The author has recently completed an evaluation of an engineering failure assessment approach developed at the Central Electricity Generating Board (CEGB) of the United Kingdom and became aware of certain limitations of this approach. The CEGB failure assessment approach, referred to in the United Kingdom as R-6, is an engineering approach to the elastic-plastic fracture mechanics assessment of structural components.The use of the R-6 diagram allows a straightforward prediction of maximum load which a given structure can safely withstand.In addition, for a given load and either a postulated or actual defect size, the margin of safety of the structure can be directly determined from the diagram. The original concept of the R-6 approach came from the CEGB Dowling and Townley two-criterion approach [i] which states that structures will fail by either of two mechanisms: brittle fracture or plastic collapse. Harrison, Loosemore, and Milne of CEGB [2] reformulated the two-criterion approach into what is known today as R-6.The author used the R-6 failure assessment diagram to predict the ductile tearing behavior of compact fracture test specimens of HYI30 steel, A533B steel, and 7075-T651 aluminum [3]. While it is known that the R-6 assessment curve is based upon the elastic-perfectly plastic Dugdale strip yield model, the author used a simple approach [3] to account for strain hardening effects of the A533B steel. This approach was to take the average of the yield and ultimate strengths of the material for the flow stress. While this simple approach gave fairly accurate predictions of maximum load values for the displacement-controlled compact specimens, loads after maximum load were poorly predicted. Prediction of maximum load for load-controlled structures is adequate, as this corresponds to the instability point of the load-controlled structure. For displacement-controlled structures or specimens, the maximum load point does not correspond to the instability point, and therefore it is of interest to know the complete load history. For HYI30, a material which exhibits very little strain hardening, the R-6 diagram predicted the behavior of the HYI30 compact specimens for the complete load history quite well. On the basis of this work [3], it was concluded that while the current R-6 curve could be used for materials with little or no strain hardening, prediction of specimen or structural behavior for materials with significant strain hardening would require an alternative assessment curve which accounts for these strain hardening effects.(The 7075-T651 specimens responded in an essentially valid linear elastic manner.)The essence of the failure assessment diagram is that it is a w~iversal failure curve that has been normalized with respect to the linear elastic fracture mechanics (LEFM) behavior of the specific structural configuration at one extreme, and the limit load of the structure Int Journ of Fracture 16 (1980) R74 at the other extreme.In terms of the J-integral, the equation of the failure assessment curve can ...