top of page

ISO 604
Compressive Properties of Plastics

Universal Grip ISO 604 multi-cavity specimen mould

Downloads: ISO 604 Multi-Cavity Specimen Mould – Universal Grip Data Sheet (PDF)

Written by Chris Thomas, Universal Grip Company · Reviewed by Frank Bertini · Updated October 24, 2026

ISO 604 is the international standard for the compressive properties of plastics. A short prism, cylinder or tube is compressed between two parallel plates at a constant speed until it yields, breaks or reaches an agreed strain, and the force-deformation curve gives the compressive strength, the compressive yield stress, the stress at a given strain and the compressive modulus. It is the compression counterpart of the tensile test in ISO 527-1 and the flexural test in ISO 178, and the ISO twin of ASTM D695.

The standard is short and the equipment looks simple, which is why most of the pages that rank for it give no numbers at all. This guide covers the current edition and its status; the two preferred specimens, 50 × 10 × 4 mm for modulus and 10 × 10 × 4 mm for strength, and why there are two; the small specimens; the buckling criterion that decides the specimen length; the speed rule that gives 1, 1 or 5 mm/min; the 0.05 to 0.25 % chord modulus and the ± 1 µm extensometer it demands; the plate flatness, alignment and squareness tolerances; the procedure; every formula with a checked worked example; the compliance correction of Annex C; the common errors; a numeric comparison with ASTM D695; and the platens, extensometers, machines and specimen-preparation tooling Universal Grip supplies. The compression testing page gives the lab overview and the ISO standards hub lists every ISO method we equip.

In this guide: What ISO 604 Measures and Where It Applies · Edition and Status · Specimens: Type A, Type B and the Small Types · The Buckling Criterion and Choosing a Length · Apparatus: Plates, Alignment, Force and Strain · Test Speed: the 0.02l, 0.1l and 0.5l Rule · Conditioning and Procedure · Formulas and a Worked Example · Common Errors and How Large They Are · ISO 604 vs ASTM D695: the Numbers · Equipment for ISO 604: Platens, Extensometers, Machines, Specimen Preparation · FAQ

What ISO 604 Measures and Where It Applies

ISO 604 applies to rigid and semi-rigid thermoplastics, thermosets and thermotropic liquid-crystal polymers, including filled and reinforced compounds with fibres up to 7.5 mm long before processing. It is not normally used for textile-reinforced materials, for fibre-reinforced composite laminates (ISO 14126), for rigid cellular plastics (ISO 844) or for sandwich structures. The properties defined are:

  • Compressive stress σ: force divided by the original cross-section, in MPa.

  • Compressive strain ε: change in gauge length divided by the gauge length, measured with an extensometer; nominal compressive strain εc is the change in plate separation divided by the specimen length.

  • Compressive modulus Ec: the chord slope between ε1 = 0.0005 and ε2 = 0.0025 (0.05 to 0.25 % strain), the same window as the tensile and flexural moduli in ISO 527 and ISO 178.

  • Compressive yield stress σy: the first stress at which strain increases without an increase in stress.

  • Compressive strength σM: the maximum stress sustained, and stress at break σB if the specimen fractures.

  • Stress at x % strain σx: reported for materials with no yield point, with x set by the product standard or by agreement and below the strain at strength.

  • Strains at yield, strength and break and the ISO 2602 statistics for the set.

Compression of a short block differs from tension in two ways that shape the standard: a long specimen buckles before it fails, and a short specimen is restrained by friction at the plates and gives a modulus that cannot be read from the crosshead. ISO 604 answers both with two specimens and a buckling criterion rather than with a support jig.

Edition and Status

ISO 604:2002 is the third edition and the current one. It replaced the 1993 edition, has no amendments, and was last reviewed and reconfirmed in September 2022; no revision is in progress. The European adoption is EN ISO 604:2003 and the Japanese adoption JIS K 7181. One honest weakness of the 2002 text is that its precision clause says the interlaboratory data have not been established, so there are no published repeatability or reproducibility values for compressive strength or modulus to compare a lab against; the five-specimen statistics and the apparatus tolerances carry the whole burden of comparability.

