ISO 4578
Floating-Roller Peel Resistance of Adhesive Bonds

Downloads: ISO 4578 Floating-Roller Peel Fixture for Structural Adhesives – Universal Grip Data Sheet (PDF)

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

ISO 4578 is the international floating-roller peel test for high-strength adhesive bonds. A thin, flexible adherend (usually 0.5 mm metal sheet) is bonded to a stiffer rigid adherend, then peeled off around a free-turning 25 mm roller at a constant crosshead speed. The result is the average force per unit width needed to keep the peel front moving, reported in kN/m, together with the maximum and minimum forces and the failure pattern.

The method is used by structural adhesive formulators, aerospace and automotive bonding engineers, and quality labs that need a repeatable number for toughened epoxies, acrylics and other structural adhesives on metal. This guide covers the current edition, specimen requirements, apparatus, procedure, the calculation with a worked example, common errors, and how ISO 4578 compares with ASTM D3167. Universal Grip builds a dedicated floating-roller peel fixture for this standard (part ref. UGC-ISO4578-PF).

In this guide: What ISO 4578 Measures · Current Edition, Scope and History · Test Specimens · Apparatus: Machine, Load Cell, Fixture and Grip · ISO 4578 Test Procedure Step by Step · Calculation of Peel Resistance · Worked Example · Interpreting Results and Failure Modes · ISO 4578 vs ASTM D3167 and Related Standards · Test Report Requirements · ISO 4578 Equipment from Universal Grip · FAQ

What ISO 4578 Measures

ISO 4578 measures peel resistance: the force needed to separate a flexible adherend from a rigid adherend along a bond line, divided by the bonded width. Unlike a lap-shear test such as ISO 4587, which loads the whole overlap in shear, a peel test concentrates stress in a narrow zone at the crack front. That makes it very sensitive to adhesive toughness, surface preparation and cure, which is exactly why structural adhesive specifications pair a shear test with a peel test.

In the floating-roller geometry the rigid adherend is held in a fixture of two free-running rollers that hangs from the load cell. The unbonded end of the flexible adherend is bent at a right angle and clamped in the opposite grip. As the crosshead separates, the flexible adherend is pulled away and bends around a roller, so the bending radius at the peel front is set by the roller rather than by the stiffness of the adherend. The rigid panel slides through the fixture as the peel advances, so the peel front stays at the roller throughout the test. This is what the word floating refers to.

The standard's scope notes that the floating roller may give more consistent numerical data than other peel methods, but also that the flexible adherend should not be expected to conform exactly to the roller surface. The number you get is therefore a comparative peel resistance under defined conditions, not an intrinsic material property such as fracture energy.

• Primary result: average peeling force per unit width, kN/m (numerically equal to N/mm)

• Also reported: maximum and minimum peeling force for each specimen, kN/m

• Also reported: failure type, designated per ISO 10365

• Typical users: adhesive qualification, process control of surface treatments, aerospace and automotive structural bonding

Current Edition, Scope and History

The current edition is ISO 4578:1997, the third edition, published by ISO/TC 61/SC 11 (Plastics, Products). It cancelled and replaced the second edition, ISO 4578:1990, which was technically revised; the first edition dates from 1979. ISO reviewed and confirmed the 1997 edition in 2023, so it remains the version to cite. Check the ISO catalogue before quoting the standard in a contract, because a confirmation can be followed by a revision.

The scope covers bonds between one rigid adherend and one flexible adherend, prepared and tested under the conditions the standard specifies. It is written around metal-to-metal bonds, but other flexible adherends are allowed if they are made thicker to limit deformation of the rigid adherend. It is aimed at structural adhesives; pressure-sensitive tapes and flexible-to-flexible laminates are normally tested with 90°, 180° or T-peel methods instead (see the related standards section below).

Note: Several of the standard's normative references have since been replaced. Surface preparation was referenced to ISO 4588 (metals) and ISO 13895 (plastics); both were replaced by ISO 17212, now in its 2012 edition. ISO 10365 (failure patterns) is now in its 2022 edition. Use the current documents unless your specification fixes the older ones.

Test Specimens

Specimens may be made individually or cut from larger bonded panels. Either way, each consists of a flexible and a rigid adherend, prepared and bonded according to the adhesive manufacturer's recommendations. When cutting from panels, use a method that does not damage the bond; milling and band sawing are the usual choices. The unbonded end of the flexible adherend is bent perpendicular to the rigid adherend so it can be clamped in the grip.

