ISO 844
Compression Properties of Rigid Cellular Plastics

Downloads: ISO 844 Compression Platens for Rigid Cellular Plastics – Universal Grip Data Sheet (PDF)

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

ISO 844 is the international standard for the compressive properties of rigid cellular plastics: the insulation boards, structural core foams and moulded rigid foams whose load-bearing performance is sold on one number, the compressive stress at 10 % deformation. A square or round prism is compressed between two flat parallel platens at a rate of 10 % of its thickness per minute until it reaches its maximum force or 10 % relative deformation, and the compressive strength or the stress at 10 %, and optionally the compressive modulus, are calculated from the force and the original cross-section.

This guide covers the current edition, ISO 844:2026, and what the 2014, 2021 and 2026 revisions changed; the preferred 100 × 100 × 50 mm prism and the rules for other sizes and thicknesses; the speed rule and how it scales; Procedure A with platen displacement against Procedure B with an extensometer and the gauge-length table that goes with it; platen requirements and why a spherical seat is locked for this test; the formulas with a checked worked example; the precision data; the common errors; a numeric comparison with ASTM D1621 and EN 826; and the platens, extensometers and machines Universal Grip supplies for the test.

In this guide: What ISO 844 Measures · Editions: 2007, 2014, 2021, 2026 · Specimens: the 100 × 100 × 50 mm Prism and the Alternatives · Test Speed: 10 % of Thickness per Minute · Procedure A and Procedure B: Where the Strain Comes From · Platens, Alignment and Capacity · Step-by-Step Procedure · Formulas and Worked Example · Precision · Common Errors and How Much They Cost · ISO 844 vs ASTM D1621 vs EN 826 · Equipment for ISO 844 · FAQ

What ISO 844 Measures

ISO 844 applies to rigid cellular plastics, the closed- and open-cell foams whose cell walls are stiff enough to carry load: expanded and extruded polystyrene, polyurethane and polyisocyanurate boards, phenolic foam, PVC and PET structural cores, and rigid moulded foams. Flexible foams go to ISO 3386 and the tensile test for flexible foams is ISO 1798. The standard reports:

• Compressive strength σm: the maximum force divided by the original cross-section, in MPa, when the maximum occurs before 10 % relative deformation. A foam that shows a clear yield followed by cell-wall buckling and a falling or flat curve has a compressive strength.

• Compressive stress at 10 % relative deformation σ10: the force at 10 % deformation over the original cross-section, reported when there is no maximum before 10 %. This is the value on most insulation and core data sheets.

• Compressive modulus E: the slope of the steepest linear part of the stress-strain curve below the proportional limit, after toe correction.

The 2026 edition retitled the standard to 'compressive properties' and formalised two vocabularies: Procedure A, where deformation comes from platen displacement, gives the nominal strain, the stress at 10 % nominal strain and a nominal modulus; Procedure B, where an extensometer on the specimen measures strain directly, gives the conventional strain and modulus. The strength and the 10 % stress are what buyers compare; the modulus is what structural designers use for sandwich panels, and it is the value that depends most on which procedure was run.

Editions: 2007, 2014, 2021, 2026

Most pages on the web still describe the 2007 fifth edition, which had a fixed 50 mm specimen thickness, crosshead-only deformation and results in kPa. The sixth edition of 2014 introduced the extensometer procedure. The seventh edition, ISO 844:2021, set the minimum thickness at 20 mm, added the extensometer gauge-length table by thickness, added interlaboratory precision for the extensometer procedure and an annex on modulus. The eighth edition, ISO 844:2026, published in June 2026 and adopted as EN ISO 844:2026 in July, changed the title to 'Determination of compressive properties' and introduced the nominal (Procedure A) and conventional (Procedure B) terminology for strain and modulus. The normative references are ISO 291 for conditioning, ISO 1923 for dimensions, ISO 7500-1 for force and ISO 9513 for extensometers. EN 826, the CEN thermal-insulation method, is a close relative but not identical.

Specimens: the 100 × 100 × 50 mm Prism and the Alternatives

The preferred specimen is a right prism 100 ± 1 mm square and 50 ± 1 mm thick, with the base normal to the direction in which the product is compressed in service. The standard allows, but does not prefer, any square or circular base between 25 and 230 cm², so a 50 mm cube (25 cm²) is the smallest legal specimen, not the most common one. The rules that go with it:

• Thickness: minimum 20 ± 1 mm. Products with integral moulded skins that stay on in service are tested at full product thickness provided it is at least 10 mm and no more than the width or diameter.

