ASTM D378

What the test does

D378 is a family of tests, not one, all aimed at flat rubber conveyor belting. The central one is plain tension: a full-thickness specimen, dumbbell or rectangular strip, is cut from the belt, clamped by serrated wedges and pulled along the running direction until the carcass parts.

The rest surround it: strips peeled apart for ply and cover adhesion, a slit specimen pulled so a tear runs along the carcass, a short length hung from its edges to see whether it sags into a troughed idler set, and fasteners pulled from clamped specimens. Cover rubber goes to the general rubber methods; a burn chamber handles the flame test.

What it measures, and why it matters

Results are force per unit width — pounds per inch of width, or kN/m — not stress, because a belt is bought by width and rating rather than cross-section. Breaking strength sets the rating and, with a safety factor, the working tension the conveyor is designed around. Elongation, taken at break and at a reference load of a tenth of rated breaking strength, fixes the take-up travel needed; misjudge it and the belt runs slack or bottoms the take-up.

Elastic modulus, from low-frequency cycling rather than the break curve, describes how tension distributes along a long centre and how far the belt sags between idlers. Adhesion predicts ply separation, which is how textile-carcass belts usually die, and tear resistance whether a gouge from tramp material arrests or runs.

Specimen

Specimens are cut full-thickness, in the running direction for tensile work, and no sooner than five days after manufacture. Reference marks sit 100 mm apart on the dumbbell and no closer on the rectangular specimen, where they are kept clear of the clamps so grip pressure does not disturb the marked length.

Conditioning is real, not nominal: 23 ± 2 °C at 50 ± 5 % relative humidity, with at least three days' exposure beyond routine quality control. Where a minimum requirement exists one specimen may suffice; otherwise the median of three is reported, and referee work the mean of five.

What the machine must be capable of

No capacity is prescribed; the method asks only for a machine able to apply stress sufficient for the belting tested. As a guide, specimens from belting rated 160 to 1000 PIW part at roughly 0.7 kN to 4.5 kN at 25 mm wide and double that at 50 mm, while the 150 mm specimens used for mechanical fastenings can pass 20 kN. Carcass tear carries a rule of its own: the machine is chosen so the force lands in the upper 90 % of its rated capacity — a small frame, or a second low-range load cell.

Rate is specified per test. Most of the tensile work, breaking strength and elongation included, runs at 100 ± 10 mm/min; the carcass tear test is deliberately slower and continues until the tear has run a set distance. The modulus clause is the outlier — a few hundred sinusoidal cycles at a fraction of a hertz, between two low percentages of minimum breaking load — so it needs a dynamically capable frame, not a plain screw-driven one, with strain read over a gauge of at least 100 mm.

Grips decide whether the day is wasted: transversely serrated wedge faces, held flat and square, with rosin and coarse emery cloth expressly permitted against slippage. D378 states no accuracy class, but the cover-rubber and adhesion methods it calls up require conformance to Practices E4 — ±1 % of reading, equivalent to ISO 7500-1 Class 1. Ageing and immersion sub-tests run hot, for tens of hours in reference oils.

What goes wrong in practice

Slippage is the standing problem. Thick, plasticised belting creeps out of the wedges under load and inflates elongation before anyone notices; the tell is a trace that softens instead of climbing to a clean break.

Jaw breaks are the other side of it: clamp hard enough to stop the creep and the specimen fails at the grip line, below true carcass strength. On wide specimens the failure walks, cords parting one after another from an edge because the clamp sits out of square, and the trace staircases rather than dropping once.

Delamination is easy to misread: if plies separate before the carcass parts, the recorded strength is the adhesion's. Running tear work on the large breaking-strength frame is the last common error — a tear force read in the bottom few per cent of the cell's range is noise dressed as data.

Related and equivalent standards

ISO 283 is the nearest counterpart, covering full-thickness tensile testing of textile conveyor belting; it reports in kN/m but differs in specimen and conditioning detail, so figures should not travel between the two unchecked. Steel-cord and steel-cable belting is referred out to the ISO series for those constructions. Two confusions are worth naming: D412 is the tensile method for vulcanised rubber, used here only on cover compound, and D413 the adhesion method called up for peel — neither a belt standard.