1.1 本试验方法包括使用静态或准静态加载条件下的预裂纹试样，在高温下测定金属中的蠕变裂纹萌生（CCI）和蠕变裂纹扩展（CCG）。本试验方法中提出的解决方案针对基材（即同质性能）和具有不均匀微观结构和蠕变性能的混合基材/焊缝材料进行了验证。CCI时间， t 0.2 ，即达到初始裂纹扩展δ所需的时间 一 一、 = 0.2 mm，从第一次施加力开始，CCG率， 一 ˙ 或 da/dt 表示为与断裂力学参数相关的蠕变裂纹扩展幅度， C *或 K 具有 C *定义为裂纹尖端应力的稳态测定，主要来自 C * （t） 和 C t ( 1- 17 ) . 2. 以这种方式导出的裂纹扩展被确定为可用于建模和寿命评估方法的材料特性 ( 17- 28 ) . 1.1.1 裂纹扩展相关参数的选择 C *, C * （t） , C t 或 K 取决于材料蠕变特性、试样的几何形状和尺寸。蠕变裂纹扩展试验中通常观察到两种类型的材料行为；蠕变延性 ( 1- 17 ) 蠕变脆性 ( 29- 44 ) . 在蠕变延性材料中，蠕变应变占主导地位，蠕变裂纹扩展伴随大量时间- 依赖于裂纹尖端的蠕变应变，裂纹扩展速率通过以下稳态定义进行关联: C t 或 C * （t） ，定义为 C *（参见 1.1.4 ). 在蠕变脆性材料中，蠕变裂纹扩展发生在低蠕变延性下。因此，随时间变化的蠕变应变与裂纹尖端局部的伴随弹性应变相当或受其支配。在这种稳态蠕变-脆性条件下， C t 或 K 可以选择作为相关参数 ( 8- 14 ) . 1.1.2 在任何一个试验中，可能存在两个裂纹扩展行为区域 ( 12 , 13 ) . 初始瞬态区域，弹性应变占主导地位，蠕变损伤发展，稳态区域，裂纹随时间成比例增长。 本标准涵盖稳态蠕变裂纹扩展速率行为。此外，还提出了具体建议 11.7 关于如何根据初始裂纹扩展期处理瞬态区域。在稳态期间，以下各项之间存在唯一的相关性: da/dt 以及合适的裂纹扩展速率相关参数。 1.1.3 在蠕变延性材料中，当整个未开裂韧带发生蠕变变形时，会发生广泛蠕变。这种条件不同于小规模蠕变和过渡蠕变条件 ( 1- 10 ) . 在大范围蠕变的情况下，与裂纹长度和未开裂韧带尺寸相比，蠕变变形主导的区域在尺寸上是显著的。 在小范围蠕变中，只有裂纹尖端局部未开裂韧带的一小部分经历蠕变变形。 1.1.4 扩展蠕变区的蠕变裂纹扩展速率与 C * （t） -积分。这个 C t 参数将小尺度蠕变和过渡蠕变区域中的蠕变裂纹扩展速率关联起来，并根据定义降低到 C * （t） 在广泛的蠕变区域 ( 5. ) . 因此，在本文件中 C *用作稳态扩展蠕变区的相关参数，而 C * （t） 和/或 C t 是描述小尺度蠕变、瞬态和稳态蠕变中瞬时应力状态的参数。 建议推导的函数 C *对于中所示的不同几何形状 附件A1 如中所述 附件A2 . 1.1.5 初始裂纹扩展尺寸δ的工程定义 一 一、 用于量化裂纹发展的初始阶段。该距离为0.2 mm。它已显示 ( 41- 44 ) 试验开始时存在的初始阶段可能是试验时间的一个重要阶段。在这一早期阶段，裂纹尖端经历了损伤发展以及达到稳定状态之前的应力重新分布。建议将该初始裂纹扩展期定义为: t 0.2 在δ处 一 一、 =0.2 mm，稳态 C *当裂纹尖端处于大范围蠕变和 K 对于蠕变-脆性条件。的值 C *和 K 应根据定义为的最终指定裂纹尺寸进行计算 一 o + δ 一 一、 哪里 一 o 是起始裂纹的初始尺寸。 1.1.6 CCI和CCG试验的推荐试样为标准紧凑拉伸试样C（T）（见 图A1.1 )在恒定载荷条件下，销承受张力。U形夹设置如所示 图A1.2 （参见 7.2.1 详细信息）。本程序中测试有效的其他几何形状如所示 图A1。 3. . 这些是张力CS（T）中的C型环、张力M（T）中的中间裂纹试样、单边切口张力传感器（T）、单边切口弯曲传感器（B）和双边切口张力传感器（T）。在里面 图A1.3 ，显示了用于测量力线位移（FLD）和裂纹口张开位移（CMOD）处位移的试样侧槽位置以及电位降（EPD）输入和输出引线的位置。拉伸试样的建议载荷为销载荷。配置和尺寸范围如所示 表A1.1 属于 附件A1 , ( 43- 47 ) . 样本选择将在中讨论 5. 9 . 1.1.7 裂纹尖端的应力状态可能会影响蠕变裂纹扩展行为，并可能导致平面试样中的裂纹前沿隧穿。试样尺寸、几何形状、裂纹长度、试验持续时间和蠕变特性将影响裂纹尖端的应力状态，是决定裂纹扩展速率的重要因素。试样及其侧槽的建议尺寸范围如所示 表A1.1 在里面 附件A1 . 已经表明，在该范围内，开裂率不会因材料和荷载条件的范围而变化 ( 43- 47 ) . 这表明，对于相对较短的测试持续时间（少于一年），约束水平在这些几何形状测试中观察到的正常数据分散范围内没有变化。 然而，建议在实验室施加的限制范围内，对不同几何形状、试样尺寸、尺寸和裂纹尺寸启动器进行试验。在所有情况下，应通过测试标准对上述数据进行比较 C（T） 如有可能，取样本。显然，通过测试上述更广泛的试样类型和条件，可以提高材料裂纹扩展数据的可信度。 1.1.8 材料不均匀性、残余应力和温度下的材料退化、试样几何形状和低力长时间试验（主要大于一年）可能会影响裂纹萌生和扩展特性的速率 ( 42- 50 ) . 在存在残余应力的情况下，当从具有残余应力场特征的材料或受损材料或两者中提取试样时，这种影响可能非常显著。例如，焊接件或厚铸件、锻造、挤压、部件、塑性弯曲部件和复杂部件形状，或其组合，其中完全消除应力是不可行的。从含有残余应力的此类部件上取下的试样也可能含有残余应力，这些残余应力的范围和分布可能因试样制造而改变。试样本身的提取部分缓解并重新分布了残余应力模式； 然而，除非进行焊后热处理（PWHT），否则剩余量仍可能在随后的试验中产生重大影响。否则，残余应力叠加在外加应力上，导致裂纹尖端应力强度不同于仅基于外力或位移的应力强度。不考虑拉伸残余应力效应将产生 C *低于预期的值有效地产生了相对于常数的更快开裂速率 C *. 这将产生用于寿命评估的保守估计和用于设计目的的非保守计算。 还应注意，试样加工过程中的变形也可能表明存在残余应力。 1.1.9 还应考虑蠕变和裂纹扩展引起的残余应力松弛。本标准中不包括处理这些变化的具体余量。然而，计算方法 C *本文介绍了使用试样的蠕变位移率来估计 C *固有地考虑了上述影响，如已测量的瞬时蠕变应变所反映的。然而，在分析这些类型的测试作为相关参数时，仍应格外小心 K 和 C *如所示 附件A2 尽管预计高温下的应力松弛可能部分抵消残余应力的影响。 附件A4 提供推导所需的正确计算 J 和 C* 对于需要考虑不匹配因素的焊接件测试。 1.1.10 样本配置和尺寸，不包括 表A1.1 在恒定力下进行测试将涉及进一步的有效性要求。这是通过比较来自推荐测试配置的数据来实现的。然而，如果将数据与从标准试样（如中所述）获得的数据进行比较，则该方法适用于使用其他几何形状 表A1.1 )并验证了合适的相关参数。 1.2 以国际单位制表示的数值应视为标准。括号中的英寸-磅单位仅供参考。 1.3 本标准并非旨在解决与其使用相关的所有安全问题（如有）。本标准的用户有责任在使用前制定适当的安全、健康和环境实践，并确定监管限制的适用性。 1.4 本国际标准是根据世界贸易组织技术性贸易壁垒（TBT）委员会发布的《关于制定国际标准、指南和建议的原则的决定》中确立的国际公认标准化原则制定的。 ====意义和用途====== 6.1 蠕变裂纹扩展速率表示为稳态的函数 C *或 K 表征材料在大范围蠕变变形或脆性蠕变条件下的裂纹扩展阻力。关于在蠕变裂纹扩展数据分析中采用断裂力学方法的基本原理的背景信息，请参见 ( 11 , 13 , 30- 35 ) . 6.2 高温下的侵蚀性环境会显著影响蠕变裂纹扩展行为。在研究和生成设计数据时，必须注意正确选择和控制温度和环境。 6.2.1 表示CCI时间， t 0.2 和CCG率， da/dt 作为适当断裂力学相关参数的函数，对于本文件中给出的几何形状和尺寸范围内的裂纹尖端相同应力状态，通常提供与试样尺寸和平面几何形状无关的结果（参见 附件A1 ). 因此，适当的相关性将能够交换和比较从各种样本配置和加载条件获得的数据。此外，该特性使蠕变裂纹扩展数据能够用于在高温下运行的工程结构的设计和评估，其中蠕变变形是一个问题。 