• Parallelism: the distance between the loaded faces may vary by no more than 1 % across the specimen.

• No stacking. Specimens are never piled up to reach a thickness; stacked pieces slip and shear.

• Skins: non-integral skins and facings are removed; the presence and location of integral skins goes in the report.

• Anisotropy: foams are weaker across the rise direction than along it, often by a factor of two or more, so an anisotropic material gets an extra set of specimens cut in the second direction.

• Cutting: by band saw, hot wire or slicer without crushing or tearing the cell structure; measure the three dimensions per ISO 1923, each as the mean of three readings.

• Number: as the product specification says, otherwise at least 5.

• Comparability: results are compared only between specimens of the same geometry and thickness; a 50 mm result and a 20 mm result on the same board are different numbers.

Conditioning is at least 6 hours at one of three atmospheres: 23 ± 2 °C and 50 ± 10 % RH (ISO 291 class 2), 23 ± 5 °C and 50 (+10, −20) % RH, or 27 ± 5 °C and 65 (+10, −20) % RH, and the test is run in the same atmosphere.

Test Speed: 10 % of Thickness per Minute

The crosshead speed is set as close as possible to 10 % of the initial specimen thickness per minute. For the preferred 50 mm prism that is 5 mm/min, the figure most pages quote as if it were fixed; for a 20 mm insulation board it is 2 mm/min, for 25 mm it is 2.5 mm/min, for 30 mm it is 3 mm/min and for a 100 mm block it is 10 mm/min. The rule keeps the strain rate at 10 % per minute whatever the thickness, so that the 10 % deformation point arrives one minute into every test and the rate-sensitive cell-wall buckling is loaded the same way. ASTM D1621's 2.5 ± 0.25 mm/min per 25.4 mm of thickness and EN 826's d/10 per minute (± 25 %) are the same rule in other clothes.

The test continues until the maximum force has been passed and the 10 % point has been reached, whichever is needed for the result; for the modulus the complete force-deformation curve is recorded, at 100 Hz or better on a modern machine, so that the steepest linear section can be found afterwards. Force is measured with an ISO 7500-1 class 1 system fixed to one platen.

Procedure A and Procedure B: Where the Strain Comes From

Procedure A takes deformation from the relative movement of the platens, which must be recorded continuously to ± 5 % or ± 0.1 mm, whichever is the more accurate. It is simple and it is what most quality-control labs run, but the record includes the seating of the platen on a slightly uneven face, machine compliance and crushed surface cells, which is why its modulus is called nominal. Procedure B measures strain on the specimen with an ISO 9513 class 1 extensometer, usually a short-travel clip-on with stop studs clamped at the mid-height, over a gauge length chosen from the specimen thickness h0:

• 20 to 32 mm thick: gauge length 15 ± 1.0 mm or 12.5 ± 1.0 mm.

• over 32 to 42 mm: 20 ± 1.0 mm.

• over 42 to 52 mm: 25 ± 1.0 mm (the preferred 50 mm prism).

• over 52 to 72 mm: 35 ± 1.0 mm.

• over 72 mm: 50 ± 1.0 mm.

The gauge length is measured to ± 0.1 mm. Platens used for Procedure B must leave room for the extensometer arms on thin specimens, which is one reason the standard specifies that platens cover the whole specimen rather than a particular diameter. Strength and stress at 10 % are the same in both procedures, because they depend only on force and area; the modulus is not, and a report must say which procedure produced it. A video extensometer is a practical Procedure B device on foams because it adds no mass and no clip force to a cell wall that may fail at 100 kPa.

Zero deformation is found the same way in both procedures: extrapolate the steepest straight part of the curve back to zero force and take that intercept as the origin. If there is no straight section, or the intercept would be negative, the origin is the deformation at a compressive stress of 250 ± 10 Pa. On a 100 × 100 mm specimen 250 Pa is only 2.5 N, which is why the force channel needs a load cell sized to the foam, not to the frame.

Platens, Alignment and Capacity

The platens are two flat, parallel, polished, non-deformable plates, square or circular, large enough to cover the whole specimen, with a surface finish of Ra 0.4 µm or better and, in steel, at least 8 mm thick in the loading direction. For a 100 mm square specimen 'cover the whole specimen' means the platen must span the 141 mm diagonal; a 100 mm round platen does not, and the standard says nothing about a 100 mm diameter. The historic text, repeated in the 2021 edition's apparatus clause, is that neither plate is self-aligning: the test is run between fixed parallel platens, and a spherical seat that is free to tilt is a feature of EN 826 and ASTM D1621, not of ISO 844. In practice the solution is a lockable seat, aligned once with the platens closed and then locked for ISO 844 work, which the Universal Grip platens below provide.

