, for thick plate; check the dimensions in E8/E8M-25. ISO covers thin products, from 0.1 to 3 mm, in Annex B, and Type 2 in that annex is 20 mm wide, with an 80 mm gauge length and a 120 mm parallel length (recommended; the minimum is 90 mm). Above 3 mm, the proportional rule generally applies again (Annex D).

The diagram below compares the two pairs: on the left, the same 12.5 mm round specimen with the 50 mm gauge length of E8 and the 62.5 mm gauge length of E8M and ISO; on the right, the 12.5 × 50 mm sheet specimen of E8 and the 20 × 80 mm ISO Type 2. The orange marks delimit L₀.

For those who machine the test specimen, the next diagram gives the complete dimensions, with tolerances, of the two standard E8 formats: the sheet-type specimen and the 12.5 mm round specimen, with the 50 mm gauge length of E8 and the 62.5 mm gauge length of E8M. They were taken from the 2009 edition of the standard, the one we consulted in PDF. The current edition is E8/E8M-25, and the dimensions should be checked against it before releasing a manufacturing drawing, as the footnote of the image below indicates.

In all cases, the area used is measured, not nominal. A 1% error in S₀ becomes a 1% error in every stress value, so the diameter, or the width and thickness, is measured with a micrometer at more than one location.

Speed control: ISO methods A1, A2 and B and E8 methods A, B and C

The yield strength of many steels depends on the strain rate: speeding up the test increases the measured value. That is why both standards specify the speed up to yield more strictly than after it. They use different names for similar ideas, and the letters do not match.

ISO 6892-1 has two methods:

• Method A, strain rate control. In A1, the machine controls strain in closed loop, using the extensometer signal. In A2, it keeps the crosshead speed constant, calculated from the parallel length: v_c = L_c × ė, where L_c is the parallel length and ė is the target rate. The rates come in ranges, all with a tolerance of ±20%: range 1, 0.00007 s⁻¹; range 2, 0.00025 s⁻¹; range 3, 0.002 s⁻¹; range 4, 0.0067 s⁻¹. For R_p and R_eH, ranges 1 or 2 are used; for R_eL and A_e, ranges 2 or 3, by crosshead; for tensile strength and elongation, ranges 2, 3 or 4 (range 2 recommended before yield and range 4 after).
• Method B, stress rate control. In the elastic range, 6 to 60 MPa/s for materials with a modulus of elasticity of at least 150 GPa, such as steel, and 2 to 20 MPa/s below that, such as magnesium, aluminum alloys, brass and titanium (examples given by the standard itself).

The ISO test report summarizes the choice in a code: "A224" means method A, range 2 for R_eH or R_p, range 2 for R_eL and range 4 for R_m and elongation.

In method A, range 2 is recommended up to yield and also during the yield plateau, when R_eL is measured; after yield, range 4 is recommended, and range 3 can be used for both phases. In method B, the stress rate is only imposed in the elastic range: during yielding, the strain rate does not exceed 0.0025 s⁻¹ (for R_eL, it stays between 0.00025 and 0.0025 s⁻¹) and, after yield, it goes up to 0.008 s⁻¹. The diagram below shows what applies in each phase of the test in both methods.

E8/E8M has three methods for the portion up to yield. The values below are those of E8/E8M-09, the edition we consulted; the current edition is E8/E8M-25, and the rates should be checked against it before setting up a test method:

• Method A, stress rate: 1.15 to 11.5 MPa/s.
• Method B, strain rate measured by the extensometer: 0.015 ± 0.006 mm/mm/min.
• Method C, crosshead speed: 0.015 ± 0.003 mm/mm/min multiplied by the length of the reduced section.

After yield, for materials with an expected elongation above 5%, E8/E8M-09 accepts 0.05 to 0.5 mm/mm/min.

