ISO 813
Rubber-to-Rigid Substrate Adhesion, 90° Peel Method

ISO 813 90° Peel Fixture for Rubber-to-Metal Adhesion by Universal Grip

Downloads: ISO 813 90° Peel Fixture for Rubber-to-Metal Adhesion – Universal Grip Data Sheet (PDF)

Written by Chris Thomas, Universal Grip Company · Reviewed by Frank Bertini · Updated October 27, 2026

ISO 813 is the international test for the adhesion of vulcanized or thermoplastic rubber to a rigid substrate, usually metal. A 25 mm wide, 6 mm thick rubber strip bonded to a 25 mm square area on a rigid plate is pulled off at 90° at 50 mm/min. The adhesion strength is the maximum force divided by the strip width, in N/mm, reported together with a coded description of where the bond failed.

Rubber compounders, bonding agent suppliers and makers of anti-vibration mounts, bushings, rollers and rubber linings use ISO 813 to select compounds and adhesive systems and to control bonding processes. This guide covers the current 2019 edition, test piece preparation, apparatus, the procedure, the calculation with a worked example, failure codes, common errors and the comparison with ASTM D429 Method B. Universal Grip makes a dedicated 90° peel fixture for this standard (part ref. UGC-ISO813-PF).

In this guide: What ISO 813 Measures · Current Edition and Scope · Test Pieces: Dimensions and Preparation · Apparatus · ISO 813 Test Procedure Step by Step · Calculation of Adhesion Strength · Worked Example · Failure Codes and How to Interpret Them · Common Errors and Practical Tips · ISO 813 vs ASTM D429 and Related Standards · Test Report Requirements · ISO 813 Equipment from Universal Grip · FAQ

What ISO 813 Measures

ISO 813 measures the force needed to separate a rubber strip from a rigid substrate with the peel held at, or as close as possible to, 90° to the bond plane. Because the bonded area is a small 25 mm × 25 mm square and the rubber is a thick 6 mm strip, the test is short and severe: the peel front crosses the whole bond in one pull, and the highest force reached during that pull is the result.

The number is only half of the answer. Every ISO 813 result is paired with a failure code that says whether the rubber tore (the bond was stronger than the rubber) or one of the interfaces in the bonding system let go. In rubber-to-metal bonding, a high value with failure in the rubber is the goal; a high value with interfacial failure can still signal a process problem.

Result: adhesion strength in N/mm = maximum force ÷ test piece width
Failure type: coded R, RC, CP, PS, CS, D or S, with the percentage of each
Test pieces: four per test
Typical uses: comparing rubber compounds and bonding agents, developing materials, and controlling manufacturing

Current Edition and Scope

The current edition is ISO 813:2019, the sixth edition, published in July 2019 by ISO/TC 45/SC 2 (Rubber and rubber products, Testing and analysis). It replaced ISO 813:2016 as a minor revision that corrected a typographical error in Table B.1 of Annex B. ISO reviewed and confirmed the 2019 edition in 2025, so it remains the version to cite.

The 2016 fifth edition was itself a minor revision that updated the normative references; the 2010 fourth edition replaced the 1997 third edition. The technical method has been stable across these editions. One small detail did change: the 1997 text told the operator to strip about 1.5 mm of rubber with a knife before loading, while the current text says about 2 mm.

Scope and limits

Covers vulcanized and thermoplastic rubber bonded to a single plate of rigid material, peeled at 90°
Intended mainly for laboratory-prepared test pieces made under standard conditions
Used to select rubber compounds or adhesive systems, develop materials and control manufacturing processes
Not suitable for high-hardness rubbers, typically above 85 IRHD, which cannot bend through 90° in a controlled way
Normative references in the current edition: ISO 5893 (test machines), ISO 18899 (calibration) and ISO 23529 (preparation and conditioning of test pieces)

Check hardness before testing hard compounds; see our hardness testing page if you need to confirm IRHD or Shore values.

Test Pieces: Dimensions and Preparation

Dimensions

Rubber strip thickness in the test area: 6 mm ± 0.1 mm
Rubber strip width: 25 mm ± 0.1 mm
Rubber strip length: 125 mm
Bonded area: 25 mm long × 25 mm ± 0.1 mm wide, approximately centered on the substrate
Substrate width: 25 mm ± 0.1 mm
Substrate length: 60 mm ± 1 mm
Substrate thickness: enough to prevent deformation during the test; a minimum of 1.5 mm is recommended for metal, and plastics may need more

The rubber is longer than the substrate on purpose. The overhanging tail gives the grip something to hold, while only the 25 mm square in the middle of the plate is bonded.

