top of page

ISO 8067
Tear Strength of Flexible Cellular Polymeric Materials

Universal Grip ISO 8067 foam tear test specimen cutting die

Downloads: ISO 8067 Foam Tear Test Specimen Cutting Die – Universal Grip Data Sheet (PDF)

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

ISO 8067 is the international standard for the tear strength of flexible cellular polymeric materials: polyurethane, latex and other soft foams used in seating, bedding, automotive trim, packaging and medical products. A shaped test piece is pulled apart at a constant rate on a tensile testing machine, the maximum tearing force is recorded, and the tear strength is the force divided by the thickness of the piece. It is the foam industry's companion to the tensile test of ISO 1798 and the rubber tear test of ISO 34-1, and the ISO counterpart of ASTM D3574 Test F.

This guide covers the current edition, ISO 8067:2026, published on 20 October 2026 and replacing the 2018 edition that every equipment page still quotes; the two test pieces (Method A trouser and Method B angle) and when to use each; the 50 ± 5 mm/min referee rate and the 50 to 500 mm/min range; the three-piece minimum and the 20 % spread rule; the median result and the N/cm, N/m and N/mm units that trip up lab reports; conditioning and the 72-hour rule after manufacture; a worked calculation; the common errors, including why trouser results scatter three times more than angle results; a side-by-side with ASTM D3574 Test F; and the cutting dies, grips and machines Universal Grip supplies for the test.

In this guide: What ISO 8067 Measures and Why Foam Needs Its Own Tear Test · Editions: 1989, 2008, 2018 and the New ISO 8067:2026 · Test Pieces: Method A Trouser and Method B Angle · Test Speed, Number of Pieces and the Machine · Conditioning and the 72-Hour Rule · Step-by-Step Procedure · Calculation, Units and a Worked Example · Precision: Why Trouser Results Scatter Three Times More Than Angle Results · Common Errors and How Much They Cost · ISO 8067 vs ASTM D3574 Test F vs ISO 34-1 · Grips for Foam: Low Pressure, Wide Faces, No Serrations · Equipment for ISO 8067 from Universal Grip · FAQ

What ISO 8067 Measures and Why Foam Needs Its Own Tear Test

Tear strength is the resistance of a material to the propagation of a cut or tear. For a foam it is one of the three properties that decide whether a cushion survives sewing, stapling, handling and years of use: tensile strength and elongation (ISO 1798), tear strength (ISO 8067) and compression properties (ISO 3386 and ISO 2439). A foam with good tensile strength but poor tear strength splits along a staple line or at a sewn seam, which is why automotive seating specifications from GM, Ford, BMW, Daimler, PSA and Volkswagen call up ISO 8067 alongside ISO 1798.

The method cannot simply borrow the rubber tear test. Foam is 90 to 98 % air, its cells collapse under a serrated jaw, it stretches a long way before the tear runs, and the force is small, typically single-digit to tens of newtons. ISO 8067 therefore uses thick test pieces (nominally 25 mm), a slow referee speed, the maximum tearing force rather than an average, and the median of the pieces rather than the mean, which keeps one badly cut piece from dominating the result. The result is normalised by thickness only, not by width or cross-section, because the tear runs through the thickness.

  • Method A, trouser test piece: a rectangular block with a cut along its length that divides one end into two legs; the legs are pulled apart in the grips and the tear runs along the remaining uncut length. The maximum force over a marked 25 mm of tear is recorded.

  • Method B, angle test piece: a die-cut piece with a 90° re-entrant angle and no nick; the tear starts at the apex of the angle under tension and runs across the piece. The maximum force is recorded.

The two methods do not give the same number for the same foam and results from one must never be compared with the other; the specification, not the lab, chooses the method.

Editions: 1989, 2008, 2018 and the New ISO 8067:2026

The first edition of 1989 was three pages long and had a single method for foam thicker than 24 mm. The 2008 second edition introduced the two methods A and B that are still used. The 2018 third edition added test pieces with skin to the specimen clause, so that foams tested with a moulded skin face could be reported as such. The fourth edition, ISO 8067:2026, was published on 20 October 2026 and withdraws the 2018 edition. It is a minor revision: the one listed technical change is in the test report clause, which now requires the location and number of surfaces with skin to be recorded for every test piece.