Specimens: Type A, Type B and the Small Types

Specimens are prisms, cylinders or tubes compressed along their long axis. The preferred sizes, all with ± 0.2 mm on width and thickness, are:

  • Type A, for modulus: 50 ± 2 mm long, 10 mm wide, 4 mm thick; length-to-thickness ratio 12.5. Long enough to carry an extensometer, too slender to reach a strength without buckling.

  • Type B, for strength: 10 mm long, 10 mm wide, 4 mm thick; length-to-thickness ratio 2.5. Short enough to reach yield or fracture without buckling, too short for a modulus.

  • Type 1, small, for strength: 6 × 5 × 3 mm, from Annex A, for when only small pieces are available.

  • Type 2, small, for modulus: 35 × 5 × 3 mm with a 15 mm gauge length recommended.

Both preferred specimens are cut from the 10 × 4 mm centre section of the ISO 20753 type A1 multipurpose bar, the same bar that gives the ISO 527-2 tensile specimen and the ISO 178 flexural bar, so a single moulding tool serves all three tests. They may also be moulded to ISO 293, 294-1 or 295 or machined to ISO 2818. Two machining requirements matter more than the nominal dimensions: the end faces must be perpendicular to the long axis within 0.025 mm, and the sides must be parallel. Injection-moulded bars carry 1 to 2° of draft, which on a 4 mm thickness is enough to seat a Type B block on one edge, so moulded specimens are machined square. At least five specimens are tested for an isotropic material and ten for an anisotropic one, five along and five across the principal axis; a specimen that fails at an obvious flaw is replaced.

The Buckling Criterion and Choosing a Length

ISO 604 does not use a slenderness-ratio rule. It requires that the maximum nominal strain reached in the test, εc*, satisfies εc* ≤ 0.4 (x / l)², where x is the thickness or diameter and l the length. Annex B derives it from Euler buckling: a prism buckles at roughly εb ≈ 0.8 (h / l)² and a cylinder at 0.6 (d / l)², and the 0.4 factor keeps the test strain at half the prism value. Applied to the preferred specimens:

  • Type A (x/l = 4/50): εc* ≤ 0.4 × 0.0064 = 0.26 %, which is almost exactly the 0.25 % top of the modulus window. Type A is a modulus specimen and nothing else; the standard's guidance that modulus specimens should have x/l above 0.08 puts Type A at the limit.

  • Type B (x/l = 4/10): εc* ≤ 0.4 × 0.16 = 6.4 %, so a Type B block can be taken to yield on most engineering plastics; the guidance for strength specimens is x/l of at least 0.4, which gives about 6 % usable strain.

  • Ductile or strain-sensitive materials: choose the length so that εc* is two to three times the strain you expect to reach, which for a polyolefin yielding near 10 % means a block shorter than 10 mm or a thicker section.

This is why there are two specimens. A 50 mm bar gives a clean modulus from a 25 to 125 µm extensometer reading but would bow sideways long before yield; a 10 mm block reaches yield but any modulus taken from its crosshead displacement is dominated by plate friction, end effects and machine compliance. Testing one specimen for both, which some labs do to save time, violates the criterion on one property or the other.

Apparatus: Plates, Alignment, Force and Strain

  • Compression plates: hardened steel, polished faces flat within 0.025 mm, parallel to each other and perpendicular to the load axis; the load must be applied axially within 1 part in 1,000. A self-aligning spherical seat on one plate is the usual way to meet the parallelism requirement with a real machine.

  • Force: ISO 7500-1 class 1, ± 1 % of reading.

  • Strain: an extensometer accurate to ± 1 % of the strain interval it measures. For the 0.2 % modulus interval on a 50 mm Type A gauge that is 0.1 mm × 1 % = ± 1 µm, which rules out crosshead displacement and most general-purpose clip-ons; strain gauges must resolve 2 × 10⁻⁵. Nominal strain from plate separation is allowed when it is corrected for machine compliance per Annex C.