Moulding and bonding

Moulds: either a multi-cavity mould for several test pieces or a single-piece mould; the inside length parallel to the substrate is 125 mm, and the depth is set (for example with spacer plates) to leave 6 mm ± 0.05 mm for the rubber
Substrate surfaces: clean, dry and free of dust; prepared as the adhesive system requires
Masking: both ends of the substrate are masked with pressure-sensitive tape so that only the 25 mm bonded area is exposed
Rubber: cleaned as required, cut to fill the mould with enough volume to develop full bonding pressure
Assembly: substrate at the bottom of the mould; in multi-piece moulds the substrate strips are spaced about 3 mm apart
Cure: vulcanize or bond under pressure for the time and temperature the system requires; remove carefully and do not stress the bonds while hot
Separation: cut multi-piece slabs apart; edges may be buffed flush with the substrate without overheating or going below the width tolerance

Post-vulcanization bonding, where cured rubber is bonded to the plate afterwards, is also permitted. In that case the bonding temperature, pressure and surface preparation must be reported.

Number and conditioning

Number of test pieces: four
Minimum time between moulding and testing: 16 h
Conditioning immediately before test: at least 16 h at 23 °C ± 2 °C or 27 °C ± 2 °C
Other test temperatures: condition long enough to reach equilibrium, per ISO 23529

Apparatus

Testing machine and load cell

Use a universal testing machine that meets ISO 5893, measures force to class 1 accuracy and runs the moving grip at 50 mm/min ± 5 mm/min. Calibrate per ISO 18899; Annex B of the standard gives the calibration schedule. Pick a load cell so the expected maximum force falls well inside its calibrated range, and use an electronic data system fast enough to catch the true peak, because the result is a single maximum value.

Fixture

The fixture holds the substrate to one crosshead so that the pull on the rubber stays as nearly perpendicular to the bond plane as possible. Figure 1 of the standard shows an example design. The substrate is held flat, with the bonded face toward the grip, and the rubber tail hangs at right angles to it.

Grip

The grip holds the free end of the rubber strip without letting it slip and without causing it to break at the jaw. A screw clamp with a knurled or serrated face is common for 6 mm rubber.

Temperature chamber

For testing above or below laboratory temperature, an environmental chamber that holds the chosen temperature within the tolerances of ISO 23529 is used, with the fixture and grip inside.

ISO 813 Test Procedure Step by Step

1. Measure the width of each test piece in the bonded area and record it.
2. Mount the fixture and grip; zero the force with the fixture installed.
3. Place the test piece symmetrically in the fixture, with the end to be pulled toward the operator.
4. With a sharp knife, strip the rubber from the substrate for about 2 mm at the starting edge of the bond, taking care not to tear the rubber.
5. Clamp the free end of the rubber strip in the grip, aligned with the pull.
6. For non-standard temperatures, keep the test piece in the chamber until it reaches the test temperature.
7. Separate at 50 mm/min ± 5 mm/min until separation is complete.
8. Record the maximum force. A continuous force record over the full length is optional but recommended.
9. If the rubber tends to tear instead of peeling, cut it back to the substrate with the knife and continue.
10. Inspect the fracture surfaces and estimate the percentage of each failure type.

Other separation speeds may be used to examine the viscoelastic behavior of the bond, as described in informative Annex A. In that case, speeds stepped in logarithmic increments are recommended.

Calculation of Adhesion Strength

For each test piece:

Adhesion strength (N/mm) = maximum force (N) ÷ test piece width (mm)

The divisor is the width of the strip, not the bonded area; this is a force-per-width peel value, not a stress. The value used is the maximum force, not an average of a plateau. That is a key difference from adhesive peel standards such as ISO 4578 or ISO 8510-1, where an average over a peel length is reported, and it is a point several published summaries of ISO 813 get wrong.

To convert to US units, 1 N/mm equals 5.710 lbf/in.

Worked Example

A natural rubber compound is bonded to grit-blasted steel with a primer and cover cement. Four test pieces are tested at 23 °C. Each strip measures 25.0 mm wide.

Test piece 1: 312 N ÷ 25.0 mm = 12.48 N/mm, failure R 80 %, RC 20 %
Test piece 2: 298 N ÷ 25.0 mm = 11.92 N/mm, failure R 60 %, RC 40 %
Test piece 3: 325 N ÷ 25.0 mm = 13.00 N/mm, failure R 100 %
Test piece 4: 305 N ÷ 25.0 mm = 12.20 N/mm, failure R 90 %, RC 10 %

The four values sum to 49.60 N/mm, so the mean is 49.60 ÷ 4 = 12.40 N/mm (about 70.8 lbf/in). The median, the average of the two middle values 12.20 and 12.48, is 12.34 N/mm. Report all four individual values with their failure codes, and state which summary statistic you use if your specification calls for one.