The comparison has a point few people notice: 0.015 min⁻¹ divided by 60 is exactly 0.00025 s⁻¹, ISO range 2. The nominal value is the same; what changes is the tolerance. In E8/E8M-09, method C allows ±0.003 min⁻¹, that is, ±20%, like ISO; method B allows ±0.006 min⁻¹, about ±40%. In terms of control, E8 method B corresponds to ISO A1, and method C to A2. The stress rate ranges, however, only overlap between 6 and 11.5 MPa/s: a steel test at 20 MPa/s meets ISO and does not meet E8.

The chart below shows the ranges converted into crosshead speed for a parallel length of 75 mm, the one in the worked example further on. Before yield, ISO range 2 and E8 method C fall in the same interval, and E8 method B is wider. After yield, ISO range 4, around 30 mm/min, fits within the E8 range. E8 method B is shown converted to crosshead speed for comparison only: in the standard, its rate is the one measured by the extensometer.

With crosshead control (A2 or method C), the actual rate in the test specimen is lower than the nominal rate, because part of the displacement is deformation of the machine and the grips; for rate-sensitive materials, closed-loop control by extensometer is more repeatable. Yielding itself is covered in the article on the elastic limit.

Yield, tensile strength and elongation: what each standard calls what

The properties are the same; the names and some criteria differ. ISO uses symbols (R_p0.2, R_eH, A); ASTM uses spelled-out terms (offset yield strength, elongation).

Yield strength

In a material without a yield point, both standards use the offset method: draw a line parallel to the elastic portion, offset by a specified plastic strain, and read the stress where it intersects the curve. With 0.2%, ISO calls the result R_p0.2; E8 calls it the 0.2% offset yield strength. ISO also defines R_t, the stress at a specified total extension, which corresponds to the E8 extension-under-load method used in some tube and bar specifications.

In steel with a yield point, such as rolled low-carbon steel, the force drops after the first peak and fluctuates along a plateau. ISO defines two values: R_eH, the upper yield strength, which is the stress at the peak before the first drop, and R_eL, the lower yield strength, the lowest stress on the plateau, disregarding the initial transient effect. The length of the plateau is A_e. E8 has the same two values under other names: upper yield strength (UYS) and lower yield strength (LYS), read from the recorded diagram with the same logic (first peak; lowest value excluding transients), and the plateau length is the yield point elongation (YPE), per E8/E8M-09. Which value goes on the certificate is set by the product specification, not by the test standard.

Tensile strength, elongation and reduction of area

Tensile strength is the same in both standards: R_m = F_m/S₀, the maximum force divided by the original cross-sectional area. Reduction of area, the percentage decrease in area at the neck, is Z in ISO and reduction of area in E8; the phenomenon is covered in the article on necking.

Elongation after fracture is A = (L_u − L₀)/L₀ × 100, with L_u measured by fitting the two halves together. ISO writes the gauge length into the symbol: A (proportional, 5.65√S₀), A_11.3, A_50mm, A_80mm. E8 reports "elongation in 50 mm" or "in 4D". ISO also defines elongation at maximum force: A_gt, total, including the elastic part, and A_g, the plastic part only. Both measure ductility before necking and do not depend on the gauge length in the same way as A.

The two maximum-force values come from the same point on the curve: A_gt is the total extension read there, and A_g is what remains after subtracting the elastic part, using a line parallel to the slope of the elastic portion. The same logic applies at fracture: A_t is the total extension at the moment of fracture, and A is the plastic part, the one measured by fitting the halves together. The diagram below locates all four on the curve.

Rounding

ISO 6892-1 recommends (section 22) rounding the results, unless the product standard says otherwise: stresses to the nearest whole MPa, A_e to 0.1%, other elongations to 0.5% and reduction of area to 1%. E8 says to follow the material specification and ASTM E29; in their absence, E8/E8M-09 recommends stresses to 1 MPa up to 500 MPa, to 5 MPa between 500 and 1,000 MPa and to 10 MPa above that, and reduction of area to 0.5% below 10% and to 1% above. Close to the specified minimum, the rounding rule decides acceptance and must be stated in the report.

Machine and extensometer: ISO 7500-1, ISO 9513, ASTM E4 and ASTM E83

The test standards refer to their own standards for force and strain measurement, and each family has its own: a calibration certificate from one family does not prove compliance with the other.

