top of page

ASTM D903
180 Degree Peel Strength Test

180° Peel Fixture for ASTM D903
Universal Grip ASTM D903 vise grip for universal testing machines

Downloads: Standard Test Method for Peel or Stripping Strength of Adhesive Bonds - PDF · ASTM D903 Vise Grip – Universal Grip Data Sheet (PDF)

Written by Chris Thomas, Universal Grip Company · Reviewed by Frank Bertini · Updated October 17, 2026

ASTM D903 is the standard test method for the peel or stripping strength of adhesive bonds. A flexible strip is bonded to a second member, folded back on itself and pulled at 180° while the machine records the load needed to strip it away at a fixed separation rate. The result, reported as load per unit width, is the most common number used to compare adhesives, surface treatments, primers and cure schedules for flexible-to-rigid and flexible-to-flexible bonds: tapes, labels, laminated films, coated fabrics, rubber-to-metal and foam bonds all get tested this way.

This guide covers the full specimen geometry (both members, not just the flexible strip), why the crosshead runs at 305 mm/min to produce a 152 mm/min separation rate, the 7-day conditioning requirement, how to grip and align the specimen, how to draw the best average load line and convert it to peel strength, the failure modes you will see, and how D903 differs from the T-peel, tape and 90° peel methods that are often confused with it. It also lists the grips, peel fixtures and machines Universal Grip supplies for the test. One caution up front: D903 has no precision and bias statement, so the only way to know your scatter is to run the full set of specimens the standard asks for.

In this guide: What ASTM D903 Measures · Specimen Dimensions and Materials · Surface Preparation and Bonding · Conditioning: 7 Days, Not 24 Hours · Test Speed: Why the Crosshead Runs at 305 mm/min · Machine Requirements and Load Cell Selection · Grips, Peel Fixtures and Alignment · Step-by-Step Procedure · Calculating Peel Strength and Reading the Curve · Failure Modes and What They Tell You · ASTM D903 vs D1876, D3330, D6862, D1781 and ISO 8510 · Common Errors · Equipment for ASTM D903 · FAQ

What ASTM D903 Measures

D903 measures the comparative peel or stripping characteristics of an adhesive bond on specimens of standard size, prepared, conditioned and tested under defined conditions. The standard defines peel or stripping strength as the average load per unit width of bond line required to separate progressively one member from the other at an angle of approximately 180° and a separation rate of 152 mm (6 in) per minute.

Two words in that definition drive everything else. Progressively means the bond is unzipped along its length, not broken all at once, so the load trace is a long plateau rather than a single peak, and the result is an average, not a maximum. Comparative means the number ranks adhesives and processes against each other; it is not a design allowable, because peel load depends on the stiffness and thickness of the flexible member and on the exact angle at the peel front as much as on the adhesive itself.

The only document D903 references is ASTM D907, the terminology standard for adhesives. Everything about surface preparation, cure and bonding is left to the adhesive manufacturer or to the parties running the test, which is why the report section asks you to record those details with every data set. Peel testing in general is covered on our peel testing applications page, and the fixtures for every peel geometry are compared in our peel test fixtures guide.

Specimen Dimensions and Materials

Most descriptions of D903 stop at a one-inch strip. The standard actually specifies two members with different lengths, and the unbonded tail of the flexible member is what gets folded back and gripped.

  • Flexible member: 25 mm (1 in) wide by 304.8 mm (12 in) long. It must be flexible enough to bend back through 180° without cracking.

  • Second member (flexible or rigid): 25 mm (1 in) wide by 203.2 mm (8 in) long.

  • Bonded length: 152.4 mm (6 in) at one end of the flexible member, with the bonded area ending at one end of the second member.

  • Unbonded portions: the free 152.4 mm of the flexible member lies face to face with the unbonded 50.8 mm of the second member until the test begins; the flexible tail is folded back for gripping.

