The fracture toughness of ferritic steels has been characterized by numerous different parameters. It is not the purpose here to discuss history, so the main parameter is linear-elastic KIc and its use with respect to the elastic-plastic Kjc. The KIc parameter has, in the past, been one of the most commonly used parameters, also for structural steels, but its limita­tions for describing the transition behavior controlled by both cleavage and ductile cracking are widely recognized (discussed later in detail) today. Many high-alloyed quenched and tempered steels, which exhibit practically no plastic deformation, still have moderate fracture toughness, KIc, and can be used and no additional benefit is achieved by using an elastic-plastic parameter like KJc. For these steels, the measurement of fracture toughness at one or a few temperatures is all that is necessary. For low — alloyed structural steels, which typically exhibit a pronounced ductile-brittle transition and may be loaded in a wide temperature range, the situation is different. In this case, an elastic-plastic parameter is needed. An example of such an application is an irradiated RPV where safety and performance have to be demonstrated in accident and abnormal service conditions which are more severe (i. e., at lower tem­peratures) than in normal operation. The integrity analyses must be based on the material fracture toughness covering the entire transition tempera­ture range. In the mid and lower temperature por­tions of the transition curve, cleavage fracture has to be explicitly considered, and the parameter KJc and Master Curve analysis of data are the preferred methods to determine fracture toughness. It should be noted that application of the methodology is not just limited to ferritic steels, but the fracture is generally cleavage or a stress-controlled type of mechanism.

When a cracked material is loaded, a plastic zone will develop at the crack tip. The size of this plastic zone depends on the crack tip loading (stress inten­sity) and the material yield strength. The radius of the plastic zone (rpl) can be expressed for mode I loading (tension loading perpendicular to crack plane) in a simple form as follows: ±( Kl

2я ffys

where KI is the stress-intensity factor and sys is mate­rial yield strength.

The plastic zone size is thus a measure of plasticity at the crack tip and can be used to assess the applica­bility of different fracture toughness parameters. In predominantly elastic cases, the plastic zone size can be very small, and the material may be analyzed using LEFM. The parameter KIc is generally determined without correction term for plasticity, when the plas­tic zone size is small in relation to the specimen dimensions. Otherwise, an elastic-plastic parameter such as the J-integral should be used.

For a compact specimen [C(T)] geometry and load­ing, the value of Kj can be calculated directly from the load (Pt) and the original crack length (a0/ W):

Pi

vWrnffi where W is specimen width, B is thickness, Bnet is net thickness, and function f (a0/W) is defined dependent on specimen crack depth.

For elastic-plastic conditions where a large plastic zone has developed and some stable crack growth can occur, one normally cannot use the LEFM parameter KIc and eqn [3] without at least some correction term(s). The use of EPFM is more practical to deter­mine directly the J-integral which takes into account the plastic component of the work done to the speci­men or component. The fracture toughness equiva­lence Kj, denoted as Kjc, can then be converted from the J-integral, which is first divided into elastic and plastic components:

Jc — Je + Jp [4]

where the elastic component of J is calculated from the elastic K (Ke) as follows:

J — [5]

where E is the elastic modulus and n is Poisson’s ratio.

The plastic component of J (Jp) is calculated from the total and elastic work using the measured load versus load line displacement or crack opening data. The value of Jc is converted to Kjc as follows:

K* 4 Jc rb [6]

Despite the limitations of LEFM as discussed earlier, one of the basic standards applied in the past in fracture toughness testing has been American Society for Testing and Materials (ASTM) E 399 which
defines the methodology for Kjc determination. This standard has recently been revised and issued as part of ASTM E 1820-08 (Annex 5),2 but the basic approach is essentially the same as in the previous versions of ASTM E 399. Because these standards, as well as a corresponding linear-elastic European stan­dard, BS 7448, are still being used to assess some ferritic steels, it is important to know how they differ from the Master Curve approach used in ASTM E 1921-08.3 In particular, the scope of the application requires discussion since it affects how Kjc data should be analyzed with respect to Kjc.

