Standard 25 mm
Pneumatic Vice Action Grip
Flat vice-action faces grip a thin flexible adherend along its full width without cutting it — a peel test fails at the grip long before it fails at the bond if the jaws bite.
SpecificationsTesting standard
ASTM D903
Standard Test Method for Peel or Stripping Strength of Adhesive Bonds
Written and technically reviewed by Dak System Inc. engineeringLast reviewed
ASTM D903 measures the peel strength of an adhesive bond between a rigid and a flexible adherend. The flexible member is folded back through 180° and stripped off the bond at a constant rate, and the result is force per unit width read from the running portion of the trace — reported together with the observed failure mode, which is often the more useful half.
At a glance
- Test type
- Peel & adhesion — a bonded joint is pulled apart
- Published by
- ASTM
- Edition
- D903-98(2025)
- Material
- Adhesives, tapes & bonded joints
- Runs on
- Series 7200 and Series 9000
What the test does
Two adherends are bonded over a defined length, one rigid and one flexible. The unbonded end of the flexible member is folded back on itself through 180° and clamped in the moving grip; the rigid member is held in the opposing grip or carried on a support plate. The crosshead separates at a constant rate and strips the flexible member off the bond while force is recorded continuously against travel.
What it measures, and why it matters
The result is peel strength — force per unit width of bond, reported from the running portion of the trace rather than a single peak, alongside the observed failure mode. Adhesive suppliers use it for lot release and shelf-life monitoring, because a drifting average peel value catches a bad batch before it reaches a customer. Formulators use it to rank candidate adhesives against a named adherend pair. In failure investigations, the split between adhesive and cohesive failure tells you whether to fix surface preparation or the adhesive itself.
The bonded assembly
- Width
- Nominally 25 mm
- Bonded length
- About 152 mmWith enough unbonded flexible material at one end to double back and reach the grip.
- Adherend pairing
- One rigid, one flexibleThe flexible member must tolerate bending through 180° without cracking — so the PAIRING decides whether the geometry is usable at all, not just the adhesive.
- Cut edges
- Clean, to widthA ragged or nicked edge starts a tear that reads as low peel strength.
- Conditioning
- Nominally 23 °C and 50 % RHA conditioning room or desiccator is part of the required apparatus, not an optional extra.
- Discard and replace
- Tearing through the flexible adherend, delaminating at a cut edge, or peeling outside the bond line
- Record the failure mode
- On every specimenDakAdhesive failure at the interface and cohesive failure through the glue line point at different fixes — surface preparation or the adhesive itself. The force number alone does not distinguish them.
Test speed
- Rate of jaw separation
- 152.4 mm/min (6 in./min), constantThe method calls for a power-driven machine of the constant-rate-of-jaw-separation type. The peel front advances at about half that rate, so a 152 mm bond needs on the order of 300 mm of usable stroke.
- Read from the running portion
- Not the initial peakThe first spike is the peel front starting, not the bond's resistance to being peeled.
Calculations
Peel strengthP
P = F_avg / w
- F_avg
- average force over the running portion of the peel, N
- w
- specimen width, mm
Force per unit WIDTH — N/mm or lbf/in. Measure the width on each specimen rather than assuming it from the cutting die, since it is the divisor.
How the test runs
- 01Prepare the adherends with the surface treatment the test plan specifies — it is usually what the test is really comparing.
- 02Bond the assembly over the specified length, leaving enough flexible material free at one end.
- 03Cure to the adhesive's schedule.
- 04Cut cleanly to width, and measure the width of each specimen.
- 05Condition in the standard laboratory atmosphere.
- 06Fold the free flexible end back through 180°.
- 07Clamp the flexible end in the moving grip and the rigid member in the opposing grip or on its support plate.
- 08Separate the jaws at a constant 152.4 mm/min (6 in./min), recording force against travel.
- 09Let the peel front reach a steady state before taking data.
- 10Average the running portion, excluding the initial peak.
- 11Record the failure mode — adhesive, cohesive, or adherend failure.
- 12Discard anything that tore through the flexible member or peeled outside the bond.
The fixture this method needs
What the report has to contain
- Reference to ASTM D903 and the edition
- Adhesive identification, batch and cure schedule
- Both adherends, and the surface preparation used on each
- Bond dimensions and specimen width as measured
- Conditioning and test atmosphere
- Rate of grip separation
- Peel strength as force per unit width, from the running portion
- FAILURE MODE — adhesive, cohesive, or adherend
- Number of specimens, mean and standard deviation
- Any specimen discarded, with the reason
What the machine must be capable of
Forces are low. A 25 mm strip carrying a weak pressure-sensitive adhesive may peel at around a newton, while a tough rubber-to-metal bond can demand a few hundred, so a 50 N to 1 kN load cell on a single-column frame covers essentially all of this work. Resolution at the bottom of that range matters more than capacity: a cell sized for the frame rather than the specimen buries a soft peel trace in noise. The method calls for a power-driven machine of the constant-rate-of-jaw-separation type, and separation is fixed at 152.4 mm/min (6 in./min); peel is rate-sensitive, so a machine that cannot hold that rate under a fluctuating load shifts the number.