  • Thickness: enough to withstand the tensile pull without breaking or stretching, but not more than 3 mm (1/8 in). Typical thicknesses are 1.6 mm (1/16 in) for metals and plastics, 3 mm (1/8 in) for wood, 1.9 mm (0.075 in) for rubber, and 627.4 g/m² (18.5 oz/yd²) for cotton duck.

  • Number of specimens: at least ten per adhesive. The standard says to cut five 25 mm strips from a bonded panel about 152 mm wide, so two panels give the minimum set.

  • Flaws: discard any specimen with visible voids, misalignment or damage before testing rather than after; a low result from a flawed specimen is not a valid data point.

The standard asks for members that are relatively non-extensible. If the flexible member stretches under the peel load (thin films, light fabrics, soft foams), the measured load includes the energy going into stretching the backing, and the peel front angle drifts away from 180°. Back such materials with a stiffer carrier (adhesive-coated cloth or a thin rigid film is common) and report the backing in the test report, as section 5.2 requires.

Bond width is the denominator of the result, so measure it. A nominal 25 mm strip cut by hand can be 24 to 26 mm, which is a 4 to 8 % error before the adhesive has done anything. A specimen cutting die or a strip cutter with a fixed blade spacing removes that error and gives clean edges that peel evenly; ragged edges start tears in the flexible member that show up as spikes in the load trace.

Surface Preparation and Bonding

D903 does not specify a surface preparation, an adhesive application method or a cure cycle. It says to follow the adhesive manufacturer's recommendations and to record what was done. That makes the preparation record part of the result: the same adhesive on solvent-wiped aluminum, on abraded aluminum and on anodized aluminum will give three different peel strengths, and the number is meaningless without the preparation that produced it.

  • Prepare the second member and the flexible member as the adhesive data sheet directs (degrease, abrade, prime, dry) and note each step.

  • Apply the adhesive to the 152.4 mm bond area only. Mask the free tails so they do not stick together; a strip of release film between the unbonded portions is the usual trick.

  • Control bondline thickness. Peel strength rises with bondline thickness for most structural adhesives, so use the manufacturer's spread rate or shims and note it.

  • Cure as directed (time, temperature, pressure). If pressure is applied with a roller, record the roller mass and number of passes; for pressure-sensitive products this step dominates the result.

  • Cut strips from the bonded panel after cure, not before. Cutting before bonding leaves edge beads and misaligned members.

Conditioning: 7 Days, Not 24 Hours

Section 7 of D903 requires conditioning the bonded specimens for 7 days at 23 ± 1 °C (73.4 ± 1.8 °F) and 50 ± 2 % relative humidity, or until they reach equilibrium with that atmosphere, unless the adhesive manufacturer specifies otherwise. Note the tolerances: ± 1 °C and ± 2 % RH, tighter than the ± 2 °C / ± 5 % RH used in many plastics standards. Several guides on the web quote a 24-hour conditioning period; that is not in the standard and will give different results for any adhesive that is still curing or still absorbing moisture after a day.

If you do not have a conditioning room, the standard allows a desiccator cabinet with a saturated solution of calcium nitrate, which holds about 51 % RH at 23 °C. Special conditioning (heat aging, humidity soak, water immersion, salt spray) is permitted by agreement between the parties and must be reported. Test as soon as practical after removing the specimen from conditioning, preferably in the same atmosphere, because peel strength of many adhesives changes within minutes as the surface temperature and moisture content drift. Our environmental chambers let you condition and test at the same temperature when the test specification calls for hot or cold peel.

Test Speed: Why the Crosshead Runs at 305 mm/min

D903 specifies a separation rate of 152 mm (6 in) per minute. In a 180° peel, both the flexible member and the second member are moving relative to each other, and each millimetre of crosshead travel strips only half a millimetre of bond line, because the flexible member is folded back on itself and the peel front advances at half the crosshead speed. To get 152 mm/min of bond separation, the standard therefore requires the power-actuated grip to travel at 305 mm (12 in) per minute. This is the single most common setup error in D903: labs that enter 152 mm/min as the crosshead speed are running the test at half the specified rate, and because peel strength is rate-dependent for most adhesives, their results are not comparable to anyone else's.