假设了相似性的概念，这意味着不同尺寸的裂纹受到相同的标称应力 C * （t） , C t 或 K 如果满足特定裂纹扩展速率相关参数的有效性条件，则每单位时间的裂纹扩展增量相等。看见 11.7 详细信息。 6.2.2 试样尺寸、几何形状和材料延展性变化引起的裂纹尖端约束效应可能会影响 t 0.2 和 da/dt . 例如，在相同的 C * （t） , C t 在蠕变过程中，韧性材料通常随厚度的增加而增加。因此，在为实验室测试选择试样厚度、几何形状和尺寸时，有必要记住部件尺寸。 6.2.3 中提到的不同几何形状 1.1.6 对于获得与几何形状和尺寸无关的蠕变裂纹扩展速率数据，可能有不同的尺寸要求。因此，在比较时有必要考虑这些因素 da/dt 不同几何形状的数据或使用实验室数据预测部件寿命时的数据。由于这些原因，本标准的范围仅限于使用中所示的试样 附件A1 这些样本的验证标准见 11.7 . 然而，如果使用C（T）几何形状以外的试样生成蠕变裂纹扩展数据，则 da/dt 如有可能，应将获得的数据与标准C（T）测试得出的测试数据进行比较，以验证数据。 6.2.4 与同等条件下的速率相比，在试验开始时观察到蠕变裂纹以不同速率增长 C * （t） , C t 或 K 维持先前蠕变裂纹扩展的裂纹值 ( 12 , 13 ) . 该区域被识别为“尾部”。这种瞬态条件“尾”的持续时间随材料和最初施加的力水平而变化。这些瞬态是由于初始弹性载荷后裂纹尖端应力场的快速变化和/或由于初始阶段，在此期间，蠕变损伤区在裂纹尖端演化，并以自相似方式扩展，从而进一步扩展裂纹 ( 12 , 13 ) . 该区域与该周期后的稳态裂纹扩展分离，具有独特的特征 da/dt 对 C * （t） , C t 或 K 关系这种瞬态区域，尤其是在蠕变脆性材料中，可能在整个寿命的很大一部分中存在 ( 35 ) . 本标准提供了将该区域量化为初始裂纹扩展期的标准（见 1.1.5 )并将其与稳态裂纹扩展速率数据并行使用。看见 11.8.8 了解更多详细信息。 6.3 本试验方法的结果可用于以下方面: 6.3.1 使用分析和数值技术建立裂纹孕育期和扩展的预测模型 ( 18- 21 ) . 6.3.2 在可能发生蠕变变形的高温持续加载条件下，确定蠕变裂纹发展和扩展对剩余部件寿命的影响 ( 23- 28 ) . 注1: 对于这种情况，必须在代表性载荷和应力状态条件下生成实验数据，并结合适当的断裂或塑性坍塌标准、缺陷表征数据和应力分析信息。 6.3.3 建立材料选择标准和损伤容限应用的检查要求。 6.3.4 定量确定冶金、制造、工作温度和加载变量对蠕变裂纹扩展寿命的单独和综合影响。 6.4 从该试验方法中获得的结果是针对蠕变失效的裂纹主导状态设计的，不应适用于具有广泛蠕变损伤的结构中的裂纹，这有效地将裂纹扩展减少到集体损伤区域。允许在裂纹尖端周围的小区域内出现局部损伤，但不允许在与裂纹尺寸或剩余韧带尺寸相当的区域内出现局部损伤。此处的蠕变损伤由晶界空化的存在来定义。蠕变裂纹扩展主要由晶间时间相关裂纹的扩展来定义。如果试样尺寸选择错误，则可能出现裂纹尖端分支和裂纹扩展方向偏差- 开槽并形成几何形状（参见 8.3 ). 几何选择的标准在中进行了讨论 5.8 .

1.1 This test method covers the determination of creep crack initiation (CCI) and creep crack growth (CCG) in metals at elevated temperatures using pre-cracked specimens subjected to static or quasi-static loading conditions. The solutions presented in this test method are validated for base material (i.e. homogenous properties) and mixed base/weld material with inhomogeneous microstructures and creep properties. The CCI time, t 0.2 , which is the time required to reach an initial crack extension of δ a i = 0.2 mm to occur from the onset of first applied force, and CCG rate, a ˙ or da/dt are expressed in terms of the magnitude of creep crack growth correlated by fracture mechanics parameters, C * or K , with C * defined as the steady state determination of the crack tip stresses derived in principal from C * (t) and C t ( 1- 17 ) . 2 The crack growth derived in this manner is identified as a material property which can be used in modeling and life assessment methods ( 17- 28 ) . 1.1.1 The choice of the crack growth correlating parameter C *, C * (t) , C t , or K depends on the material creep properties, geometry and size of the specimen. Two types of material behavior are generally observed during creep crack growth tests; creep-ductile ( 1- 17 ) and creep-brittle ( 29- 44 ) . In creep ductile materials, where creep strains dominate and creep crack growth is accompanied by substantial time-dependent creep strains at the crack tip, the crack growth rate is correlated by the steady state definitions of C t or C * (t) , defined as C * (see 1.1.4 ). In creep-brittle materials, creep crack growth occurs at low creep ductility. Consequently, the time-dependent creep strains are comparable to or dominated by accompanying elastic strains local to the crack tip. Under such steady state creep-brittle conditions, C t or K could be chosen as the correlating parameter ( 8- 14 ) . 