Capacity is modest for insulation foams and surprisingly large for structural cores. A 100 × 100 mm specimen at the 0.1 to 0.3 MPa of PU and PIR boards needs 1 to 3 kN; EPS and XPS at 0.07 to 0.7 MPa need up to 7 kN; a 3 MPa PET or PVC structural core at the same size needs 30 kN. A 10 kN platen pair on a 5 kN or 10 kN frame covers insulation and most moulded foams; a 25 kN frame with 30 kN platens covers structural cores. At the other end, a 250 Pa zero-point rule on an EPS board means a 500 N or 1 kN load cell, or a high-resolution 5 kN cell, is needed to see the toe at all.

Step-by-Step Procedure

• Condition the specimens at least 6 h at 23 ± 2 °C / 50 ± 10 % RH (or one of the alternatives) and test in the same atmosphere.

• Measure length, width and thickness per ISO 1923, three readings each, and calculate the original cross-section A0 and thickness h0.

• Check platen parallelism and lock the seat; centre the specimen on the lower platen using the etched centring lines.

• For Procedure B, attach the extensometer at mid-height with the gauge length from the table; for Procedure A, zero the displacement channel.

• Set the crosshead speed to 10 % of h0 per minute (5 mm/min for 50 mm) and the data rate to 100 Hz or better.

• Bring the upper platen to a light contact, start the test and record force and deformation continuously.

• Run past the maximum force, or to beyond 10 % relative deformation if no maximum appears; stop before the densification region unless the full curve is wanted.

• Find the zero-deformation origin by extrapolating the steepest linear section (or at 250 Pa), then read Fm and the deformation at Fm, or F10, and the slope for the modulus.

• Calculate σm or σ10 and E for each of the 5 specimens, report the mean to three significant figures and the individual values if they spread more than 10 %.

Formulas and Worked Example

With F in newtons and A0 in mm², stresses come out in MPa:

• Compressive strength: σm = Fm / A0, reported with the relative deformation at Fm, when Fm occurs before 10 % deformation.

• Compressive stress at 10 %: σ10 = F10 / A0, reported when there is no maximum before 10 %.

• Relative deformation: ε = Δh / h0 × 100, from platen displacement (nominal) or extensometer (conventional), measured from the corrected origin.

• Compressive modulus: E = Δσ / Δε over the steepest linear section, with ε as a ratio, not a percentage.

Worked example on a PIR insulation board, 100.2 × 99.8 × 50.1 mm, Procedure A, 5 mm/min. A0 = 100.2 × 99.8 = 10,000 mm². The curve shows no maximum before 10 %, and the force at 10 % relative deformation (5.01 mm from the corrected origin) is 3,200 N, so σ10 = 3,200 / 10,000 = 0.320 MPa, i.e. 320 kPa. For the modulus the steepest straight section runs from 400 N at 0.25 mm to 1,600 N at 0.85 mm: Δσ = 1,200 / 10,000 = 0.120 MPa and Δε = 0.60 / 50.1 = 0.01198, so E = 0.120 / 0.01198 = 10.0 MPa. Five specimens at 312, 320, 325, 318 and 329 kPa give a mean of 321 kPa, reported as 0.321 MPa, with a spread of 5 % so the individual values need not be listed.

A structural PVC foam core would give the other result type: 100 × 100 × 50 mm, maximum force 24,500 N at 6.2 % deformation followed by a drop as the cells buckle, so σm = 24,500 / 10,000 = 2.45 MPa at εm = 6.2 %, and the stress at 10 % is not reported because the strength came first.

Precision

The interlaboratory programme behind the standard, ten laboratories testing foams of 95 to 230 kPa strength and 2,500 to 8,500 kPa modulus, gives a repeatability limit of 2 % and a reproducibility limit of 9 % for strength or stress at 10 %, and 8 % and 25 % respectively for the modulus. The modulus figures are the reason Procedure B exists: platen displacement includes seating and compliance that differ from machine to machine. The 2021 edition added separate precision data for the extensometer procedure. Means are reported to three significant figures, with individual values whenever the spread exceeds 10 %.

Common Errors and How Much They Cost

• Running 5 mm/min on every thickness. A 20 mm board at 5 mm/min is tested at 25 %/min, two and a half times the standard rate; rate-sensitive foams read 5 to 10 % stronger. Set 10 % of h0 per minute.