The rate dependence runs in both directions. Viscoelastic adhesives (pressure-sensitive acrylics, rubber-based tapes, hot melts) usually show higher peel strength at higher rates until the failure mode flips from cohesive to adhesive, after which the strength can drop sharply. Rigid structural adhesives are less rate-sensitive but still not immune. The practical rule is simple: run at the standard rate, state the crosshead speed in the report, and never compare a 305 mm/min result with one from a 50 mm/min or 100 mm/min test such as ISO 8510-2.

Machine Requirements and Load Cell Selection

The apparatus clause has four requirements. The machine must measure load to within ± 1 %, the grips must clamp the specimen firmly without slipping, one grip must be power-driven at 305 mm/min, and the machine must produce an autographic record of load against separation; a peak-hold reading is not enough, because the result is an average over the plateau. The maximum load should fall between 15 % and 85 % of the machine's rated capacity. For a pendulum-type machine the standard adds that the maximum-load pawl must be disengaged so the pendulum swings freely, which is a reminder that it was written when dial testers were common; a modern electromechanical frame with a load cell meets the clause directly.

The 15 to 85 % rule applies to the load cell, not the frame, and peel loads are small. Typical 180° peel loads on a 25 mm strip run from under 2 N for a light label adhesive to 100 to 300 N for a structural adhesive on a rubber or fabric backing. A 100 N or 250 N load cell suits most tapes and labels; a 1 kN cell covers structural and rubber-to-metal bonds; a 5 kN or larger cell is wrong for nearly every D903 specimen because the signal sits below 15 % of range. Resolution matters as much as range: aim for a cell that resolves 0.1 % of the expected plateau load or better, so the average is not quantized.

Whatever the capacity, the crosshead must have enough travel. Peeling at least half of the 152.4 mm bond at a 2:1 crosshead-to-separation ratio takes more than 152 mm of crosshead movement before you add the initial 25 mm hand-separated tail; a frame with 250 mm or more of usable stroke is the practical minimum. See our universal testing machines and tensile testers for frames that meet the travel, rate and accuracy requirements.

Grips, Peel Fixtures and Alignment

The standard says only that the grips must clamp without slipping and that the specimen should be aligned in the plane of the clamp. In practice the gripping arrangement decides whether the peel angle stays at 180° and whether the flexible member tears at the jaw instead of peeling.

Flexible-to-rigid specimens

When the second member is rigid (metal, wood, rigid plastic), clamp it in the lower grip along its unbonded 50.8 mm end so that the bond line is parallel to the load axis. Clamp the folded-back tail of the flexible member in the upper grip. The thickness of the rigid member offsets the peel line from the load axis by half its thickness; a 180° peel fixture plate with an offset mount, or simply shimming the rigid member in a vise grip, puts the peel line back on the axis so the flexible member runs straight up to the upper grip. A roller grip on the flexible member is the usual choice for films and tapes, because the self-tightening action grips thin material without the edge damage that serrated jaws cause.

Flexible-to-flexible specimens

When both members are flexible, D903 is still a 180° peel (the T-peel at 90° each side is ASTM D1876, covered below). Grip the free tail of the flexible member in the upper grip and the unbonded end of the second member in the lower grip, with the bonded section hanging between them and the second member supported so that it does not fold forward. Light pneumatic grips with flat rubber-faced jaws hold fabrics, films and foams evenly across the 25 mm width at a repeatable clamping pressure; manual screw grips work but the clamping force varies from operator to operator, which shows up as jaw breaks.

Alignment checks

  • The bond line must be parallel to the load axis in both views; a specimen clamped at a slight angle peels from one edge first and gives a sawtooth trace.

  • The flexible member must not touch itself or the fixture as it folds back; rubbing adds friction load that appears as peel strength.