• Force, ISO family: ISO 6892-1 requires a machine verified to ISO 7500-1:2018 (ABNT NBR ISO 7500-1:2023 in Brazil), class 1 or better. The class applies only within the verified range of each load cell.
• Force, ASTM family: E8 requires a machine verified to ASTM E4-24, whose criterion is an error of up to ±1% within the verified range. E4 itself warns that, once calibration uncertainty is added, the actual deviation can exceed 1%.
• Extensometer, ISO family: ISO 9513:2012 (ABNT NBR ISO 9513:2013 in Brazil), with classes 0.2, 0.5, 1 and 2. ISO 6892-1 requires class 1 for R_p and R_t and accepts class 2 for properties at extensions above 5%, such as A_gt (section 9). R_eH and R_eL are determined from force alone.
• Extensometer, ASTM family: ASTM E83-25, with classes A, B-1, B-2, C, D and E. E8/E8M-09 requires class B-2 or better for yield strength.

In both families, yield strength cannot be taken from crosshead displacement: it includes grip slack and machine deflection. The extensometer gauge length must also fit within the specimen gauge length: a 50 mm extensometer suits the 12.5 mm E8 round specimen; for the ISO 62.5 mm, yield is measured with a compatible extensometer and final elongation from the gauge marks. For R_p, ISO recommends an extensometer gauge length between half of L₀ and about 90% of the parallel length; for properties at or after maximum force, close to L₀ (8.3).

Summary: ASTM E8/E8M vs. ISO 6892-1 side by side

The table brings together what the previous sections explained, for quick reference when setting up the test method or reading a test report:

Item	ASTM E8/E8M-25	ISO 6892-1:2019 and NBR ISO 6892-1:2024
Temperature	10 to 38 °C	10 to 35 °C (23 ± 5 °C controlled)
Units	E8: inch-pound; E8M: SI	SI
12.5 mm round specimen	L0 = 50 mm (4D, E8) or 62.5 mm (5D, E8M)	L0 = 5.65√S0 ≈ 62.5 mm (5d0)
Sheet	12.5 × 50 mm (sheet); 40 × 200 mm (plate)	up to 3 mm: Annex B, type 2, 20 × 80 mm
Speed up to yield	A: 1.15 to 11.5 MPa/s; B: 0.015 ± 0.006 min⁻¹; C: 0.015 ± 0.003 min⁻¹ (E8/E8M-09)	A1/A2: 0.00025 s⁻¹ ±20% (or 0.00007); B: 6 to 60 MPa/s (E ≥ 150 GPa)
Speed after yield	0.05 to 0.5 min⁻¹ (E8/E8M-09)	A: 0.002 or 0.0067 s⁻¹ ±20% (range 4 recommended); B: up to 0.008 s⁻¹
Yield	offset, extension under load, diagram	Rp, Rt, ReH, ReL
Elongation	with the gauge length stated (50 mm, 4D)	A, A11.3, A80mm, Ag, Agt
Rounding	per the specification; otherwise 1/5/10 MPa depending on the range (E8/E8M-09)	recommended: 1 MPa; Ae 0.1%; A 0.5%; Z 1%
Machine	ASTM E4 (±1%)	ISO 7500-1, class 1
Extensometer	ASTM E83, class B-2 for yield (E8/E8M-09)	ISO 9513, class 1 for Rp and Rt; class 2 above 5%

Worked example: the same test specimen under both standards

A lot of cold-drawn 1020 steel bar needs test reports for two customers, one requiring ISO 6892-1 and the other ASTM E8. The lab machines a round specimen with d₀ = 12.50 mm (measured) and a 75 mm parallel length, scribes marks every 12.5 mm so both gauge lengths can be measured, and tests with an extensometer. Readings: force at the 0.2% offset of 42.95 kN and maximum force of 51.55 kN. After fracture, with the halves fitted together, the 62.5 mm gauge length measures 71.9 mm and the 50 mm one, centered on the fracture, measures 58.6 mm.