Dimensions

• Flexible adherend thickness (metal): 0.5 mm ± 0.02 mm

• Flexible adherend thickness (other materials): thicker than 0.5 mm, to reduce deformation of the rigid adherend

• Rigid adherend thickness: 1.6 mm ± 0.1 mm

• Width: 25 mm ± 0.5 mm (preferred), or any other convenient width the fixture accepts, stated in the report

• Lengths: per Figure 2 of the standard; the bond line must be long enough for the peel length requirements described below

Because the standard asks for force over at least 115 mm of bond line after discarding the first 25 mm of peel, and the average excludes the last 20 mm as well, a practical design target is about 160 mm of peelable bond line plus the unbonded gripping tail. That satisfies every reading of the length clauses and leaves margin for a slightly late start.

Number of specimens and bond line thickness

• Number: as given in the material specification, otherwise at least five

• Edge specimens cut from the edges of a panel: identify them and report their results separately

• Bond line thickness: measure the total thickness of the bonded pair and subtract the thicknesses of the two adherends, for every specimen

• Ageing, if required: choose conditions from ISO 9142

Conditioning

Condition and test in a standard laboratory atmosphere from ISO 291 (normally 23 °C and 50 % relative humidity) unless the adhesive user and manufacturer agree on something else. Record the conditioning time and atmosphere in the report.

Apparatus: Machine, Load Cell, Fixture and Grip

Testing machine

Use a constant-rate-of-traverse universal testing machine that can record force against crosshead movement. ISO 4578 sets three practical requirements that many lab write-ups skip:

• The peel forces measured must fall between 15 % and 85 % of the load cell's capacity

• Force must be indicated to within 1 %

• Calibration: ISO 7500-1 for metal adherends, ISO 5893 for rubber or plastic adherends

• Recording: autographic (continuous) force versus crosshead movement; an inertialess electronic system is recommended

The 15 % to 85 % window matters. A toughened epoxy peeling at around 160 N on a 25 mm strip sits at 16 % of a 1 kN load cell but only 3.2 % of a 5 kN cell, so a frame's main high-capacity cell is often the wrong choice. Many labs use a 500 N or 1 kN cell for this test, and a single-column benchtop frame is usually sufficient.

Floating-roller fixture

The fixture specified in the standard is mild steel with dual roller bearings and 25 mm diameter rollers that must roll freely. The standard stresses that the roller geometry and bearings are critical and must be maintained carefully: a sticky or worn bearing adds friction that is recorded as peel force. The fixture should be free to align itself with the line of pull.

Grip for the flexible adherend

The flexible adherend is held in a self-aligning grip whose jaws firmly hold the outer 25 mm of the strip. Serrated wedge grips or screw-action grips sized for thin sheet are typical. The grip must not let the 0.5 mm strip slip, and it should not tear the strip at the jaw edge.

ISO 4578 Test Procedure Step by Step

• 1. Measure and record the width of each specimen and the bond line thickness (total thickness minus both adherend thicknesses).

• 2. Fit the floating-roller fixture to the load cell and the grip to the opposite crosshead; zero the force with the fixture hanging free.

• 3. Insert the rigid adherend into the fixture so that the bonded flexible adherend wraps around the roller on the peel side, as shown in Figure 1 of the standard.

• 4. Clamp the outer 25 mm of the bent, unbonded flexible adherend in the grip, keeping the strip straight and centered on the pull line.

• 5. Remove slack without preloading the bond.

• 6. Start the test at 100 mm/min ± 5 mm/min unless the material specification gives another speed.

• 7. Record force against crosshead movement over at least 115 mm of bond line, disregarding the first 25 mm of peel.

• 8. Watch the rigid adherend. If it bends or distorts, the result is not valid; the standard says to redesign the rigid adherend with more stiffness.

• 9. Reject any specimen that fails outside the peeling zone defined by the fixture.

• 10. After the test, examine both fracture surfaces and record the failure type per ISO 10365.

Note: For a flexible adherend that does not stretch, crosshead travel is approximately equal to peel distance in this geometry, so the force versus crosshead movement trace can be read directly as force versus peel length.

Calculation of Peel Resistance

From the force trace, take the portion covering at least 115 mm of peeling, disregarding the first 25 mm and the last 20 mm. Determine the average force over that region and divide by the specimen width:

Peel resistance (kN/m) = average peeling force (N) ÷ width (mm)

Because 1 N/mm equals 1 kN/m, the arithmetic needs no conversion factor. The standard allows several ways to find the average: planimeter, the gravimetric method (cutting out and weighing the chart area), a best-fit straight line drawn by eye, or other methods such as computer analysis. Today almost every lab uses software that averages the sampled data over the evaluation window. Also record the maximum and minimum forces in the window for each specimen, divided by width.

Report the method used to determine the average, because a peak-picking average, a true integral average and a best-fit line can give slightly different values on a stick-slip trace.