  • Grip the flexible member symmetrically across its width, so both edges peel together.

  • Account for the tare: the weight of a fixture, an offset plate or a hanging member registers on the load cell and must be zeroed or subtracted (section 8.2).

  • Hand-separate about 25 mm of the bond before clamping so that the peel front is already established when the crosshead starts.

Universal Grip builds dedicated 180° peel fixtures with an offset specimen plate and a matching upper grip, in adapters for Instron, MTS, ZwickRoell, Shimadzu, Mark-10 and other machines; see our tensile grips overview for the full range of vise, pneumatic, roller and wedge grips that can serve as the upper grip.

Step-by-Step Procedure

  • 1. Measure. Record the width of each strip to 0.1 mm at three points along the bond, and the thickness of each member. Note the conditioning history.

  • 2. Separate the start. Peel the free end of the flexible member back by hand for about 25 mm so that the peel front is established.

  • 3. Clamp. Fix the second member (or its unbonded tail) in one grip and the folded-back flexible tail in the other. Align the bond line with the load axis and make sure the flexible member runs straight to its grip without touching anything.

  • 4. Zero. Zero the load reading with the specimen clamped and slack, so that tare weight is removed; if a fixture plate hangs on the cell, subtract its weight.

  • 5. Set the rate. 305 mm/min crosshead speed (giving 152 mm/min separation). Confirm the machine has at least 160 mm of travel left.

  • 6. Run. Start the crosshead and record load against crosshead displacement. Peel until the average load is clearly established and at least half of the bonded length (76 mm or more) has separated; peeling the whole bond is better when the specimen allows it.

  • 7. Inspect. Stop, remove the specimen and record the failure mode (see below) and anything unusual: a tear in the flexible member, slippage in a grip, a dry spot in the bond.

  • 8. Repeat for at least ten specimens per adhesive. Discard a result only for a cause you can name (jaw break, flaw, slip), and say so in the report.

Calculating Peel Strength and Reading the Curve

A 180° peel trace has three parts: a rise and often a spike as the hand-separated front starts to move, a long plateau with noise on it as the bond unzips, and a drop as the bond runs out. Section 9 tells you to draw the best average load line through the plateau, correct it for tare, and report the load per unit width at 152.4 mm/min. The start and end transients are not part of the average. Software can do this as a mean of the sampled load between two cursors set well inside the plateau; some labs quote the average of a set number of peaks and troughs, which is acceptable as long as the method is stated and used consistently, but it is not what the standard describes.

Peel strength = average plateau load ÷ bond width. The standard writes the units as kilograms per millimetre (pounds per inch), which is a legacy of its 1949 origin; current practice reports N/mm, kN/m, N/25 mm or lbf/in. Pick one, state it, and never mix them in one table. The conversions: 1 N/mm = 1 kN/m = 25 N per 25 mm = 5.71 lbf/in; 1 lbf/in = 0.175 N/mm.

Worked example. A 25.2 mm wide strip of rubber-backed adhesive on an aluminum plate gives a plateau that averages 12.4 N between 40 mm and 120 mm of separation after a 15.1 N start spike. The fixture tare was zeroed before the run. Peel strength = 12.4 N ÷ 25.2 mm = 0.49 N/mm = 0.49 kN/m = 12.3 N per 25 mm = 2.8 lbf/in. The 15.1 N start spike is not used. If the ten specimens average 0.51 N/mm with a range of 0.44 to 0.58 N/mm, the report gives 0.51 N/mm as the result with the maximum and minimum, and lists the individual values if requested.

The report (section 10) must include the adhesive and specimen identification, surface preparation, bonding and cure details, conditioning and the atmosphere at test, the number of specimens, the test speed, the average, maximum and minimum peel strength, individual values and curves on request, and the type of failure. Because D903 has no precision and bias statement, there is no published repeatability to compare your scatter against; your own ten-specimen standard deviation is the only measure you have, and it belongs in the report.

