Determination of constraint limits for cleavage initiated toughness data

The proper constraint limits for cleavage initiated toughness data within the ductile-to-brittle transition regime have been studied extensively using both numerical analysis and analysis of experimental data. Historically, the experimentally based constraint limits have supported less conservative limits. This study conducts analysis of existing and new experimental data developed using data sets targeted to exhibit constraint loss toughness enhancement. Constraint herein is quantified in terms of the scaled specimen deformation level, more commonly known as M. It is expected that data with low M values will exhibit the greatest affect of constraint loss. Large data sets are therefore developed and extracted from the literature that include data with a large range of M levels and at least the required number of uncensored results with M > 30 for valid T0 measurement as per ASTM E1921-02. Differences in the calculated T0 values using censoring limits of 5–500 are then determined. The onset of Tlim differences due to constraint loss is examined by simply increasing the censoring limit, Mlim, utilized in determining the indexing temperature, Tlim, and evaluating differences between T0 and the Tlim values obtained using higher constraint limits. Bias resulting from this censoring procedure is examined using a Monte-Carlo analysis and shown to be small. ement of a high constraint T0 in bend specimens is shown to require Mlim > 200. As Mlim increases from 30 to 200 in bend specimens, the corresponding Tlim can increase by approximately 15 °C. Further increases in Mlim do not result in substantial increases in Tlim. This evolution buttresses previous numerical findings by Dodds and co-workers [Specimen size requirements for fracture toughness testing in the ductile-to-brittle transition regime, J Test Eval 1991;191:123–34; Size and deformation limits to maintain constraint in KIc and Jc testing of bend specimens. In: Kirk M, Ad Bakker, editors. Constraint effects in fracture theory and applications: second volume. ASTM STP 1244, 1995; Numerical investigation of 3-D constraint effects on brittle fracture in SE(B) and C(T) specimens, Int J Fract 1995;74:131–61] and provides a strong justification for changes to ASTM E1921-02 if a conservative, geometry insensitive, and transferable reference temperature, T0, is to be determined using this standard. Possible short-term changes to ASTM E1921-02 could include raising Mlim and requiring an upward shift of SE(B) T0 values. A more desirable solution is to adjust individual KJc toughness values before evaluation of T0 to eliminate specimen geometry bias.