In this example the bond is mostly stronger than the rubber, with some rubber-to-cover-cement failure. If a later batch gave a similar force but 100 % RC failure, the bonding process would need attention even though the number looks acceptable.

Note: These values are illustrative only. Adhesion strength depends on the compound, the bonding system, surface preparation, cure and test temperature.

Failure Codes and How to Interpret Them

ISO 813 designates the location of failure with these symbols, and the report gives the percentage of each:

R: failure in the rubber
RC: failure at the rubber–adhesive interface
CP: failure at the adhesive–primer interface (when a primer is used)
PS: failure at the primer–substrate interface (when a primer is used)
CS: failure at the adhesive–substrate interface (when no primer is used)
D: failure at the rubber–substrate interface in direct adhesion, with no adhesive
S: failure in the substrate

What the codes usually point to

R: the bonding system is stronger than the rubber; the force is then limited by the rubber's tear strength and stiffness
RC: the cover cement did not bond fully to the rubber, often from contamination, compound bloom, scorch before bonding or cure mismatch
CP: incompatibility between cover cement and primer, or poor intercoat timing
PS or CS: substrate preparation problems such as residual oil, oxide or inadequate grit blasting
D: in direct-bonding compounds, an issue with the compound's adhesion chemistry or cure
S: substrate too weak or too thin for the bond strength

Because rubber tearing caps the force, a better bonding system cannot raise the value beyond the rubber's own strength. When comparing bonding agents, compare failure codes as well as forces.

Common Errors and Practical Tips

Averaging the force trace instead of reporting the maximum force
Dividing by bonded area (625 mm²) instead of width, which gives a stress in N/mm² that is not an ISO 813 result
Starting with the bonded square off-center under the pull line, which makes the angle drift further from 90° as the peel front crosses the 25 mm bond
Skipping the 2 mm knife start, so the first peak reflects crack initiation through the fillet rather than steady peel
Letting the rubber tear and continue tearing instead of cutting it back to the substrate
Testing too soon after moulding (less than 16 h) or without 16 h conditioning
Substrates thinner than about 1.5 mm that bend during the pull
Masking tape not applied, so adhesive bonds beyond the 25 mm area and inflates the result

Centering matters more than it looks. If the peel front sits 12.5 mm to one side of the pull line with 100 mm of free strip, the pull is about 7.1° off 90° (the arctangent of 12.5/100); if the piece is loaded with the front 25 mm off line, the error grows to about 14.0°. Placing the test piece symmetrically, as the standard requires, keeps the front within about 12.5 mm either side of the line through the whole test.

ISO 813 vs ASTM D429 and Related Standards

The closest ASTM method is ASTM D429, Method B (90° stripping test, rubber part assembled to one metal plate). The geometry is nearly the same, and the two are often run on the same fixture:

Rubber thickness: 6 mm ± 0.1 mm in ISO 813; 6.3 mm (0.250 in) in ASTM D429 Method B
Rubber width and length: 25 mm × 125 mm in both
Metal plate: 60 mm × 25 mm in both; ASTM specifies 1.6 mm thick, ISO 813 recommends at least 1.5 mm
Bonded area: 25 mm × 25 mm in both
Speed: 50 mm/min in both (2 in/min in ASTM)
Result: maximum force per width in both, N/mm or lbf/in
Failure codes: ISO 813 uses R, RC, CP, PS, CS, D and S; ASTM D429 Method B uses R, RC, CP and M (metal–prime cement)
Number of pieces: four in ISO 813

The 0.3 mm difference in rubber thickness changes the bending stiffness of the strip slightly, so specify one standard and do not mix data sets without checking correlation.

Related standards

ISO 814: rubber-to-metal adhesion by the two-plate tension method, for bonds loaded normal to the interface
ASTM D413: rubber adhesion to flexible substrates such as fabric
ISO 4578: floating-roller peel of structural adhesive bonds between metals
ISO 23529 and ISO 18899: test piece conditioning and machine calibration referenced by ISO 813

Test Report Requirements

The report should allow anyone to repeat the test. It includes:

A reference to ISO 813
Identification of the sample, the rubber, the substrate and the bonding system
A description of the test piece, its components and the bonding method, including bonding conditions for post-vulcanization bonding
Preparation dates and the conditions the test pieces were kept in before testing
Test temperature and test speed
Adhesion strength for each test piece in N/mm
Failure types with the percentage of each
Test date and any unusual observations

Precision data for interlaboratory reproducibility have not been widely established for this method, so set acceptance limits from your own historical results and keep test temperature, speed and preparation identical across a comparison series.