  • Dimensions: micrometer reading to 0.01 mm; semi-rigid materials measured under a flat foot at 20 ± 3 kPa.

  • Speeds: Table 1 lists 1, 2, 5 and 10 mm/min with ± 20 % tolerance and 20 mm/min with ± 10 %.

The 1:1,000 axiality and 0.025 mm flatness are tighter than most labs check. A Type B block that is 0.05 mm out of square on one end face sits on a line, not a face, and the first 0.1 mm of crosshead travel is spent bedding it in while the force rises non-linearly; on a brittle thermoset that bedding-in can shave 10 to 20 % off the apparent strength. Squaring the specimen ends and verifying the plates with a straightedge and feeler gauge, or a spherical seat that is unlocked at the preload and locked for the test, removes it.

Test Speed: the 0.02l, 0.1l and 0.5l Rule

The speed is tied to the specimen length so that the nominal strain rate is constant whatever the specimen: v = 0.02 l for modulus, v = 0.1 l for the strength of brittle materials and v = 0.5 l for the strength of materials that yield, with l in mm and v in mm/min, rounded to the nearest Table 1 speed. For the preferred specimens:

  • Type A modulus: 0.02 × 50 = 1 mm/min, a nominal strain rate of 2 %/min, so the 0.25 % top of the window is reached in about 8 seconds.

  • Type B strength, brittle material: 0.1 × 10 = 1 mm/min, 10 %/min nominal.

  • Type B strength, yielding material: 0.5 × 10 = 5 mm/min, 50 %/min nominal, which is the rate most engineering thermoplastics are run at.

'1 mm/min for modulus and 5 mm/min for strength' is therefore correct only for yielding materials; a brittle phenolic or a highly filled thermoset is run at 1 mm/min for strength too. The speed and the specimen type go into the result designation, for example 'Compressive test ISO 604 /A/ 1' for a modulus test on Type A at 1 mm/min, because the yield stress of a viscoelastic material rises with rate and a result without its speed cannot be compared.

Conditioning and Procedure

Specimens are conditioned and tested in the ISO 291 standard atmosphere, 23 ± 2 °C and 50 ± 10 % RH, unless the material standard says otherwise; humidity conditioning may be omitted for materials that are insensitive to it. The procedure is then:

  • Measure width and thickness (or diameter) at three points along the length to 0.01 mm and use the mean cross-section; measure the length to 1 %.

  • Place the specimen centred between the plates with its long axis on the load axis; for Type A attach the extensometer at mid-length.

  • Apply the preload: for modulus a stress no higher than 5 × 10⁻⁴ times the expected modulus (0.05 % prestrain), for strength no more than 1 % of the expected strength.

  • Run at the speed from the rule above and record force against extensometer strain (Type A) or against plate separation corrected for compliance (Type B).

  • Stop at fracture, at yield, or at the agreed strain; the maximum nominal strain must not exceed the buckling limit.

  • Note any lubricant or abrasive disc used between specimen and plate, which is a report item because it changes the end restraint and therefore the strength of a short block.

The Annex C compliance correction is the same idea as in ISO 178: the load train is loaded against a stiff steel reference, its deflection per newton CM is recorded, and the specimen deformation is ΔL = s − CM × F where s is the crosshead or plate travel. It lets a Type B test report a nominal strain at yield, but it does not rescue a modulus from a 10 mm block, because end effects and friction are not compliance.

Formulas and a Worked Example

  • Compressive stress: σ = F / A, with A the mean original cross-section in mm².

  • Compressive strain: ε = ΔL0 / L0 from the extensometer gauge length; nominal strain εc = ΔL / L from the plate separation and the specimen length. The denominator is always a length along the load axis, never the thickness.

  • Compressive modulus: Ec = (σ2 − σ1) / (ε2 − ε1) with ε1 = 0.0005 and ε2 = 0.0025, or the least-squares slope over that interval.

  • Results: stresses and modulus to three significant figures, strains to two; mean, standard deviation and 95 % confidence interval per ISO 2602.

