Limitations of some methods of establishing fracture-toughness data

Abstract

Several methods of measuring fracture-toughness values have been assessed with emphasis on their practical limitations. Uncertainties in the evaluation procedures, considerations of specimen size, and scatter in material properties, lead to restrictions on the usefulness of all these methods. The methods examined were: (i) K_IC according to ASTM E399–78; (ii) J-integral testing according to the ASTM Recommended Procedure (March 1977); (iii) crack-opening displacement (COD) using an extrapolation technique; (iv) a simple equivalent-energy approach; (v) instrumented Charpy testing; (vi) an empirical method according to Begley and Logsdon. The success of the methods was judged by their ability to give data comparable with the ASTM E399 procedure in the linear-elastic regime, and a continuously rising extrapolation thereof at higher temperatures. The experimental results obtained from a rotor steel at low temperatures show that the COD method yielded values which cannot be determined with sufficient accuracy. At higher temperatures, where valid ASTM K_IC data can no longer be obtained, excessively conservative values are provided by the instrumented Charpy and empirical methods, whereas the J-integral, COD, and equivalent-energy methods yield realistic data. Additional tests on a pressure-vessel steel further demonstrate the value of the latter techniques. Since the J-integral method can also be used in considering extended slow crack growth in components, it appears to be a particularly useful procedure. However, further refinement of the method is required before it can be considered universally applicable.