Failure Modes and What They Tell You

  • Adhesive (interfacial) failure: the adhesive comes away cleanly from one member, leaving a bare surface. Usually a surface-preparation, contamination or primer problem rather than a weak adhesive; the number measured is the strength of the interface, not the adhesive.

  • Cohesive failure: the adhesive layer itself splits, leaving adhesive on both members. This is the result most adhesive manufacturers want to see, because it means the interfaces are stronger than the bulk adhesive and the measured value reflects the adhesive.

  • Substrate or stock break: the flexible member tears or the second member delaminates before the bond fails. The bond is stronger than the measured load; report it as a substrate failure, not a peel strength, and consider a stiffer backing.

  • Mixed failure: patches of adhesive and cohesive failure along the bond. Report the approximate percentage of each; a changing mix along the length often follows a bondline thickness gradient.

  • Backing stretch or zippering: a sawtooth trace with large amplitude usually means stick-slip peeling, common with pressure-sensitive adhesives at this rate; a steadily rising plateau usually means the flexible member is stretching and the peel angle is closing. Both are reasons to re-check the backing and the alignment.

Photograph both peeled surfaces of every specimen. Failure mode is a required report item, it is the fastest diagnostic you have when results scatter, and it is what a customer will ask for first when a bond fails in service.

ASTM D903 vs D1876, D3330, D6862, D1781 and ISO 8510

Peel standards differ in angle, rate, specimen and what is being qualified. Choosing the wrong one, or comparing results across them, is a frequent source of disputes between suppliers and customers.

  • ASTM D1876 (T-peel): two flexible adherends peeled apart in a T, each leg at roughly 90°, at 254 mm/min crosshead. Used for flexible-to-flexible laminates and bonded foils. Not interchangeable with D903 values.

  • ASTM D3330 (pressure-sensitive tape): 180° (Methods A to E) or 90° (Method F) peel of 24 mm tape from a stainless steel panel, applied with a 2 kg roller, at 305 mm/min, reported in N/100 mm or oz/in. The tape method for qualifying tapes, not bonded assemblies.

  • ASTM D6862 (90° peel): flexible-to-rigid at 90° with a sliding or translating fixture that keeps the peel angle constant, at 254 mm/min. Better for stiff flexible members that cannot bend through 180° cleanly.

  • ASTM D1781 (climbing drum): peel of a facing from a honeycomb or laminate core with a drum fixture; used for sandwich panels in aerospace.

  • ASTM D3167 (floating roller): flexible-to-rigid peel over a roller fixture at 152 mm/min, popular for structural film adhesives.

  • ISO 8510-2: the ISO 180° flexible-to-rigid method, run at 100 mm/min, so its results are not directly comparable with D903 at 152 mm/min separation. ISO 8510-1 is the 90° version; ISO 11339 is the T-peel equivalent of D1876.

  • ASTM F88: seal strength of flexible barrier materials; it looks like a peel but qualifies heat seals in packaging, not adhesive bonds.

For shear rather than peel, the companion method is ASTM D1002 lap shear. Many adhesive data sheets quote both, because an adhesive that is strong in shear can be weak in peel, and peel is what the bond sees at an edge, a corner or a flexing joint.

Common Errors

  • Crosshead set to 152 mm/min instead of 305 mm/min, which runs the test at half the specified separation rate.

  • Conditioning for 24 hours instead of 7 days, or at ± 2 °C / ± 5 % RH instead of ± 1 °C / ± 2 % RH.

  • Reporting the peak load instead of the best average load line through the plateau.

  • Not measuring strip width, or dividing by a nominal 25 mm when the strip is 24 mm.

  • Flexible member stretching or touching itself, closing the peel angle and adding friction.

  • Rigid member not offset, so the peel line is off the load axis and the flexible member runs at an angle.

  • Load cell too large (plateau below 15 % of range) or too coarse in resolution.

  • Fewer than ten specimens, or discarding low results without a stated cause.