1. Area: S₀ = π × 12.50² / 4 = 122.72 mm².
2. Gauge lengths: ISO, L₀ = 5.65 × √122.72 = 5.65 × 11.078 = 62.59 mm, marked as 62.5 mm (5d₀); E8M, 5D = 62.5 mm; E8, 4D = 50.0 mm.
3. Yield: 42,950 / 122.72 = 349.99, rounded to 350 MPa, the same under both standards.
4. Tensile strength: 51,550 / 122.72 = 420.07, rounded to 420 MPa, the same under both standards.
5. Elongation per ISO (and E8M): A = (71.9 − 62.5) / 62.5 × 100 = 15.04, rounded to 15.0%.
6. Elongation per E8 (4D): (58.6 − 50.0) / 50.0 × 100 = 17.2%, reported as elongation in 50 mm.

Stresses do not depend on the gauge length and come out identical, with values consistent with a cold-drawn 1020 steel. Elongation changes by 2.2 percentage points without any change in the material, for the reason explained in the next section.

Crosshead speed, for L_c = 75 mm. ISO range 2 (A2): 75 × 0.00025 = 0.01875 mm/s = 1.125 mm/min, with ±20%, from 0.90 to 1.35 mm/min. E8 method C: 0.015 × 75 = 1.125 mm/min, from 0.90 to 1.35 mm/min. After yield, ISO range 4 gives 75 × 0.0067 = 0.5025 mm/s, about 30 mm/min, and E8 accepts from 0.05 × 75 = 3.75 to 0.5 × 75 = 37.5 mm/min. A program of 1.125 mm/min up to yield and 30 mm/min afterward meets both standards, but each report cites its own. If the lab prefers stress rate, the range common to both standards is 6 to 11.5 MPa/s, which for this area gives 0.74 to 1.41 kN/s.

Machine capacity. The maximum force was 51.6 kN. A 50 kN machine has no margin even for rolled 1020 (380 MPa × 122.72 mm² = 46.6 kN, 93% of capacity), and the cold-drawn bar exceeds it. This specimen calls for the 100 to 200 kN range. If the lab only has the 50 kN machine, it can use the smaller 6 mm E8 round specimen (S₀ = 28.27 mm²), which fractures at 420 × 28.27 = 11.9 kN, with a 24 mm gauge length under E8 and 30 mm under ISO (5.65 × √28.27 = 30.0 mm).

Why elongation under one standard is not valid under the other

Elongation after fracture has two components. The first is uniform: up to maximum force, the entire gauge section stretches evenly, and this component, as a percentage, is the same for any gauge length. The second is localized: after maximum force, deformation concentrates in the neck, which adds a few millimeters to the length, almost independently of the gauge length. Divided by a short gauge length, this fixed component weighs more; by a long one, it weighs less. That is why, in the same material, A_50mm is greater than A_62.5mm, which is greater than A_80mm or A_200mm.

The chart below shows this effect with the numbers from the example, using an illustrative model: a uniform component of 6.4% and a neck that adds about 5.4 mm. The curve passes through 17.2% at the 50 mm gauge length and 15.0% at 62.5 mm, and would drop to about 9% at a 200 mm gauge length.

The consequence for buyers and inspectors is direct: an ASTM certificate showing 17% in 50 mm does not prove that the material meets an ISO specification requiring 15% in 5.65√S₀, nor the other way around, because the two numbers measure different things. There is a conversion standard: ISO 2566-1:2021 converts elongations between gauge lengths for carbon and low-alloy steels with R_m from 300 to 700 MPa, and ISO 2566-2 covers austenitic steels. The conversion has limits: part 1 does not apply to cold-worked steel, such as the cold-drawn bar in the example, nor to quenched and tempered steel. Outside these limits, the safe route is to test to the standard the specification requires, or to mark both gauge lengths, as in the example.

ASTM A370: why the steel certificate cites it

Anyone receiving steel plate, tube, rebar or fasteners under an American specification sees A370 on the certificate, not E8. A370-26 brings together the methods and definitions for mechanical testing of steel, stainless steel and related alloy products: tension, bend, hardness (Brinell, Rockwell and portable) and impact. It has ten annexes for specific products, such as bar, tube, fasteners and wire, and ASTM product specifications cite it to demonstrate the required properties.