In practice this means a result on a skinned piece can no longer be reported as a bare 'tear strength' figure; a report written to ISO 8067:2026 must say which faces carried skin. Dies, speeds, pieces and the calculation are unchanged, so a lab running the 2018 method only has to update its report template and its certificate reference. Specifications that cite 'ISO 8067:2018' remain technically satisfied by a 2026 test, but the certificate should cite the edition actually used. The CEN adoption is EN ISO 8067 and the British version BS EN ISO 8067; GB/T and JIS adoptions follow the ISO text.

Test Pieces: Method A Trouser and Method B Angle

ISO 8067 defines the two test piece geometries in its figures. The dimensions below are the ones that govern the result; the full drawing with tolerances is in the standard and should be the basis of any die.

  • Method A trouser piece: a rectangular block, in practice about 150 mm long by 25 mm wide by 25 mm thick, with a cut made along the centreline from one end for 45 to 55 mm so that two 'legs' are formed. Two lines are marked on the uncut part 25 mm apart; the tear is run through this marked length and the maximum force within it is the result. The cut must be central, square to the faces and made with a sharp blade in one stroke; a wandering cut biases the tear toward the thinner leg.

  • Method B angle piece: a flat piece die-cut from foam sheet 10 to 15 mm thick with a 90° angle in one edge; there is no nick at the apex. The piece is held by its two tab ends and the tear initiates at the apex when the local stress exceeds the foam's tear resistance. Because there is no razor cut, the result includes the tear-initiation energy, which is why angle values are higher and less scattered than trouser values.

  • Thickness is measured at the tear line to ISO 1923 with a contact gauge at low pressure, because a foam gauge foot that sinks into the surface reads thin and inflates the result. The thickness used in the calculation is the actual measured value, not the nominal die depth.

  • Skin: pieces may be cut with a moulded or cast skin on one or more surfaces where the product is tested as made. ISO 8067:2026 requires the report to state the location and number of skinned surfaces, because a skin can double the apparent tear strength of a low-density foam.

  • Orientation: foams rise in one direction and their cells are elongated along it. The direction of the tear relative to the rise direction must be stated; specifications usually call for pieces cut parallel and perpendicular to rise.

Cutting matters more here than in almost any other test. A hollow die driven through 25 mm of foam in one stroke gives square, uncrushed faces; a band saw or hot wire gives a wavy face that starts the tear early or late. A die must be tall enough to clear the full thickness without the base compressing the foam before the edge cuts, and it needs an ejector so the piece is released without being stretched. The Universal Grip ISO 8067 cutting die below is built for exactly this.

Test Speed, Number of Pieces and the Machine

  • Rate of jaw separation: the standard permits 50 to 500 mm/min. The referee rate for the trouser piece (Method A) is 50 ± 5 mm/min; for the angle piece the faster end of the range, 500 ± 50 mm/min, is the usual referee practice quoted by equipment makers and the rate most automotive specifications call for. Use the rate in the material specification, state it in the report, and never compare results taken at different rates: tear force in polyurethane foam rises with rate, typically by several percent per doubling.

  • Number of test pieces: at least three per sample and direction. If the spread of results exceeds about 20 % of the median, or the tear stops and restarts so that the force trace is not a clean tearing record, test additional pieces.

  • Machine: a constant-rate-of-extension tensile machine with a force system verified to ISO 7500-1 class 2 or better; because tearing forces are often below 50 N, a class 1 system on a 100 N to 1 kN load cell is the sensible choice so that a 10 N result sits well above 1 % of range. Extension measurement is not required; the result is a force.

  • Data rate: the maximum force is a peak on a saw-tooth trace, so sample at 50 Hz or faster; a slow channel clips the peaks and under-reads the tear strength.

  • Grips: must hold the 25 mm legs without crushing them and without letting them pull out; see the grip section below. The gap between the jaw faces must open to at least 25 to 30 mm to accept a full-thickness leg.

