ASTM D1876

What the test does

Two flexible strips bonded together over most of their length are pulled apart. The short unbonded ends are bent away from each other and clamped in opposing grips, so the assembly forms a T. The crosshead separates the grips at a constant rate and the bond splits progressively along the joint, while force and head travel are recorded continuously through the separation.

What it measures, and why it matters

The reported quantity is peel resistance: the average force per unit specimen width sustained while the peel front advances steadily, taken over a fixed length of peel and excluding the initial load spike. It is a lot-release number for tapes, films and laminates, and a screening number during adhesive formulation. It also grades surface preparation, since the same adhesive on a degreased and on an untreated adherend separates by a wide margin. It ranks bonds; it is not a design allowable for a joint of different geometry.

Specimen

The specimen is a pair of flexible adherends — thin sheet metal, film or laminate — bonded over most of their length with a short unbonded length left free at one end. Those free ends are bent back in opposite directions to form the T. Strips are commonly 25 mm wide, and because the result is force per unit width, that width is measured on each specimen rather than assumed from the cutting die. Both arms must be flexible enough to bend without taking a permanent set; a stiff arm converts the test into a bending test and the number is no longer peel. Specimens are conditioned in the standard laboratory atmosphere of about 23 °C and 50 % relative humidity for at least 24 h before test. Replicate counts and panel dimensions are fixed in the standard's own text.

What the machine must be capable of

Forces are modest. A film laminate peels at tens of newtons on a 25 mm strip and a toughened structural adhesive on thin metal arms at a few hundred, so a 100 N to 1 kN load cell covers almost all of this work, with 5 kN reserved for outliers. Capacity matching is the binding constraint rather than frame size: a 20 N peel plateau read on a 5 kN cell sits in the noise, and the fine structure of the trace is lost. The force-accuracy class required is set in the standard's own text.

The method fixes head speed at 254 mm/min (10 in./min), which separates the bond at 127 mm/min because both arms move. Peel is rate-sensitive, so the speed must be held constant, not merely averaged. Stroke matters more than it does in tension: recording the required 127 mm of peel after the initial peak needs roughly 250 to 300 mm of usable travel, and a short-stroke bench runs out mid-test.

No extensometer is used — the record is force against head travel. No dedicated peel fixture is called for either; the two bent arms go into plain opposing tension grips, mechanical vice-action or pneumatic side-action, with non-slip faces. Alignment is the critical detail: the arms must sit symmetric about the load axis. Testing is done in the same 23 °C, 50 % RH atmosphere used for conditioning.

What goes wrong in practice

Arm yield is the commonest invalidation — a metal adherend that bends permanently absorbs work that is then reported as adhesion. Asymmetric clamping runs the peel front crooked across the width, and the trace drifts downward as the effective bonded width shrinks. Stick-slip peel gives a sawtooth trace as the crack alternately arrests and runs; the mean is meaningful, single points are not. Including the initial peak in the average inflates the result, sometimes by a factor of two.

Related and equivalent standards

ISO 11339 is the nearest counterpart and covers the same flexible-to-flexible T-peel geometry; the two are routinely cited together. Specimen dimensions, conditioning and reporting are specified independently in each document, so results should not be quoted across them without checking both. Where one adherend is rigid, the T geometry cannot form and the work falls outside this method entirely — a different peel configuration is required.