The mechanism of caustic cracking of carbon steels—I. Influence of electrode potential and film formation

The anodic behaviour of a 0·1% carbon steel wire in 10M sodium hydroxide solution at 121°C has been studied at different electrode potentials under static conditions, and while the wire yields at strain rates between 1·5 and 436%/min.The static material forms a fairly coherent and adherent film of magnetite in the potential range ca. ?0·60 to ?0·85 V(she), while the anodic current density falls to ca. 0·2 mA/cm² (at ?0·70 V) in ca. 40 min. When such a specimen is made to yield at 436%/min, the overall current density continuously rises to over 8 mA/cm² at the bared metal area produced by the cracking of the oxide film. This current density is sufficient to account for the rate of crack propagation measured microscopically on specimens strained at 1·5%/min. Equivalent results are reported for potentials over the whole range ?0·60 to ?0·85 V(she).On either side of this potential range little or no adherent film is formed, the anodic current density does not fall below ca. 1 mA/cm² and on straining rises to no more than ca. 5 mA/cm²; and no cracks are produced.The results strongly support the theory that crack propagation occurs because bared metal at the yielding advancing edge of a stress-raising crack can dissolve several hundred times as fast as that at the static crack sides, which are continuously protected by the growth of film. The crack thus maintains its acuity, and the raised stress at its edge maintains ductile yielding of the metal so that the electrochemical crack advancement proceeds without the need for any mechanical cracking.A comparison with previously reported work on the same steel in hot concentrated nitrate solutions shows that the range of anodic current densities found on bared metal, and the corresponding range of crack propagation rates, are about an order of magnitude less in the present hot concentrated hydroxide solution.