Stroke is the constraint people miss. A 180° peel advances the peel front at roughly half the crosshead rate, so stripping a 152 mm bond needs on the order of 300 mm of usable travel. No extensometer is used — peel is a force-per-width measurement, not a strain measurement.
The fixture must hold the peel angle at 180° as the front travels down the bond. A fixed pair of grips lets the rigid adherend swing as the free length changes, so the angle drifts and the recorded force drifts with it; a support plate or roller arrangement that translates with the peel front holds it. Jaw faces must grip a 25 mm strip without slipping or cutting into it.
What goes wrong in practice
Peel angle drift is the commonest error: the rigid adherend rotates during the run and the trace slopes for geometric reasons that look like a real change in the bond. Stick-slip peel produces a saw-toothed trace whose peaks are not the peel strength — average the running region, and report that the trace was unstable. Confusing adhesive-interface failure with a cohesive split misdirects the corrective action entirely, so inspect both peeled faces before recording a number.
D903 or D1876
Two peel tests, and the difference is what the adherends are.
| ASTM D903 | ASTM D1876 | |
|---|---|---|
| Adherends | One rigid, one flexible | Both flexible |
| Geometry | 180° peel | T-peel |
| The bend | Flexible member folds back on itself | Both arms bend away from each other |
| Reports | Force per width, running portion | Force per width, average over a fixed peel length |
| Fails if | The flexible member cracks when folded | Either arm is too stiff to bend |
Neither is a design allowable. Peel results depend heavily on the geometry, the peel angle and the stiffness of the adherends, so they rank bonds and detect process drift — they do not predict the strength of a joint of different shape.
Questions we are asked about this test
What is ASTM D903?
It is the ASTM method for peel or stripping strength of adhesive bonds. Two adherends are bonded, one rigid and one flexible; the flexible member is folded back through 180° and stripped off at a constant rate, and the result is the force per unit width taken from the running portion of the trace, reported with the observed failure mode.
Why is the result read from the running portion rather than the peak?
Because the initial spike is the peel front getting started — it reflects how much energy it takes to initiate separation at one point, not how the bond resists being peeled. The steady running force is the reproducible quantity, and it is what the method asks for.
Why does the failure mode matter as much as the number?
Because it tells you what to fix. Adhesive failure at the interface points at surface preparation; cohesive failure through the glue line points at the adhesive itself; adherend failure means the bond outlasted the substrate. Two specimens can give the same peel force by different mechanisms, and only one of them is a problem you can solve with better cleaning.
What is the difference between D903 and D1876?
The adherends. D903 bonds one rigid member to one flexible one and peels the flexible one back through 180°. D1876 bonds two flexible adherends and bends both away from each other to form a T. If both of your adherends are flexible, D903's geometry cannot be built.
Can I use peel strength as a design allowable?
No. Peel results depend strongly on the geometry, the peel angle and the stiffness of the adherends, so a figure measured on a 25 mm strip at 180° does not transfer to a joint of different shape. Peel tests rank adhesives, monitor lots and detect process drift — all things they do well — but they are not structural design data.
Why did my specimen tear instead of peeling?
Either the flexible adherend was not flexible enough to fold through 180° without damage, or a ragged cut edge started a tear that ran ahead of the peel front. Both invalidate the specimen. The adherend pairing is part of what makes this geometry usable, so a material that cracks when folded needs a different peel method rather than a different operator.
Running ASTM D903 on the Series 7200 and Series 9000
Dak verifies against whichever standard the method names, and where a class applies our frames sit a class tighter than it asks.
| The method asks for | Dak supplies | |
|---|---|---|
| Capacity | Low forces on a 25 mm strip — anything from about a newton for a weak pressure-sensitive bond up to a few hundred newtons for a tough rubber-to-metal bond, so a 50 N to 1 kN load cell on a single-column frame covers essentially all of this work. | Load cells from 1 kg to 60 ton on the Series 7200, and 0.5 to 100 kN on the Series 9000 |
| Force accuracy | unknown | ISO 7500-1 Class 0.5 — the method sets no class of its own |
| Gripping | 180° peel arrangement — rigid adherend clamped low or on a support plate, flexible tab folded back into the moving grip | Our peel and adhesion fixtures, built to the specimen |
| Environment | Condition and test in the standard laboratory atmosphere, nominally 23 °C and 50 % RH (published tolerance band unconfirmed); a conditioning room or desiccator is part of the required apparatus | 3009 series chambers, −150 °C to +400 °C — temperature only |
This page describes the method as practised. The governing text is the current edition from the issuing body. Tell us what you are testing and we will answer with the machine, the fixture and a quotation.