Conditioning and the 72-Hour Rule

Test pieces are conditioned for at least 16 hours in one of the two standard atmospheres, 23 ± 2 °C and 50 ± 5 % relative humidity or, for tropical laboratories, 27 ± 2 °C and 65 ± 5 % RH, and tested in the same atmosphere. Foam properties continue to develop after manufacture as the polymer cures and blowing-agent residues leave, so ISO 8067 follows ISO 1798 in requiring that tests are not carried out less than 72 hours after manufacture. A shorter interval is allowed only where it has been shown that results after 16 hours and after 48 hours agree within ± 10 %; for routine production control labs often qualify a 16-hour interval on that basis and record it.

Pieces are cut after conditioning, or cut and then conditioned for the full 16 hours, never cut from a conditioned block and tested at once from a different room. For low-density foams the thickness measurement is taken after conditioning as well, because the gauge reading changes with the foam's recovery state.

Step-by-Step Procedure

  • 1. Choose the method from the material specification (A trouser or B angle), the direction relative to foam rise, and whether skin faces are to be left on. Record all three.

  • 2. Cut at least three pieces with the appropriate die in a press, with the die edge sharp and the stroke completed in one movement. For Method A, make the 45 to 55 mm centre cut with a fresh razor blade guided by a jig, then mark the 25 mm tear length on the uncut section.

  • 3. Condition for at least 16 hours at 23 °C / 50 % RH (or 27 °C / 65 %), and check the piece is at least 72 hours old unless a shorter interval has been qualified.

  • 4. Measure thickness at the tear line to ISO 1923 with a low-pressure contact gauge, to 0.1 mm.

  • 5. Mount. Method A: one leg in each grip, each leg gripped over its full width with about 15 to 20 mm inside the jaws and the cut centred between the grips so the uncut section hangs square. Method B: the two tab ends in the grips with the angle apex centred on the load axis. Close the jaws at the lowest pressure that prevents pull-out.

  • 6. Run at the specified rate (50 ± 5 mm/min for the trouser referee test, 500 ± 50 mm/min where the specification calls for it) until the tear has passed through the marked 25 mm (A) or the piece has torn through (B). Record the force trace.

  • 7. Read the maximum force within the marked tear length (A) or the maximum force of the test (B).

  • 8. Reject any piece where the tear left the centreline and ran into a leg, where a leg pulled out of or tore at the jaw, or where the cut was visibly off-centre; replace it.

  • 9. Calculate the tear strength of each piece as maximum force divided by thickness, take the median, and if the spread exceeds about 20 % test further pieces.

  • 10. Report per the standard's clause 10: material, method, rise direction, rate, conditioning, thickness, individual and median results, and, for ISO 8067:2026, the location and number of skinned surfaces on each piece.

Calculation, Units and a Worked Example

Tear strength T = Fmax ÷ d, where Fmax is the maximum tearing force in newtons and d the thickness of the test piece in the direction of the tear, in millimetres (or centimetres or metres, which is where the unit confusion starts). The standard's unit is force per unit thickness; ISO-style reports and most European equipment pages give it in N/cm, ASTM D3574 and many North American reports give it in N/m, and software often defaults to N/mm. All three describe the same quantity: 1 N/mm = 10 N/cm = 1,000 N/m, and 1 N/cm = 0.571 lbf/in.

  • Worked example (Method A, seating polyurethane, 28 kg/m³): three trouser pieces, thicknesses 25.1, 24.8 and 25.3 mm, maximum forces within the marked 25 mm of 9.8, 11.2 and 10.1 N. Tear strengths: 9.8 ÷ 25.1 = 0.390 N/mm; 11.2 ÷ 24.8 = 0.452 N/mm; 10.1 ÷ 25.3 = 0.399 N/mm. The median is 0.399 N/mm = 3.99 N/cm = 399 N/m. The spread is (0.452 − 0.390) ÷ 0.399 = 15.5 %, under 20 %, so three pieces suffice.

  • Same foam, Method B angle piece, 12.0 mm thick: maximum force 18.6 N → 18.6 ÷ 12.0 = 1.55 N/mm = 15.5 N/cm = 1,550 N/m. Roughly four times the trouser value for the same material, which is normal: the angle piece includes the energy to initiate the tear, the trouser piece only propagates one. The two numbers are not comparable and should never appear in the same table without the method named.