A worked example on a polycarbonate. Modulus, Type A: the specimen measures 10.0 × 4.0 mm, so A = 40.0 mm², and a 50 mm extensometer is fitted. At 0.05 % strain (25 µm on the gauge) the force is 60 N and at 0.25 % (125 µm) it is 300 N, giving σ1 = 60 / 40 = 1.5 MPa and σ2 = 300 / 40 = 7.5 MPa, and Ec = (7.5 − 1.5) / (0.0025 − 0.0005) = 6.0 / 0.002 = 3,000 MPa. The nominal strain at the top of the window is 0.25 %, inside the 0.26 % buckling limit. Strength, Type B: a block measured at 10.02 × 4.01 mm has A = 40.2 mm²; run at 5 mm/min it yields at a maximum force of 3,400 N, so σy = σM = 3,400 / 40.2 = 84.6 MPa, reported as 84.6 MPa with the designation 'ISO 604 /B/ 5'. The nominal strain at yield from the compliance-corrected plate travel is, say, 0.65 mm on 10 mm = 6.5 %, which exceeds the 6.4 % limit for Type B, so a lab wanting the strain at yield on this material would use a shorter or thicker block.

Common Errors and How Large They Are

  • Modulus from a Type B block or from crosshead travel. Plate friction, end bedding and load-train compliance make the apparent modulus 30 to 60 % low and non-repeatable; only an extensometer on a Type A specimen, resolving 1 µm, gives an ISO 604 modulus.

  • Strength from a Type A bar. Above about 0.3 % strain a 50 mm bar bows; the force maximum you record is a buckling load, not a compressive strength, and can be half the true value.

  • Speed chosen by habit. 5 mm/min on a brittle thermoset instead of 1 mm/min raises the strength by a few percent and is not an ISO 604 value; 1 mm/min on a ductile polyolefin takes a minute to yield and reads low.

  • Ends not square. 0.05 mm of out-of-squareness on a 10 mm block produces a non-linear toe and a strength 10 to 20 % low on brittle materials; the standard's limit is 0.025 mm.

  • Plates worn or not parallel. Dished or scored plate faces load the block on an edge; verify flatness to 0.025 mm and use a spherical seat.

  • Stress at x % reported without the x. σ10 and σ5 on the same material differ by 20 % or more; the strain must be in the designation.

  • Anisotropic material tested in one direction. Glass-filled compounds differ by 20 to 40 % between flow and cross-flow; ten specimens, five each way, are required.

  • Strain divided by thickness. Nominal strain is plate travel over specimen length; dividing by the 4 mm thickness instead of the 10 mm length multiplies every strain by 2.5.

  • Comparing with ASTM D695 numbers. The ASTM block is 25.4 mm long at 1.3 mm/min, about 5 %/min nominal, against 50 %/min for Type B; yield stresses of ductile plastics differ accordingly.

ISO 604 vs ASTM D695: the Numbers

ASTM D695 measures the same properties with a different specimen philosophy. The differences that move the numbers:

  • Strength specimen: ISO Type B 10 × 10 × 4 mm (40 mm²); ASTM a right prism or cylinder with length twice its width, 12.7 × 12.7 × 25.4 mm or 12.7 mm diameter × 25.4 mm (161 and 127 mm²).

  • Modulus specimen: ISO Type A 50 × 10 × 4 mm with an extensometer; ASTM a rod with a slenderness ratio of 11 to 16, typically 12.7 × 12.7 × 50.8 mm.

  • Thin materials: ISO offers the small Type 1 and Type 2 specimens; ASTM uses a 79.4 × 12.7 mm strip in an anti-buckling support jig for material under 3.2 mm.

  • Speed: ISO 1 mm/min (modulus and brittle strength) or 5 mm/min (yielding strength), scaled to the length; ASTM a fixed 1.3 ± 0.3 mm/min for everything.

  • Modulus definition: ISO chord between 0.05 and 0.25 % strain; ASTM tangent to the initial linear portion, with secant as an option.

  • Buckling: ISO gives the explicit criterion εc* ≤ 0.4 (x/l)²; ASTM relies on the 2:1 block and the support jig.