Worked Example

A toughened epoxy bonds 0.5 mm aluminum to 1.6 mm aluminum. Specimen 1 measures 25.0 mm wide. The software records 160 mm of peel; the evaluation window runs from 25 mm to 140 mm of peel, which is 115 mm long after excluding the first 25 mm and last 20 mm.

• Average force in window: 160 N

• Maximum force in window: 182 N

• Minimum force in window: 141 N

• Average peel resistance: 160 ÷ 25.0 = 6.40 kN/m

• Maximum: 182 ÷ 25.0 = 7.28 kN/m

• Minimum: 141 ÷ 25.0 = 5.64 kN/m

The five specimens give average peel resistances of 6.40, 6.10, 6.70, 6.30 and 6.50 kN/m. The sum is 32.00, so the mean is 32.00 ÷ 5 = 6.40 kN/m. Report each specimen's average, maximum and minimum, the failure type (for example, cohesive failure in the adhesive), and the averaging method. The 160 N average sits at 16 % of a 1 kN load cell, inside the 15 % to 85 % window.

Note: These numbers are illustrative only. Peel resistance depends heavily on adhesive, adherend alloy, surface treatment, bond line thickness and cure; do not use them as design values.

Interpreting Results and Failure Modes

A peel number means little without the failure pattern. ISO 4578 requires the failure type to be designated per ISO 10365, which distinguishes adhesion failure at the adhesive–adherend interface (AF), cohesion failure within the adhesive (CF), and other patterns such as mixed, near-interface and adherend failure. In structural bonding, a high peel value with cohesive failure usually indicates a sound surface treatment; a low value with clean interfacial failure usually points to contamination, a poor or aged surface treatment, or under-cure.

Reading the force trace

• Smooth plateau: steady crack growth; the average is meaningful and max/min are close

• Saw-tooth stick-slip: common with brittle or highly filled adhesives; report max and min as well as the average

• Rising or falling trend: often bond line thickness variation, uneven cure along the panel, or a fixture that is not rolling freely

• Sudden drop to near zero: an unbonded area or void; inspect the fracture surface

Common errors

• Load cell too large, so the forces fall below 15 % of capacity

• Rollers not turning freely, adding friction to the recorded peel force

• Rigid adherend bending during the test, which changes the peel geometry and invalidates the result

• Averaging from the start of the trace instead of discarding the first 25 mm and last 20 mm

• Not measuring bond line thickness, which strongly affects peel results for toughened adhesives

• Mixing edge specimens with center specimens instead of reporting them separately

• Flexible adherend kinked or work-hardened at the bend, causing fracture of the strip instead of peel

ISO 4578 vs ASTM D3167 and Related Standards

The closest ASTM equivalent is ASTM D3167, Standard Test Method for Floating Roller Peel Resistance of Adhesives (current designation D3167-10, reapproved 2025). The principle is the same, but the details differ, so data are not directly interchangeable:

• Crosshead speed: ISO 4578 uses 100 mm/min; ASTM D3167 uses 152 mm/min (6 in/min)

• Default flexible adherend: 0.5 mm in ISO 4578; 0.63 mm (0.025 in) in ASTM D3167

• Rigid adherend: 1.6 mm ± 0.1 mm in ISO 4578; 1.63 mm (0.064 in) in ASTM D3167

• Rollers: 25 mm in ISO 4578 fixtures; 25.4 mm (1 in) in ASTM D3167 fixtures

• Units: kN/m in ISO 4578; N/mm or lbf/in in ASTM practice

• Averaging window: each standard sets its own discard lengths; follow the one you are reporting against

In Europe, EN 1464 (floating-roller peel of adhesive bonds) and the aerospace method EN 2243-2 use the same fixture concept. Related adhesive peel tests on this site include the 90° peel ISO 8510-1, the T-peel tests ISO 11339 and ASTM D1876, and the 180° peel ASTM D903. Choose floating roller when one adherend is genuinely rigid and the adhesive is structural; choose T-peel when both adherends are flexible.

For rubber bonded to metal rather than adhesive between two metals, the relevant peel method is ISO 813, which peels a 6 mm rubber strip at 90° from a rigid plate.

Test Report Requirements

ISO 4578 lists the contents of the report. In summary it must include:

• A reference to ISO 4578

• Full identification of the adhesive: type, source, manufacturer's code, batch or lot, and form

• Full identification of the adherends: material, thickness, width and surface preparation

• The bonding process: application method, drying or pre-cure, cure time, temperature and pressure

• The average adhesive layer thickness after bonding

• Specimen description: individually made or cut from panels, dimensions, construction, number of panels and specimens, with edge specimens identified

• Conditioning procedure and test conditions

• Crosshead speed, if not 100 mm/min

• The method used to determine the average peeling force

• Average, maximum and minimum peeling force for each specimen in kN/m, with edge specimens reported separately

• Failure type per ISO 10365

The standard states that no precision data are available because no interlaboratory study had been run, so set acceptance limits from your own historical data rather than from a published repeatability figure.