1.1.2 In any one test, two regions of crack growth behavior may be present ( 12 , 13 ) . The initial transient region where elastic strains dominate and creep damage develops and in the steady state region where crack grows proportionally to time. Steady-state creep crack growth rate behavior is covered by this standard. In addition, specific recommendations are made in 11.7 as to how the transient region should be treated in terms of an initial crack growth period. During steady state, a unique correlation exists between da/dt and the appropriate crack growth rate relating parameter. 1.1.3 In creep ductile materials, extensive creep occurs when the entire un-cracked ligament undergoes creep deformation. Such conditions are distinct from the conditions of small-scale creep and transition creep ( 1- 10 ) . In the case of extensive creep, the region dominated by creep deformation is significant in size in comparison to both the crack length and the uncracked ligament sizes. In small-scale-creep only a small region of the un-cracked ligament local to the crack tip experiences creep deformation. 1.1.4 The creep crack growth rate in the extensive creep region is correlated by the C * (t) -integral. The C t parameter correlates the creep crack growth rate in the small-scale creep and the transition creep regions and reduces, by definition, to C * (t) in the extensive creep region ( 5 ) . Hence in this document the definition C * is used as the relevant parameter in the steady state extensive creep regime whereas C * (t) and/or C t are the parameters which describe the instantaneous stress state from the small scale creep, transient and the steady state regimes in creep. The recommended functions to derive C * for the different geometries shown in Annex A1 is described in Annex A2 . 1.1.5 An engineering definition of an initial crack extension size δ a i is used in order to quantify the initial period of crack development. This distance is given as 0.2 mm. It has been shown ( 41- 44 ) that this initial period which exists at the start of the test could be a substantial period of the test time. During this early period the crack tip undergoes damage development as well as redistribution of stresses prior reaching steady state. Recommendation is made to correlate this initial crack growth period defined as t 0.2 at δ a i = 0.2 mm with the steady state C * when the crack tip is under extensive creep and with K for creep brittle conditions. The values for C * and K should be calculated at the final specified crack size defined as a o + δ a i where a o is initial size of the starter crack. 1.1.6 The recommended specimens for CCI and CCG testing is the standard compact tension specimen C(T) (see Fig. A1.1 ) which is pin-loaded in tension under constant loading conditions. The clevis setup is shown in Fig. A1.2 (see 7.2.1 for details). Additional geometries which are valid for testing in this procedure are shown in Fig. A1.3 . These are the C-ring in tension CS(T), middle crack specimen in tension M(T), single edge notched tension SEN(T), single edge notched bend SEN(B), and double edge notched tension DEN(T). In Fig. A1.3 , the specimens’ side-grooving-position for measuring displacement at the force-line displacement (FLD) and crack mouth opening displacement (CMOD) and positions for the electric potential drop (EPD) input and output leads are shown. Recommended loading for the tension specimens is pin-loading. The configurations, size range are given in Table A1.1 of Annex A1 , ( 43- 47 ) . Specimen selection will be discussed in 5.9 . 