For tension, A370 works on the same principles as E8, which E8/E8M-25 itself lists as a related document; what A370 adds are the test specimens and rules for each steel product. For metals other than steel, such as aluminum, copper and titanium, A370 does not apply, and the tensile reference is E8/E8M.

Which standard to use

The choice is not technical, it is contractual: the test report follows the standard required by the material specification or the purchase order. In the absence of an explicit requirement, the practical rule is:

• ABNT specification (Brazilian standards), supply in Brazil or European specification: ABNT NBR ISO 6892-1:2024 (or ISO 6892-1:2019).
• Steel product under an ASTM specification (plate, tube, bar, fastener): ASTM A370, with tension by the E8 method.
• Other metal under an American standard, or a contract that cites E8: ASTM E8 or E8M, stating which.
• Customer requiring both: one test specimen with both gauge lengths marked and a speed program that satisfies both, as in the example; the reports are issued separately.

Common mistakes in applying the standards

Almost all application mistakes come from treating the two standards as interchangeable:

• Reading "E8" as always 4D. E8M is 5D; the report must state the gauge length.
• Comparing elongations from different gauge lengths (A_50mm, A, A_80mm) without conversion, or converting outside the limits of ISO 2566-1.
• Citing a superseded edition. E8-11, E8-13, ISO 6892-1:2009 and 2016 and NBR ISO 6892-1:2013 are still in circulation; the current ones are E8/E8M-25, ISO 6892-1:2019 and NBR ISO 6892-1:2024.
• Using the E8 method B tolerance in an ISO test. ±40% does not fit within the ISO ±20%.
• Measuring R_p0.2 from the crosshead or with an extensometer outside the required class (class 1 of ISO 9513; B-2 of E83).
• Changing the speed before yield has been determined. R_eH and R_p change with the rate.
• Using A370 for a nonferrous metal. It covers steel products.
• Testing at 36 or 38 °C and reporting it as the ISO normal condition. Above 35 °C, ISO requires the temperature to be recorded and reported; in E8, the range goes up to 38 °C.
• Omitting the speed code and the rounding rule from the ISO report.

Testing equipment for both standards

Both standards are run on a universal testing machine: the crosshead applies the tensile force at the programmed speed, the load cell measures the force, the extensometer measures the strain over the gauge length, and the software calculates the results. For the Biopdi M2 line, the product page states a class 1 load cell per ISO 7500-1 (ABNT NBR ISO 7500-1:2023 in Brazil), and its list of metals standards includes ISO 6892-1 (ABNT NBR ISO 6892-1:2024 in Brazil) and ASTM E8/E8M-25. The panel below shows the machine, the role each part plays in testing metals, and the fixtures used.

For thin sheet and small rounds, such as the 6 mm specimen in the example, the recommended range is the 5 to 50 kN machine; for the 12.5 mm steel round, bars and sections, the 100 to 200 kN range. The test method is set up in the BioLab software, which generates the report in PDF, HTML or CSV. How to size the machine in general is covered in our guide to the universal testing machine. To configure the machine for your standards, request a quote.

Standards cited

The table lists the standards in this guide with their current edition, checked on September 30, 2026 in the ASTM store, the ISO catalog and the Target (ABNT) listing:

Standard	Subject	Current edition
ASTM E8/E8M	tension testing of metallic materials	E8/E8M-25
ISO 6892-1	tensile testing of metals at room temperature	2019
ABNT NBR ISO 6892-1	Brazilian adoption of ISO 6892-1	2024, corrected version July 2026
ASTM A370	mechanical testing of steel products	A370-26
ISO 7500-1 and ABNT NBR ISO 7500-1	verification of machine force	2018 and 2023
ASTM E4	verification of machine force	E4-24
ISO 9513 and ABNT NBR ISO 9513	calibration of extensometers	2012 and 2013
ASTM E83	verification and classification of extensometers	E83-25
ISO 2566-1	conversion of elongation values, carbon and low-alloy steels	2021