ISO 813 Equipment from Universal Grip

Universal Grip's ISO 813 Peel Fixture (part ref. UGC-ISO813-PF) is a fixture set for measuring the adhesion of rubber bonded to a rigid plate at 90°. The upper holder clamps the metal substrate horizontally on a pin-jointed rod, while the lower clamp grips the rubber strip, so the strip is pulled at right angles to the bond line as the crosshead moves.

Plate holder clamps the rigid substrate horizontally for a true 90° peel
Pin-jointed hanger rod lets the holder self-align to the pull line
Lower clamp with knurled screw grips the rubber strip end
Set screws locate substrate plates of different thicknesses
Material: corrosion-protected steel and aluminum
Load capacity: 1 kN standard; higher capacities on request
Machine connection: pin connections, supplied with clevis or threaded adapters to suit your load frame
Compatible frames: electromechanical universal test machines

When ordering, specify the substrate plate width, length and thickness, the rubber strip width and your test machine connection. For more on testing rubber and peel methods, see our elastomer testing and peel testing pages. To configure a fixture for your plates and frame, contact us for a quote.

ISO 813 FAQ

What is ISO 813?

ISO 813 is the ISO method for determining the adhesion of vulcanized or thermoplastic rubber to a rigid substrate by a 90° peel. A 6 mm thick, 25 mm wide rubber strip bonded to a 25 mm square on a rigid plate is pulled off at 50 mm/min, and the maximum force per unit width is reported in N/mm.

What is the current edition of ISO 813?

The current edition is ISO 813:2019, the sixth edition. It replaced ISO 813:2016 with a minor correction to Annex B, and ISO confirmed it in 2025.

What are the ISO 813 test piece dimensions?

The rubber strip is 6 mm ± 0.1 mm thick, 25 mm ± 0.1 mm wide and 125 mm long. It is bonded over a 25 mm × 25 mm area centered on a substrate 60 mm ± 1 mm long and 25 mm ± 0.1 mm wide, with at least 1.5 mm substrate thickness recommended.

What test speed does ISO 813 use?

The moving grip travels at 50 mm/min ± 5 mm/min until the rubber is fully separated. Other speeds may be used to study the viscoelastic behavior of the bond, preferably in logarithmic steps.

How is ISO 813 adhesion strength calculated?

Divide the maximum force in newtons by the test piece width in millimeters to get N/mm. For example, 312 N on a 25.0 mm strip gives 12.48 N/mm.

Does ISO 813 use the average or the maximum peel force?

ISO 813 uses the maximum force recorded during separation. Averaging a plateau, as in many adhesive peel standards, does not give an ISO 813 result.

How many test pieces does ISO 813 require?

Four test pieces are tested. They must be tested at least 16 h after moulding and conditioned for at least 16 h at 23 °C ± 2 °C or 27 °C ± 2 °C immediately before testing.

What do the ISO 813 failure codes R, RC, CP and CS mean?

R is failure in the rubber, RC at the rubber–adhesive interface, CP at the adhesive–primer interface, and CS at the adhesive–substrate interface when no primer is used. The standard also uses PS (primer–substrate), D (direct rubber–substrate adhesion) and S (failure in the substrate), and the report gives the percentage of each.

Can ISO 813 be used for hard rubber?

No. The scope excludes high-hardness rubbers, typically above 85 IRHD, because they cannot be peeled through 90° in a controlled way.

What is the difference between ISO 813 and ASTM D429 Method B?

Both peel a 25 mm wide strip from a 25 mm × 25 mm bond on a 60 mm plate at 50 mm/min and report maximum force per width. ISO 813 uses 6 mm rubber and seven failure codes, while ASTM D429 Method B uses 6.3 mm (0.250 in) rubber and the codes R, RC, CP and M.

What does Universal Grip offer for ISO 813?

Universal Grip makes the ISO 813 Peel Fixture (part ref. UGC-ISO813-PF), with a horizontal plate holder on a pin-jointed hanger rod, a knurled-screw rubber clamp and a 1 kN standard capacity. Contact us with your plate size, strip width and machine connection for a quote.

Need a fixture, grips or a complete test system for ISO 813? Contact Universal Grip with your material, specimen size and machine connection and we will quote the right setup.

Literature

Pricing and Availability

ISO 813 Rubber-to-Rigid Substrate Adhesion, 90° Peel Method