ISO 4578 Equipment from Universal Grip

Universal Grip's ISO 4578 Peel Fixture (part ref. UGC-ISO4578-PF) is a floating-roller fixture for measuring the peel resistance of high-strength adhesive bonds between a flexible and a rigid adherend. The rigid panel is supported by free-turning rollers in a pivoting frame while the flexible strip is pulled away by the lower grip, holding a constant peel angle around the roller.

• Free-turning rollers on pinned shafts for low-friction peeling

• Ground rollers on bearing shafts, retained by circlips

• Pivoting side plates let the roller assembly self-align with the peel

• Rigid square spigot with cross-pin for direct load-cell mounting, to suit the customer frame

• Compact construction in hardened, corrosion-protected steel for metal-to-metal bonds

• Load capacity: 5 kN typical; higher capacities on request

• Compatible frames: electromechanical universal test machines

The fixture is used with a wedge or screw grip on the opposite crosshead; see our wedge grips and tensile grips. When ordering, specify specimen width and adherend thicknesses, required capacity, the lower grip and your test machine connection. For a complete system, pair the fixture with a universal testing machine and software that averages force over a defined peel window. For more on materials and methods, see our adhesives testing and peel testing pages.

To configure a fixture for your specimens and frame, contact us for a quote.

ISO 4578 FAQ

What is ISO 4578?

ISO 4578 is the ISO test method for determining the peel resistance of high-strength adhesive bonds by the floating-roller method. A flexible adherend, typically 0.5 mm metal, is peeled from a 1.6 mm rigid adherend around a 25 mm roller at 100 mm/min, and the average force per unit width is reported in kN/m.

What is the current edition of ISO 4578?

The current edition is ISO 4578:1997, the third edition, which replaced ISO 4578:1990. ISO last reviewed and confirmed it in 2023, so it remains valid.

What test speed does ISO 4578 use?

The crosshead speed is 100 mm/min ± 5 mm/min unless the material specification states otherwise. Any other speed must be reported.

What size are ISO 4578 specimens?

The preferred width is 25 mm ± 0.5 mm. A metal flexible adherend is 0.5 mm ± 0.02 mm thick and the rigid adherend is 1.6 mm ± 0.1 mm thick; other widths are allowed if the fixture accepts them and the width is reported.

How many specimens are required?

Test the number given in the material specification, or at least five if none is given. Specimens cut from panel edges are identified as edge specimens and reported separately.

How is peel resistance calculated in ISO 4578?

Average the force over at least 115 mm of peeling, disregarding the first 25 mm and the last 20 mm, then divide by specimen width. The result is in kN/m, which is numerically the same as N/mm; maximum and minimum forces are also reported.

What load cell should I use for ISO 4578?

Choose a load cell so the measured peel forces fall between 15 % and 85 % of its capacity, with force accuracy within 1 %. For 25 mm wide metal specimens a 500 N or 1 kN cell is often appropriate, rather than a frame's main high-capacity cell.

What is the difference between ISO 4578 and ASTM D3167?

Both use a floating-roller fixture, but ISO 4578 runs at 100 mm/min with a 0.5 mm flexible adherend, while ASTM D3167 runs at 152 mm/min (6 in/min) with a 0.63 mm (0.025 in) default flexible adherend. The two also differ in roller size and averaging conventions, so results should not be mixed.

Why is it called a floating-roller test?

The rigid adherend is held between free-turning rollers and slides through the fixture as the peel advances. This keeps the peel front at the roller, so the bend radius of the flexible adherend stays constant throughout the test.

Is ISO 4578 a 90° peel test?

Not in the usual sense. The flexible adherend is bent around a 25 mm roller, so the peel geometry is set by the roller rather than by a fixed 90° angle; for a true 90° peel of a flexible adherend from a rigid one, see ISO 8510-1.

What does Universal Grip offer for ISO 4578?

Universal Grip makes the ISO 4578 Peel Fixture (part ref. UGC-ISO4578-PF), a floating-roller fixture with free-turning ground rollers, pivoting side plates and a square spigot with cross-pin, rated 5 kN typical. Contact us with your specimen width, adherend thicknesses and machine connection for a quote.

Need a fixture, grips or a complete test system for ISO 4578? Contact Universal Grip with your material, specimen size and machine connection and we will quote the right setup.