  • Strain measurement: ISO extensometer to ± 1 % of the interval (± 1 µm on Type A); ASTM E83 class B-2.

  • Count and conditioning: both five isotropic or ten anisotropic; ISO 291 and ASTM D618 both at 23 °C / 50 % RH, ASTM with a 40 hour minimum.

  • Precision: ISO 604 states none; ASTM D695 publishes interlaboratory data.

Moduli from the two standards land close on linear materials because both take the slope at small strain on a slender specimen. Strengths and yield stresses do not transfer: the ISO Type B block at 5 mm/min is strained ten times faster than the ASTM block at 1.3 mm/min, and the 2.5 versus 2.0 length-to-width ratios change the end restraint. The claim on some reference sites that the two can be used interchangeably holds for early materials screening of modulus, not for a specification value.

Equipment for ISO 604: Platens, Extensometers, Machines, Specimen Preparation

Compression platens and alignment

ISO 604 needs hardened, polished, parallel plates larger than the specimen footprint and an axial load path. Universal Grip compression platens are ground flat and supplied in diameters from 50 to 150 mm with a self-aligning spherical seat on the upper platen, which is how the 1:1,000 axiality and the parallelism requirement are met on a production frame; the compression fixtures page covers guided and caged versions for taller specimens and the sub-press arrangements used when a frame's own alignment is not good enough. For strain on a Type B block without an extensometer, a transducer between the plates measures plate separation directly and takes the frame compliance out of the nominal strain.

Capacity

Forces are modest but not trivial: a Type B block (40 mm²) at 85 MPa needs 3.4 kN and a 300 MPa glass-filled thermoset needs 12 kN, so a 5 kN frame is marginal for reinforced compounds and 10 to 50 kN is the usual range. The 10 kN, 25 kN and 50 kN benchtop machines run the two-specimen sequence with the 0.05–0.25 % regression and the compliance correction in software; a 1 or 2 kN load cell on the Type A modulus test keeps the 60 to 300 N forces of the worked example in the class 1 range.

Extensometers

  • Type A modulus: a clip-on extensometer with a 25 or 50 mm gauge and 1 µm resolution, mounted on the 10 mm face, or a class 1 video extensometer tracking two marks on the bar, which also shows whether the bar is beginning to bow. The extensometers page compares the options.

  • Type B nominal strain: plate-separation transducer or compliance-corrected crosshead; an extensometer cannot be mounted on a 10 mm block.

  • Temperature: environmental chambers from −70 to +250 °C with platens and extensometer rated for the chamber; compressive yield of most thermoplastics falls by roughly half between 23 °C and 80 °C.

Specimen preparation: the Universal Grip ISO 604 mould

Nobody else on this subject talks about where the specimens come from, which is where the result is actually decided. The Universal Grip ISO 604 multi-cavity specimen mould, part UGC-ISO604-SC, is a hinged compression mould for a heated laboratory press: the ground steel cavity plate forms several multipurpose dumbbell bars per press cycle at a nominal 4 mm thickness, the hinged ground cover plate keeps cavity and cover aligned, and integral side handles allow the mould to be loaded and unloaded from a hot press safely. The 50 × 10 × 4 mm Type A and 10 × 10 × 4 mm Type B pieces are then machined from the bars with the end faces squared to the standard's 0.025 mm, and because the bars are compression moulded there is no draft to remove. Cavity count and geometry are made to order, including other ISO specimen types, and the mould is specified by specimen type and thickness, number of cavities and press platen size. It is specimen-preparation tooling, not a platen set; pair it with the platens above for the test itself, and see the plastics page for the related ISO 527-2 and ISO 178 tooling.

ISO 604 FAQ

What is ISO 604?

The international standard for the compressive properties of plastics. A prism, cylinder or tube is compressed between parallel plates at constant speed to yield, fracture or an agreed strain, giving compressive strength, yield stress, stress at a given strain and compressive modulus. The current edition is ISO 604:2002, reconfirmed in 2022.