  • Comparing D903 values with D3330, D1876 or ISO 8510-2 values as if they were the same test.

  • Omitting failure mode, surface preparation or cure details from the report.

Equipment for ASTM D903

A complete D903 setup needs a frame with 305 mm/min speed and adequate travel, a load cell sized to the adhesive, an upper grip that holds the flexible member without damage, a lower grip or fixture that holds the second member with the bond line on the load axis, and software that averages the plateau. Universal Grip supplies each part separately or as a system:

For a wider view of peel geometries and the fixtures that go with them, see our peel test fixtures guide and the adhesives testing page.

ASTM D903 FAQ

What is the specimen size for ASTM D903?

A flexible member 25 mm (1 in) wide by 304.8 mm (12 in) long, bonded over 152.4 mm (6 in) at one end to a second flexible or rigid member 25 mm wide by 203.2 mm (8 in) long. Neither member should be thicker than 3 mm (1/8 in). At least ten specimens are tested per adhesive.

What is the test speed for ASTM D903?

The separation rate is 152 mm (6 in) per minute. Because the flexible member is folded back on itself, the crosshead or power-driven grip must move at 305 mm (12 in) per minute to produce that separation rate.

What is the peel angle in ASTM D903?

Approximately 180°. The flexible member is folded back on itself and pulled in the direction opposite to the bond line.

How long must D903 specimens be conditioned?

Seven days at 23 ± 1 °C and 50 ± 2 % relative humidity, or until equilibrium, unless the adhesive manufacturer specifies a different conditioning. A desiccator over a saturated calcium nitrate solution (about 51 % RH) is an acceptable alternative to a conditioning room.

How is ASTM D903 peel strength calculated?

Draw the best average load line through the plateau of the load-versus-separation record, correct for tare, and divide the average load by the bond width. Report the average, maximum and minimum of the series in N/mm, kN/m, N/25 mm or lbf/in, and state the units used.

What units does ASTM D903 use?

The standard writes the result as kilograms per millimetre (pounds per inch). Modern practice reports N/mm, kN/m, N per 25 mm or lbf/in. 1 N/mm equals 1 kN/m, 25 N per 25 mm and 5.71 lbf/in.

How much of the bond must be peeled?

Enough to establish the average load clearly, and at least half of the 152.4 mm bonded length, so a minimum of about 76 mm of separation after the 25 mm hand-separated start.

What is the difference between ASTM D903 and ASTM D1876?

D903 is a 180° peel of a flexible member from a flexible or rigid member at 152 mm/min separation. D1876 is the T-peel test, in which two flexible adherends are pulled apart in a T shape at 254 mm/min crosshead speed. The geometries and rates differ, so the results are not comparable.

What is the difference between ASTM D903 and ASTM D3330?

D3330 is the pressure-sensitive tape method: 24 mm tape applied to a stainless steel panel with a 2 kg roller and peeled at 305 mm/min, reported in N/100 mm. D903 is for adhesive bonds between two members, with a 7-day conditioning, and reports load per unit width at 152 mm/min separation.

What load cell do I need for ASTM D903?

One sized so the plateau load falls between 15 % and 85 % of its range, with resolution of about 0.1 % of that load. Tapes and labels usually need a 50 to 100 N cell, coated fabrics and rubber-to-metal bonds 250 to 500 N, and structural adhesives on stiff backings 1 to 2 kN.

What grips are used for an ASTM D903 peel test?

An upper grip that holds the flexible member without damage (roller, pneumatic or vise grip depending on the material) and a lower grip or 180° peel fixture that holds the second member with the bond line on the load axis. For rigid second members the fixture offsets the member by half its thickness.

Does ASTM D903 have a precision and bias statement?

No. The standard has never been round-robin tested, so there is no published repeatability or reproducibility. Report your own standard deviation from the ten-specimen set.

Need a fixture, grips or a complete system for ASTM D903? Contact Universal Grip with your specimen details, capacity and machine connection and we will quote the right setup.

bottom of page