  • Why the median: with results 0.390, 0.399 and 0.452 N/mm the mean is 0.414 and the median 0.399; one high piece (a slightly off-centre cut that made the tear climb) moves the mean 4 % but the median not at all. With five pieces the median is the third value in rank order.

  • Load-cell check: 11.2 N on a 1 kN cell is 1.1 % of range, acceptable for a class 1 system but marginal; a 100 N or 250 N cell puts the reading at 4 to 11 % of range and tightens the force accuracy to a fraction of a percent.

Precision: Why Trouser Results Scatter Three Times More Than Angle Results

ISO 8067 includes an informative precision annex comparing the two test pieces, and the inter-laboratory figures quoted from it by equipment makers are stark: reproducibility of up to 36 % for the trouser piece against about 12 % for the angle piece, a factor of three. The trouser piece is sensitive to everything about the cut: its depth, its straightness, whether the blade crushed cells on the way through, and whether the legs were gripped evenly so the tear stayed on the centreline. The angle piece has no cut; the tear starts where the geometry says it must.

That does not make Method B 'better'. The trouser piece measures pure propagation, which is what a seam or staple line experiences once a tear has started, and many legacy specifications are written around it. It does mean that a laboratory running Method A has to control the cut like a machining operation: a jig for the razor, a fresh blade every few pieces, and the die pressed in a single stroke. Where a specification allows either method and the purpose is production control, Method B gives tighter limits for the same number of pieces.

Common Errors and How Much They Cost

  • Quoting ISO 8067:2018 on the certificate. The 2018 edition was withdrawn on 20 October 2026. Tests from that date should cite ISO 8067:2026 and include the skin-surface statement in the report.

  • Dividing by width instead of thickness. At least one major equipment page says the result is 'in relation to the width'; it is not. For a 25 × 25 mm trouser piece the mistake is invisible; for a 12 mm thick angle piece cut 25 mm wide it halves the result.

  • Mixing units. 4.0 N/cm reported as '4.0 N/m' is a factor of 100 error and a routine cause of failed supplier audits. Report N/cm or N/m and state which; convert with 1 N/cm = 100 N/m.

  • Serrated jaws on foam. Serrations cut the legs; the piece then tears at the jaw and the result is low and invalid. Use flat or rubber-faced jaws at the lowest pressure that prevents slip (see grips below).

  • Reading the mean of the peaks. ASTM-style averaging over a travel window is not the ISO method; ISO 8067 takes the maximum force within the marked length. Averaging reads 10 to 20 % low on a typical saw-tooth trace.

  • Testing young foam. A polyether foam tested at 24 hours can read 5 to 10 % different from the same block at 72 hours. Hold to the 72-hour rule or qualify the shorter interval with the ± 10 % check.

  • Off-centre or shallow cut. A cut that is 5 mm off the centreline makes the tear climb into the thick leg and the force rise steadily; the 'maximum' is then a geometric artefact. A cut shorter than 45 mm leaves too little leg to grip without the jaws touching the tear front.

  • Thickness from the nominal die depth. Foam recovers unevenly after die cutting; a 25 mm die can yield pieces measuring 24 to 26 mm. Measure each piece.

  • Comparing trouser with angle or ISO with ASTM. The pieces differ in geometry, cut and speed; ISO 8067 itself says the two methods are not comparable, and ASTM D3574 Test F uses a different slit and a different formula convention.

ISO 8067 vs ASTM D3574 Test F vs ISO 34-1

The three tear methods a foam or rubber lab meets most often, with the numbers that differ. ASTM D3574 Test F figures are from the ASTM method as quoted on our D3574 foam page; ISO 34-1 figures are from the ISO 34-1 rubber tear guide.

  • Test piece: ISO 8067 A trouser block, in practice about 150 × 25 × 25 mm with a 45 to 55 mm cut and 25 mm marked tear length; ISO 8067 B angle piece without nick, 10 to 15 mm thick. D3574 Test F: block 152 × 25 × 25 mm (6 × 1 × 1 in) with a 40 mm (1.5 in) slit; the angle piece is not permitted. ISO 34-1 (rubber): trouser 100 × 15 mm with 40 mm cut, angle and crescent pieces, 2.0 mm thick sheet.