What specimen sizes does ISO 604 use?

Two preferred specimens cut from the ISO 20753 multipurpose bar: Type A, 50 × 10 × 4 mm, for modulus, and Type B, 10 × 10 × 4 mm, for strength. Small specimens Type 1 (6 × 5 × 3 mm, strength) and Type 2 (35 × 5 × 3 mm, modulus) are in Annex A. End faces must be square within 0.025 mm; five specimens, or ten for anisotropic materials.

Why does ISO 604 use two different specimens?

Because of buckling. The 50 mm Type A bar can carry an extensometer for the 0.05–0.25 % modulus but bows before yield; the 10 mm Type B block reaches yield without buckling but gives no usable modulus from crosshead travel. The limit is εc* ≤ 0.4 (x/l)², which is 0.26 % strain for Type A and 6.4 % for Type B.

What is the test speed in ISO 604?

Scaled to specimen length: 0.02 l for modulus, 0.1 l for brittle strength and 0.5 l for yielding strength, rounded to the Table 1 speeds of 1, 2, 5, 10 or 20 mm/min. That gives 1 mm/min for Type A modulus, 1 mm/min for Type B brittle strength and 5 mm/min for Type B yielding strength.

How is compressive modulus calculated in ISO 604?

Ec = (σ2 − σ1) ÷ (ε2 − ε1) between 0.05 % and 0.25 % strain, as a chord or least-squares slope, on a Type A specimen with an extensometer accurate to ± 1 % of the interval, which is ± 1 µm on a 50 mm gauge. In the worked example 60 N and 300 N on 40 mm² give 3,000 MPa.

What are the plate and alignment requirements in ISO 604?

Hardened steel plates with polished faces flat within 0.025 mm, parallel and perpendicular to the load axis, with the load applied axially within 1 in 1,000; a self-aligning spherical seat is the normal way to achieve it. Force to ISO 7500-1 class 1.

What preload does ISO 604 allow?

For modulus a stress no higher than 5 × 10⁻⁴ times the expected modulus, keeping the prestrain below the 0.05 % start of the window; for strength no more than 1 % of the expected strength.

What is the difference between ISO 604 and ASTM D695?

ISO uses a 10 × 10 × 4 mm block for strength at 1 or 5 mm/min and a 50 × 10 × 4 mm bar with an extensometer for a 0.05–0.25 % chord modulus; ASTM D695 uses a 12.7 × 12.7 × 25.4 mm block, an 11–16 slenderness rod for a tangent modulus, a support jig for thin strips and a fixed 1.3 mm/min. Moduli are comparable; strengths and yield stresses are not.

Does ISO 604 have precision data?

No. The 2002 edition states that interlaboratory precision has not been established, so there are no published repeatability or reproducibility values; the five-specimen mean, standard deviation and 95 % confidence interval per ISO 2602 are reported instead.

How many specimens does ISO 604 require?

At least five for an isotropic material and at least ten for an anisotropic one, five parallel and five normal to the principal axis. Specimens that fail at an obvious flaw are replaced. Conditioning is ISO 291, 23 ± 2 °C and 50 ± 10 % RH.

What machine capacity is needed for ISO 604?

Cross-section times expected strength: a Type B block (40 mm²) needs 3.4 kN at 85 MPa and 12 kN for a 300 MPa glass-filled thermoset, so 10 to 50 kN benchtop frames cover the range and 5 kN is marginal for reinforced compounds. Type A modulus loads are 60 to 300 N and suit a 1 or 2 kN load cell.

What does Universal Grip supply for ISO 604?

Ground compression platens with a self-aligning spherical seat, guided compression fixtures, clip-on and video extensometers for the Type A modulus, environmental chambers, 10 to 50 kN benchtop machines, and the UGC-ISO604-SC hinged multi-cavity compression mould that forms the 4 mm multipurpose bars from which Type A and Type B specimens are machined.

Need compression platens, an extensometer or specimen moulds for ISO 604? Contact Universal Grip with your specimen type, expected strength and machine connection and we will quote the right setup.

bottom of page