1.1.7 The state-of-stress at the crack tip may have an influence on the creep crack growth behavior and can cause crack-front tunneling in plane-sided specimens. Specimen size, geometry, crack length, test duration and creep properties will affect the state-of-stress at the crack tip and are important factors in determining crack growth rate. A recommended size range of test specimens and their side-grooving are given in Table A1.1 in Annex A1 . It has been shown that for this range the cracking rates do not vary for a range of materials and loading conditions ( 43- 47 ) . Suggesting that the level of constraint, for the relatively short term test durations (less than one year), does not vary within the range of normal data scatter observed in tests of these geometries. However, it is recommended that, within the limitations imposed on the laboratory, that tests are performed on different geometries, specimen size, dimensions and crack size starters. In all cases a comparison of the data from the above should be made by testing the standard C(T) specimen where possible. It is clear that increased confidence in the materials crack growth data can be produced by testing a wider range of specimen types and conditions as described above. 1.1.8 Material inhomogeneity, residual stresses and material degradation at temperature, specimen geometry and low-force long duration tests (mainly greater that one year) can influence the rate of crack initiation and growth properties ( 42- 50 ) . In cases where residual stresses exist, the effect can be significant when test specimens are taken from material that characteristically embodies residual stress fields or the damaged material, or both. For example, weldments, or thick cast, forged, extruded, components, plastically bent components and complex component shapes, or a combination thereof, where full stress relief is impractical. Specimens taken from such component that contain residual stresses may likewise contain residual stresses which may have altered in their extent and distribution due to specimen fabrication. Extraction of specimens in itself partially relieves and redistributes the residual stress pattern; however, the remaining magnitude could still cause significant effects in the ensuing test unless post-weld heat treatment (PWHT) is performed. Otherwise residual stresses are superimposed on applied stress and results in crack-tip stress intensity that is different from that based solely on externally applied forces or displacements. Not taking the tensile residual stress effect into account will produce C * values lower than expected effectively producing a faster cracking rate with respect to a constant C *. This would produce conservative estimates for life assessment and non-conservative calculations for design purposes. It should also be noted that distortion during specimen machining can also indicate the presence of residual stresses. 1.1.9 Stress relaxation of the residual stresses due to creep and crack extension should also be taken into consideration. No specific allowance is included in this standard for dealing with these variations. However the method of calculating C * presented in this document which used the specimen’s creep displacement rate to estimate C * inherently takes into account the effects described above as reflected by the instantaneous creep strains that have been measured. However extra caution should still be observed with the analysis of these types of tests as the correlating parameters K and C * shown in Annex A2 even though it is expected that stress relaxation at high temperatures could in part negate the effects due to residual stresses. Annex A4 presents the correct calculations needed to derive J and C* for weldment tests where a mis-match factor needs to be taken into account. 