  • Rate: ISO 8067 50 to 500 mm/min, referee 50 ± 5 (trouser), 500 ± 50 common for angle; D3574 Test F 50 mm/min; ISO 34-1 100 mm/min trouser, 500 mm/min angle and crescent.

  • Pieces: ISO 8067 ≥ 3 (more if spread > 20 %); D3574 3; ISO 34-1 ≥ 5 per direction.

  • Result: ISO 8067 maximum force ÷ thickness, median, N/cm (or N/m); D3574 F/T × 10³ in N/m with F in N and T in mm, reported as the average; ISO 34-1 F/d in kN/m, median, with the trouser force taken as the median of the peaks per ISO 6133.

  • Conditioning: ISO 8067 23/50 or 27/65 for ≥ 16 h, ≥ 72 h after manufacture; D3574 23 ± 2 °C / 50 ± 5 % for ≥ 12 h, ≥ 12 h after manufacture; ISO 34-1 ISO 23529, ≥ 16 h after vulcanisation.

  • Comparability: none. A seating foam reading 4 N/cm (400 N/m) on ISO 8067 trouser may read 450 N/m on D3574 Test F because of the shorter slit and faster effective tear, and 15 N/cm on the ISO angle piece. Specify the method, and if a customer sends ASTM data for an ISO specification, retest.

Grips for Foam: Low Pressure, Wide Faces, No Serrations

Foam is the one material where the usual tensile-grip instinct, more clamping force, is wrong. A serrated jaw at line pressure cuts through the cell walls and the leg tears at the jaw; too little pressure and the leg slides out as the tear force rises. The standard says only that the grips must hold without slipping and without initiating failure at the jaw; the practice that satisfies both is:

  • Pneumatic grips at 1 to 2 bar with flat or rubber-faced jaws 25 to 50 mm wide, so the whole leg is clamped over its full width and the pressure is spread. Pneumatic grips keep the same pressure on every piece, which is the main reason their results repeat better than hand-tightened grips on soft materials. Our pneumatic grips are supplied with flat, rubber and serrated face sets; use the first two for foam.

  • Vise grips with flat faces for low-volume work, tightened to the same number of turns each time; a torque-limited handle or a feeler gap keeps the pressure repeatable.

  • Jaw opening of at least 25 to 30 mm to accept a 25 mm leg without pre-compressing it; most 1 kN grips open to 30 mm or more.

  • Alignment: the two legs must be gripped so that the cut lies on the load axis; a jig that locates the uncut section against a stop while the jaws close keeps the tear on the centreline and is the single most effective way to bring trouser scatter down.

Our jaw-break and slippage guide covers the pressure and face choices in more depth; the same rules apply to the ISO 1798 foam tensile test.

Equipment for ISO 8067 from Universal Grip

  • ISO 8067 foam tear test specimen cutting die (part UGC-ISO8067-SC): a hollow cutting die with ejector plate machined to the ISO 8067 test piece geometry. The tall die body clears thick foam sections in a single stroke, typically up to 25 mm; the hardened, ground and honed tool-steel cutting edge produces clean, square cut faces; the bolted ejector plate releases the cut test piece without distortion; and the heavy base block sits square under a press platen. Suitable for manual, pneumatic or hydraulic specimen presses; other foam test piece profiles (angle pieces, ISO 1798 dumbbells) on request. Specify the test piece type and dimensions, foam density and thickness, and the press when ordering.

  • Die press and arbor press: a manual arbor press cuts 25 mm seating foam cleanly with a sharp die; dense viscoelastic and latex foams and tall dies are better served by the pneumatic or hydraulic presses on the same page. The video below shows the arbor press cutting.

  • Pneumatic grips and vise grips with flat and rubber faces and 30 mm openings, 500 N to 5 kN, with adapters for any frame.

  • Machines: a 1 kN to 2.5 kN single-column frame is more than enough; the Mark-10 testers with a 100 N or 250 N gauge and our 2.5 kN benchtop UTM with a 100 N or 500 N cell are the usual choices, the UTM where the lab also runs ISO 1798 tensile and ISO 3386 compression on the same frame.

  • Environmental chambers for tear strength at the service temperatures automotive specifications add to the room-temperature test.