• Spherical seat left free. A free seat lets the upper platen follow a non-parallel face, which is what EN 826 wants and ISO 844 does not; lock it for ISO 844 and record that you did.

• Platen too small. A platen that does not cover the diagonal loads the corners of a square specimen unevenly and crushes the overhang; the result reads low and the failure is at the edges.

• Skipping the toe correction. Reading 10 % from the raw crosshead zero on a specimen with a crushed surface skin shifts the 10 % point by the seating distance, often 0.3 to 0.5 mm, which on a 50 mm prism is 0.6 to 1 % strain and a few percent of stress on a rising curve.

• Stacking thin boards. Two 10 mm boards stacked to make 20 mm slip and shear; the standard forbids it, and the 20 mm minimum (10 mm with integral skins) is the way out.

• Load cell too large. A 100 kN frame cell cannot resolve the 250 Pa (2.5 N) origin rule on a 100 × 100 mm EPS specimen; use a 500 N to 5 kN cell for insulation foams.

• Reporting a modulus without the procedure. Nominal (A) and conventional (B) moduli on the same foam can differ by 20 % or more; the 2026 edition names them separately for this reason.

• Wrong direction. Testing an insulation board flat when the product is loaded on edge, or ignoring rise direction in a block foam, changes the result by a factor of two on anisotropic materials.

ISO 844 vs ASTM D1621 vs EN 826

The three foam compression standards share the 10 % per minute strain rate and the 10 % stress concept but differ in specimen limits, end point and hardware, and the results are not interchangeable:

• Specimen base: ISO 844 preferred 100 × 100 mm, allowed 25 to 230 cm²; ASTM D1621 25.8 to 232 cm² (4 to 36 in²); EN 826 50 to 300 mm square, normally the full product thickness.

• Thickness: ISO 844 minimum 20 mm (preferred 50); D1621 minimum 25.4 mm up to the width or diameter, so 50.8 mm maximum on a 2 in square; EN 826 product thickness.

• Speed: ISO 844 10 % of h0 per minute; D1621 2.5 ± 0.25 mm/min per 25.4 mm of thickness (0.1 in/min per inch); EN 826 d/10 per minute ± 25 %.

• End point: ISO 844 and EN 826 at the maximum or 10 % deformation; D1621 to yield or 13 % strain, reporting the stress at 10 % if there is no yield.

• Platens: ISO 844 fixed parallel, Ra 0.4 µm, steel ≥ 8 mm; D1621 Ø 2 to 8 in with a spherical seat permitted; EN 826 spherical-seated upper platen.

• Deformation: ISO 844 Procedure A ± 5 % or ± 0.1 mm, Procedure B ISO 9513 class 1 with the gauge-length table; D1621 ASTM E83 class C or better.

• Specimens: ISO 844 at least 5; D1621 at least 5; EN 826 per product standard.

• Units and reporting: ISO 844 MPa to three significant figures with the procedure named; D1621 kPa or psi.

For the compression of flexible foams see ISO 3386 and ASTM D3574; for rigid plastics that are not cellular, ISO 604 and ASTM D695; for the broader compression testing programme, platens, cages and fixtures are covered on the compression fixtures page.

Equipment for ISO 844

Universal Grip ISO 844 compression platens

The Universal Grip ISO 844 platens (part UGC-ISO844-CP) are a pair of round compression platens for rigid foam and cellular plastics. The data sheet describes large ground platen faces that cover the full specimen cross-section; ground, flat, parallel contact faces in hardened steel with a corrosion-resistant finish; a bolted spherical seat on the upper platen that can be locked for fixed-platen testing, which is how it is used for ISO 844, or left free for EN 826 and ASTM D1621; a pin-and-socket shank for quick installation; and adapters to suit the load frame, clevis or threaded. Capacity is 10 kN standard with higher capacities on request. When ordering, specify the specimen cross-section, the platen diameter (at least the specimen diagonal, 142 mm or more for the preferred 100 mm square), the required capacity and the machine connection. The general compression platens range covers larger diameters and capacities for structural cores and sandwich panels.

Extensometers, machines and chambers

• Procedure B strain: a short-travel clip-on from the extensometer range with stop studs for the 12.5 to 50 mm gauge lengths, or a video extensometer that reads the gauge marks without touching a fragile cell wall and covers every gauge length in the table.

• Machines: the 5 kN and 10 kN benchtop machines for insulation and moulded foams, the 25 kN for structural cores; fit a load cell sized to the foam so that the 250 Pa origin and a 0.1 MPa board are resolved.