1.1.10 Specimen configurations and sizes other than those listed in Table A1.1 which are tested under constant force will involve further validity requirements. This is done by comparing data from recommended test configurations. Nevertheless, use of other geometries are applicable by this method provided data are compared to data obtained from standard specimens (as identified in Table A1.1 ) and the appropriate correlating parameters have been validated. 1.2 The values stated in SI units are to be regarded as the standard. The inch-pound units given in parentheses are for information only. 1.3 This standard does not purport to address all of the safety concerns, if any, associated with its use. It is the responsibility of the user of this standard to establish appropriate safety, health, and environmental practices and determine the applicability of regulatory limitations prior to use. 1.4 This international standard was developed in accordance with internationally recognized principles on standardization established in the Decision on Principles for the Development of International Standards, Guides and Recommendations issued by the World Trade Organization Technical Barriers to Trade (TBT) Committee. ====== Significance And Use ====== 6.1 Creep crack growth rate expressed as a function of the steady state C * or K characterizes the resistance of a material to crack growth under conditions of extensive creep deformation or under brittle creep conditions. Background information on the rationale for employing the fracture mechanics approach in the analyses of creep crack growth data is given in ( 11 , 13 , 30- 35 ) . 6.2 Aggressive environments at high temperatures can significantly affect the creep crack growth behavior. Attention must be given to the proper selection and control of temperature and environment in research studies and in generation of design data. 6.2.1 Expressing CCI time, t 0.2 and CCG rate, da/dt as a function of an appropriate fracture mechanics related parameter generally provides results that are independent of specimen size and planar geometry for the same stress state at the crack tip for the range of geometries and sizes presented in this document (see Annex A1 ). Thus, the appropriate correlation will enable exchange and comparison of data obtained from a variety of specimen configurations and loading conditions. Moreover, this feature enables creep crack growth data to be utilized in the design and evaluation of engineering structures operated at elevated temperatures where creep deformation is a concern. The concept of similitude is assumed, implying that cracks of differing sizes subjected to the same nominal C * (t) , C t , or K will advance by equal increments of crack extension per unit time, provided the conditions for the validity for the specific crack growth rate relating parameter are met. See 11.7 for details. 6.2.2 The effects of crack tip constraint arising from variations in specimen size, geometry and material ductility can influence t 0.2 and da/dt . For example, crack growth rates at the same value of C * (t) , C t in creep-ductile materials generally increases with increasing thickness. It is therefore necessary to keep the component dimensions in mind when selecting specimen thickness, geometry and size for laboratory testing. 