only the latter part of the test (usually less than 50% and sometimes as little as 10% of the remaining life of the test [8]) can be characterised. Consequently it is only the accelerating portion that is uniquely described by C*. This is demonstrated clearly by [7,9-10]. In this latter stage it has also been suggested that measured values of C* become equivalent to the CCG rate [11]. This is because, at large CCG rates, most of the displacement rate results from increases in the elastic compliance due to crack extension and not from creep deformation taking place within the specimen. With these concerns in mind this paper examines an alternative approach. The development of the strain-based failure assessment diagram (SBFAD) is explored, which could potentially simplify the treatment of situations involving variable stress and variable temperature. A ferritic steel representative of the simulated heat affected zone of a low alloy steel is examined at 380oC using the new SBFAD. STRAIN BASED FAILURE ASSESSMENT DIAGRAM (SBFAD) A high temperature time dependent failure assessment diagram (TDFAD) based on the well established low temperature R6 approach [12], has been developed which allows predictions of creep initiation time and times for small amounts of CCG [13]. The TDFAD uses a high temperature ‘creep toughness’ which replaces the fracture toughness used in the R6 procedure. This ‘creep toughness’ parameter may be examined directly from experimental load-displacement information or indirectly from CCG rates as a function of C*. A strain based failure assessment diagram (SBFAD) has recently been developed as an alternative to the stress based TDFAD for high temperature components [14]. Strain accumulation can be measured or calculated and could be used to provide a measure of the components continuing performance. This section very briefly describes the SBFAD. The Option 2 FAD in R6 [12] may be extended to described a TDFAD [13] given by 1/ 2 0.2 3 0.2 2 − + = T ref c r c r T ref r E L L E K ε σ σ ε (1) where Kr=Kapplied/Kmat, Lr= σref/ , ε c 0.2 σ r is the reference strain corresponding to the reference stress σref, and is the stress at 0.2% strain from the isochronous stress-strain curve. It is possible to recast eqn 1 to give K c 0.2 σ r as a function of normalised strain εr, where [ ] r r r r L K ε ε 0.5 / 1/ 2 + = (2) and εr is the ratio of the total to elastic strain at reference stress, ( ) e ref T ref ε ε / . For a structure, a relationship between the reference strains and the load line displacements is required. Preliminary studies using the EPRI handbook solutions [15] have shown that for the compact tension (CT) specimen for a/W=0.5 and 0.75, where a is the crack length and W is the specimen width the normalised strain may be related to the load line displacement by − ∆ ∆ = 1.4 2.4 e T rε (3) where ∆Τ is the total load line displacement and ∆e is the elastic load line displacement. Figure 1 shows a schematic SBFAD, with curves for Kr as a function of εr for Lr
RkJQdWJsaXNoZXIy MjM0NDE=