  • Related guides: ISO 1798 foam tensile strength, ISO 34-1 rubber tear, ASTM D624 rubber tear, our foam testing page, and the ISO standards library.

ISO 8067 FAQ

What is the current edition of ISO 8067?

ISO 8067:2026, the fourth edition, published on 20 October 2026. It replaces ISO 8067:2018. The one listed technical change is that the test report must now state the location and number of surfaces with skin on each test piece. Test pieces, speeds and the calculation are unchanged.

What are Method A and Method B in ISO 8067?

Method A uses a trouser test piece: a block with a 45 to 55 mm centre cut whose two legs are pulled apart so the tear runs along a marked 25 mm length. Method B uses an angle test piece with a 90° re-entrant angle and no nick, die-cut from 10 to 15 mm sheet, in which the tear starts at the apex. The two give different numbers and are not comparable.

What test speed does ISO 8067 use?

The permitted range is 50 to 500 mm/min. The referee rate for the trouser piece is 50 ± 5 mm/min; the angle piece is commonly run at 500 ± 50 mm/min where the specification calls for it. Use the rate in the material specification and state it in the report.

How is ISO 8067 tear strength calculated?

Tear strength equals the maximum tearing force divided by the thickness of the test piece at the tear line. For a trouser piece 25.1 mm thick with a maximum force of 9.8 N the result is 9.8 ÷ 25.1 = 0.390 N/mm, which is 3.90 N/cm or 390 N/m. The median of at least three pieces is reported.

What unit is ISO 8067 tear strength reported in?

Force per unit thickness. European reports usually give N/cm and ASTM-style reports N/m; both are correct if stated. 1 N/mm = 10 N/cm = 1,000 N/m, and 1 N/cm = 0.571 lbf/in. Reporting 4 N/cm as '4 N/m' is a factor-of-100 error.

How many test pieces does ISO 8067 require?

At least three per sample and direction. If the spread of the individual results exceeds about 20 % of the median, or the tear trace is irregular, additional pieces are tested. The result is the median, not the mean.

What are the conditioning requirements for ISO 8067?

At least 16 hours at 23 ± 2 °C and 50 ± 5 % RH (or 27 ± 2 °C and 65 ± 5 % RH), and testing not less than 72 hours after manufacture unless results after 16 and 48 hours have been shown to agree within ± 10 %.

Why do trouser results scatter more than angle results?

The informative precision annex gives reproducibility of up to about 36 % for the trouser piece against about 12 % for the angle piece. The trouser result depends on the depth, straightness and quality of the razor cut and on even gripping of the legs; the angle piece has no cut, so the tear starts where the geometry dictates.

How does ISO 8067 compare with ASTM D3574 Test F?

D3574 Test F uses a 152 × 25 × 25 mm block with a 40 mm slit at 50 mm/min, three specimens, and reports F/T × 10³ in N/m as an average; it does not permit an angle piece. ISO 8067 allows trouser or angle pieces, a 45 to 55 mm cut, 50 to 500 mm/min, and reports the median. Results are not interchangeable.

What grips should be used for foam tear tests?

Flat or rubber-faced jaws at low pressure, 1 to 2 bar on pneumatic grips or lightly tightened vise grips, with the whole 25 mm leg clamped and a jaw opening of at least 25 to 30 mm. Serrated faces cut the foam and cause jaw tears. A locating jig that keeps the cut on the load axis reduces scatter more than any other single measure.

What load cell and machine capacity does ISO 8067 need?

Tear forces are usually 5 to 50 N for seating foams and rarely exceed 200 N for dense foams, so a 100 N to 500 N load cell on a 1 to 2.5 kN single-column machine is appropriate. A 1 kN cell reading 10 N is at 1 % of range, marginal for class 1 accuracy; a 100 N cell reads the same test at 10 % of range.

Can test pieces with skin be used?

Yes. Skinned pieces have been allowed since the 2018 edition, and ISO 8067:2026 requires the report to state the location and number of skinned surfaces, because a moulded skin can double the apparent tear strength of a low-density foam. Results with and without skin must not be mixed.

Need an ISO 8067 cutting die, low-pressure grips or a complete foam tear testing system? Contact Universal Grip with your test piece type, foam thickness and press or machine details and we will quote the right setup.

bottom of page