6.2.3 Different geometries as mentioned in 1.1.6 may have different size requirements for obtaining geometry and size independent creep crack growth rate data. It is therefore necessary to account for these factors when comparing da/dt data for different geometries or when predicting component life using laboratory data. For these reasons, the scope of this standard is restricted to the use of specimens shown in Annex A1 and the validation criteria for these specimens are specified in 11.7 . However if specimens other than the C(T) geometry are used for generating creep crack growth data, then the da/dt data obtained should, if possible, be compared against test data derived from the standard C(T) tests in order to validate the data. 6.2.4 Creep cracks have been observed to grow at different rates at the beginning of tests compared with the rates at equivalent C * (t) , C t or K values for cracks that have sustained previous creep crack extension ( 12 , 13 ) . This region is identified as ‘tail’. The duration of this transient condition, ‘tail’, varies with material and initially applied force level. These transients are due to rapid changes in the crack tip stress fields after initial elastic loading and/or due to an initial period during which a creep damage zone evolves at the crack tip and propagates in a self-similar fashion with further crack extension ( 12 , 13 ) . This region is separated from the steady-state crack extension which follows this period and is characterized by a unique da/dt versus C * (t) , C t or K relationship. This transient region, especially in creep-brittle materials, can be present for a substantial fraction of the overall life ( 35 ) . Criteria are provided in this standard to quantify this region as an initial crack growth period (see 1.1.5 ) and to use it in parallel with the steady state crack growth rate data. See 11.8.8 for further details. 6.3 Results from this test method can be used as follows: 6.3.1 Establish predictive models for crack incubation periods and growth using analytical and numerical techniques ( 18- 21 ) . 6.3.2 Establish the influence of creep crack development and growth on remaining component life under conditions of sustained loading at elevated temperatures wherein creeps deformation might occur ( 23- 28 ) . Note 1: For such cases, the experimental data must be generated under representative loading and stress-state conditions and combined with appropriate fracture or plastic collapse criterion, defect characterization data, and stress analysis information. 6.3.3 Establish material selection criteria and inspection requirements for damage tolerant applications. 6.3.4 Establish, in quantitative terms, the individual and combined effects of metallurgical, fabrication, operating temperature, and loading variables on creep crack growth life. 6.4 The results obtained from this test method are designed for crack dominant regimes of creep failure and should not be applied to cracks in structures with wide-spread creep damage which effectively reduces the crack extension to a collective damage region. Localized damage in a small zone around the crack tip is permissible, but not in a zone that is comparable in size to the crack size or the remaining ligament size. Creep damage for the purposes here is defined by the presence of grain boundary cavitation. Creep crack growth is defined primarily by the growth of intergranular time-dependent cracks. Crack tip branching and deviation of the crack growth directions can occur if the wrong choice of specimen size, side-grooving and geometry is made (see 8.3 ). The criteria for geometry selection are discussed in 5.8 .