The percent reduction of area for tension specimens made in the through-thickness direction of plates (%RA^Z) is often determined to give a relative measure of the resistance to lamellar tearing. Testing variables can affect the %RA^Z measured. Results are presented of investigations designed to examine some effects of (1) tension specimen design, (2) specimen location relative to the original ingot, (3) steel microstructure, (4) material strength level, (5) specimen-to-specimen variability, and (6) final diameter measurement reproducibility. Tension specimens in the Z-direction with an 8.9 mm (0.350 in.) diameter and 50.8 to 63.5 mm (2 to 21∕2 in.) parallel section gave the highest average %RA^Z and smallest variability. Specimen location had a very marked effect on the %RA^Z, with the lowest values found along the ingot centerline and highest levels at the ingot corners and edges. Normalized ASTM A537 Class 1 steel gave higher %RA^Z than when quenched and tempered to nearly the same ultimate tensile strength. Material strength level varied by changing the tempering temperature after quenching had only a small effect on the %RA^Z. Considerable specimen-to-specimen variability in the %RA^Z can exist even within a small area of a plate and reduce the significance of a single specimen result. Final specimen diameters can be difficult to measure and reproducibility of the %RA^Z calculated by different testing personnel for an individual specimen is sometimes poor.

Results indicate that specifications covering the through-thickness tension testing of roiled steel plates should reference the test location to the original as-cast ingot or slab if the region of lowest %RA^Z is to be reported. Specifications need to address the problem of specimen-to-specimen variability and %RA^Z reproducibility when determining the number of specimens to be tested and establishing acceptance criteria.