• Fixtures: the compression fixtures page covers guided platens, cages and the ISO 604 plastics setup that shares a frame with foam work.

• Chambers: environmental chambers for compressive properties at service temperature, with the platens on extension rods.

• Related guides: ISO 1798 flexible foam tensile, ASTM D1621 and the foam testing page; the full ISO standards library.

ISO 844 FAQ

What is ISO 844?

ISO 844 is the international standard for the compressive properties of rigid cellular plastics. A square or round prism is compressed between parallel platens at 10 % of its thickness per minute, and the compressive strength or the compressive stress at 10 % relative deformation, and optionally the compressive modulus, are reported. The current edition is ISO 844:2026, the eighth, published in June 2026.

What specimen size does ISO 844 use?

The preferred specimen is a right prism 100 ± 1 × 100 ± 1 mm with a thickness of 50 ± 1 mm. Any square or circular base from 25 to 230 cm² is allowed, with a minimum thickness of 20 mm (10 mm for products with integral skins), parallel faces within 1 %, and no stacking.

What is the ISO 844 test speed?

10 % of the initial thickness per minute: 5 mm/min for the 50 mm prism, 2 mm/min for a 20 mm board, 2.5 mm/min for 25 mm and 10 mm/min for a 100 mm block. The strain rate is therefore 10 % per minute for every thickness.

What is compressive stress at 10 % deformation?

The force at 10 % relative deformation divided by the original cross-section, in MPa. It is reported when the foam shows no maximum force before 10 %; if a maximum occurs earlier, the compressive strength σm = Fm/A0 and the deformation at that point are reported instead. A 100 × 100 mm board carrying 3,200 N at 10 % has σ10 = 0.32 MPa.

What is the difference between Procedure A and Procedure B?

Procedure A measures deformation from platen displacement and gives nominal strain and modulus; Procedure B measures strain on the specimen with an ISO 9513 class 1 extensometer over a gauge length chosen by thickness (25 mm for a 50 mm specimen) and gives the conventional strain and modulus. Strength and 10 % stress are the same either way; the modulus is not.

How many specimens does ISO 844 need?

At least 5 unless the product specification says otherwise, plus a second set in the other direction for anisotropic foams. Means are reported to three significant figures, with individual values when the spread exceeds 10 %.

Is a spherical seat allowed in ISO 844?

The test is run between fixed parallel platens; a free self-aligning platen is a feature of EN 826 and ASTM D1621. A lockable spherical seat is the practical answer: align the platens, lock the seat for ISO 844 and release it for the other standards.

What platen size is needed for ISO 844?

Platens must cover the whole specimen: at least the 141 mm diagonal, so 142 mm or larger, for the preferred 100 mm square; Ra 0.4 µm or better; steel at least 8 mm thick. A 100 mm diameter platen does not cover a 100 mm square.

What machine capacity is needed for ISO 844?

Cross-section times expected stress: 100 × 100 mm insulation boards at 0.1 to 0.3 MPa need 1 to 3 kN, EPS/XPS up to 7 kN, structural PET or PVC cores at 3 MPa need 30 kN. Use a load cell small enough to resolve the 250 Pa (2.5 N) zero-point rule on the same specimen.

How is the modulus found in ISO 844?

E = Δσ/Δε over the steepest linear section of the stress-strain curve, after moving the origin to the point where that section extrapolates to zero force (or to the deformation at 250 ± 10 Pa if there is no straight section). On a 50 mm PIR specimen a slope of 1,200 N over 0.60 mm on 10,000 mm² gives E = 10.0 MPa.

What is the conditioning for ISO 844?

At least 6 hours at 23 ± 2 °C and 50 ± 10 % RH (ISO 291 class 2), or at 23 ± 5 °C / 50 (+10, −20) % RH, or 27 ± 5 °C / 65 (+10, −20) % RH, with the test run in the same atmosphere.

How does ISO 844 compare with ASTM D1621?

Same 10 % per minute strain rate (D1621 writes it as 2.5 mm/min per 25.4 mm of thickness) and the same 25.8 to 232 cm² area range, but D1621 runs to yield or 13 % strain, allows a spherical seat, requires at least 25.4 mm thickness up to the specimen width, and uses ASTM E83 class C deformation measurement; ISO 844 stops at 10 %, uses fixed platens and a minimum 20 mm thickness. Results are not interchangeable.

Need compression platens, an extensometer or a complete compression system for ISO 844? Contact Universal Grip with your specimen size, expected strength and machine connection and we will quote the right setup.