A common feature of LEFM KIc standards (such as the previous ASTM E 399) is that they express the fracture toughness as a single value, which should be a material property characterizing the resistance of a material to fracture. The value of KIc should be insensitive to specimen size, if the measured value fulfills the specified size criteria. When these condi­tions are met, the stress-strain condition at the crack tip has been thought to be predominantly plane strain, thus ensuring sufficient constraint to produce a minimum fracture toughness value for the material. The qualification to obtain a valid KIc measurement requires that relatively large test specimens have to be tested. The value of KIc is supposed to represent a lower limiting value of fracture toughness, but the method (ASTM E 1820, Annex 5) does not ade­quately cover ferritic steels where fracture by a cleav­age mechanism in the transition or lower shelf region has known statistical characteristics different from ductile initiation fracture toughness.2

Based on the current knowledge of fracture mechanics, confirmed by numerical finite element model analyses of local stress-strain fields, many of the arguments for the LEFM parameter KIc are not valid, especially for ferritic steels.4 First, for crack tip constraint, the KIc size requirement has been shown to be overly conservative, leading to highly oversized specimens. Also, it has been shown that KIc is not a size-insensitive parameter; a size adjustment, similar to that made for Master Curve specimens (discussed later in Section 4.14.1.3), should also be made to the values of KIc to correct for size effects and to make the data comparable with Kjc. Justification of this size effect argument has been demonstrated in many com­parisons made between the Kjc and the older KIc data measured with different size specimens.4 It is impor­tant to note that increasing the specimen size (both ligament and thickness) gradually diminishes the effect of the size adjustment, which means a reduced,
but still existing, dependence on specimen size even with large dimensions (discussed later in Section 4.14.1.3.3). One should also remember that the Mas­ter Curve size adjustment is valid only in the transi­tion region where cleavage fracture initiation is expected to occur, even after some stable crack exten­sion before cleavage fracture. Due to the size effect, the fracture toughness decreases in the transition region with increasing specimen size whenever cleav­age fracture is encountered. Examples of applying the statistical size adjustment are given in Section 4.14.3.

As mentioned previously, ASTM E 1820 (Annex 5) actually invalidates the KIc determination if the value exhibits transition behavior indicative of some cleav­age fracture. If KIc values characterize only the upper shelf behavior of ferritic steels, no size effect typical of the transition behavior should exist even iflater cleav­age initiation occurs. Thus, KIc determinations per­formed as per the previous ASTM standard versions do not necessarily fulfill the requirement or follow the recommendations now in ASTM E 1820. This issue is significant, because it concerns an essential qualifica­tion criterion of KIc determination. Another related issue is that ASTM E 1820-08 does not specify selec­tion of the test temperature to make sure that there is no cleavage fracture in the test. No posttest measures are required or recommended to confirm that the test data are not affected by cleavage initiation. The ques­tion arises: which of the reported KIc values should be size-adjusted consistently with Kjc data and which should not be size corrected?

Another important aspect is the 95% secant requirement (Pmax/PQ< 1.1), which limits stable crack growth to very small amounts for typical struc­tural steels exhibiting a rising tearing resistance curve.5 The methodology for KIc determination according to the ASTM E 1820 standard is thus not applicable for high-tearing resistance or high-toughness steels, which is also mentioned in Annex 5. The size require­ment for KIc is essential, since it means that the speci­men (ligament) size is the only factor which can be affected in pursuing a valid test result, if the test has to be performed at a specific temperature. If the secant requirement cannot be met, it is possible that no value of KIc can be determined at that temperature. For all of these reasons, fracture toughness testing in the tran­sition region is recommended to be made following standard ASTM E 1921, which takes into account the statistical nature of cleavage fracture.

Direct fracture toughness determination for the reactor vessel surveillance programs of nuclear power plants (NPPs) was one of the first applications of the

Master Curve methodology. Compared to the conven­tional Charpy V-notch methods, the Master Curve concept represents a new approach which makes pos­sible direct fracture toughness determination with only a few relatively small specimens, which is a more efficient use of limited test material. Using the Master Curve method allows statistical confidence to be applied to the directly measured data. The traditional practice of estimating fracture toughness from Charpy data (using correlations) is more unreliable due to large uncertainties associated with the correlations and the subsequent safety margins to meet regulatory require­ments. However, the Charpy-based methods are still in use and will continue to be used until a large amount of surveillance capsule Master Curve data is available. Existing data from correlations is discussed in Section 4.14.5, and material reference curves are discussed in Section 4.14.4.2. Use of these Charpy-based methods may remain to be the only way of estimating static fracture toughness in some cases, but is not the pre­ferred approach for future surveillance programs.