ASTM D1002, "Standard Test Method for Apparent Shear Strength of Single-Lap-Joint Adhesively Bonded Metal Specimens by Tension Loading (Metal-to-Metal)," is the most widely cited single-lap-shear protocol for industrial structural-adhesive qualification [S1][S2]. The method loads a flat metal-to-metal bonded coupon in tension along the long axis until the bond line fails, and reports the peak load divided by the bonded overlap area as the apparent shear strength in MPa or psi [S2].
Its scope is intentionally narrow: the standard is limited to rigid metal substrates (aluminum being the canonical reference), so the result is an "apparent" rather than "true" shear value, because the single-lap geometry produces a non-uniform stress field that mixes shear with peel at the overlap ends [S2][S7]. For non-metal adherends, ASTM D3163 (plastic-to-plastic) and ASTM D3164 (plastic sandwich) are the parallel test methods used in the same lap-shear form factor [S5].
Specimen geometry and substrate preparation
The reference D1002 coupon is two flat metal tabs bonded over a 12.7 mm (0.5 in) overlap, with a standard specimen width of 25.4 mm (1.0 in) and individual substrate thickness between 1.6 mm (0.063 in) and 3.2 mm (0.125 in) [S2]. Substrate selection is by agreement between purchaser and supplier, but aluminum alloy 2024-T3 is the canonical reference material in published D1002 datasets [S2][S4].
Surface preparation drives the result more than any other variable. Accepted treatments include solvent degreasing, grit blasting, and chemical etch, all of which are documented in the test report because the same adhesive can return 2x to 3x different shear values on as-received vs. etched aluminum [S2]. Specimens are conditioned at 23 ± 2°C (73.4 ± 3.6°F) and 50 ± 5% relative humidity for a minimum of 40 hours before testing unless an alternate environment is specified [S2].
Test machine setup, grips, and loading rate
The D1002 pull is performed on a calibrated universal testing machine with a load cell accurate to ±1% of the expected failure load, and self-aligning or wedge-action grips with serrated or rubber-coated jaw faces to suppress slippage [S2]. Misalignment above roughly 1° to 2° introduces peel stresses that can lower the apparent shear strength by 20% to 40% on ductile adhesives, so clevis pins or bearing-block fixtures are used where the tab stock is thin enough to deflect [S2].
The standard test speed for metal-to-metal D1002 specimens is 1.3 mm/min (0.05 in/min) for substrates up to 1.6 mm thick, rising to a higher crosshead rate for thicker adherends, with the exact rate governed by substrate stiffness rather than adhesive chemistry [S2]. Load and elongation are recorded continuously so the report can distinguish cohesive failure inside the adhesive, adhesive failure at the interface, and substrate failure, all of which are noted in the failure-mode code that accompanies the MPa figure [S1][S6].
Reporting apparent shear strength and failure mode
The reported value is the maximum load divided by the overlap area (length × width, e.g. 12.7 mm × 25.4 mm = 322.6 mm²), expressed in MPa (psi), and called "apparent" because the single-lap geometry is not in pure shear [S2][S7]. Typical room-temperature results for structural epoxies on grit-blasted 2024-T3 aluminum sit between 17 MPa and 35 MPa (roughly 2,500 to 5,000 psi), with modified acrylics and polyurethanes ranging lower and high-temperature-cure epoxies and film adhesives extending well above 35 MPa [S2][S4].
Failure mode is reported alongside the strength number using a percentage breakdown of the bond area, and the dominant code (cohesive, adhesive, substrate, or mixed) is what an engineer should compare between suppliers rather than the headline MPa value alone [S1][S6]. For procurement, common acceptance windows are 17.2 MPa minimum (a figure used in some automotive OEM material specs for aluminum lap shear) with 100% cohesive failure, but the specific acceptance criterion is set by the calling spec, not the D1002 method itself [S2].
How D1002 fits alongside the related lap-shear family
ASTM D1002 anchors a family of single-lap and sandwich-lap methods that share the same specimen shape but change the adherend: D3163 for rigid plastic-to-plastic, D3164 for plastic sandwich, D5868 for fiber-reinforced plastic (FRP) bonding, and D2344 for short-beam strength of polymer-matrix composites [S5]. All four use a similar overlap-and-pull geometry, which is why a D1002 fixture and grip stack will, with different specimen prep, run the rest of the family with no hardware change [S5].
For loading beyond a single static pull, D1002 is paired with ASTM D2919 (lap-shear fatigue), D2295 (elevated-temperature lap shear), and D5656 (thick-adherend shear), so a single bonded coupon geometry can carry a full environmental and durability matrix without re-tooling the fixture set [S2]. The reference method is also the calibration baseline for OEM incoming-inspection tests, which is why supplier datasheets will quote a "D1002 aluminum" value even when the part will be bonded to steel or composite in service [S4][S8].
Selection rules: when D1002 is the right method, and when it is not
D1002 is the right method when the production bond is a flat metal-to-metal overlap loaded in static tension or shear, and the goal is to compare adhesive lots, surface treatments, or cure cycles on a common reference [S1][S2]. It is also the right method for incoming-quality acceptance of bulk adhesive, because the small coupon size (under 80 mm long) lets a lab run 5 to 10 replicates per batch in a single day on a standard electromechanical UTM [S2][S5].
D1002 is the wrong method when the service load is peel-dominant, when the bond line is thick (above about 0.5 mm) and the adherends are compliant, or when the substrate is not a rigid metal: for those cases the apparent-shear number is misleading and the test plan should move to a thick-adherend or specialty geometry [S2][S7]. It is also the wrong method for impact or fatigue life, which need D950-type or D2919-style methods, because a single quasi-static D1002 pull says nothing about cyclic endurance or rate dependence [S2]. In practice, surface prep and cure regime drive a larger swing in D1002 numbers than the adhesive grade itself, so a surface-prep and cure-time audit should run alongside any new D1002 dataset before the result is signed off as a specification.
Common failure modes and how to read a D1002 result
A clean cohesive failure (failure inside the adhesive layer, leaving adhesive on both tabs) at 25 MPa is a stronger signal of bond quality than a 30 MPa result that shows 80% adhesive (interfacial) failure, because the higher number is being carried by mechanical keying rather than by the adhesive itself [S1][S6]. Substrate failure (the aluminum tearing before the bond fails) means the adhesive is stronger than the test fixture, so the test is no longer measuring the adhesive: the report should state this and the result should be reported as a lower bound only [S2].
Variability in D1002 is normally 5% to 15% coefficient of variation for 5 replicates on a well-controlled lab prep, and anything above 20% usually points to a surface-prep, bondline-thickness, or alignment problem rather than to the adhesive batch [S2][S6]. The 12.7 mm overlap is the geometric source of a large fraction of that scatter, so where lower variance matters, the thick-adherend geometry of D5656 is used, at the cost of a much larger and more expensive specimen [S2].
For a broader look at how this test plugs into batch-level QC, see the industrial-adhesive batch mixing and QC testing workflow, and for the upstream polymer-to-pallet view, the industrial-adhesive manufacturing chain reference lays out where in production a D1002 coupon is most representative of the bulk material.
Component reference pages worth checking: industrial adhesive, electronic test, and measurement test.