This work deals with the influence of the ratio crack length/specimen width (a/W) on crack opening displacement (COD) tests. A series of tests was performed on a weldable quality structural steel (50D, BS4360), where the ratio a/W was varied from about 0.2 to 0.8, keeping W constant. The influence of temperature was assessed by a series of tests at different temperatures over the range -90°C to room temperature.

It is concluded that critical COD, δ_c, is insensitive to a/W at low temperatures, but becomes dependent on a/W as temperatures approach the transition temperature. This behavior has been quantified and should be borne in mind in situations where the use of specimens with different a/W ratios is required.

The tests were instrumented with a clip gage at the crack mouth and a load point displacement transducer which enabled a relationship between COD and J to be determined.