Residual stresses and ductile fracture

EDF有兴趣开始博士研究残余应力对延性断裂的影响，特别是对延性撕裂阻力高达约2mm的增长。这个问题已经使用了一些不同的方法在过去进行了研究，所以它是值得回顾的艺术状态简要定义一个新的项目之前。实验证据在这里，我们有极限载荷的问题。如果裂纹萌生发生在Lr= 1附近，那么很难确定残余应力的任何影响，因为预测的影响很小，并且被实验分散所掩盖。设计和进行小规模试验是相对简单的，在低Lr下发生断裂，在那里残余应力场中几乎没有塑性引起的松弛（例如见[1]）。也可以在Lr的范围内进行试验，以检验塑性的影响。与核工业直接相关的大多数韧性材料具有高的再塑性和相对适中的屈服强度，并且裂纹扩展阻力随着撕裂的进行而增加。工程钢在极限载荷下很难设计出不发生开裂和撕裂的小尺寸试验（例如，见[2]）。因此，残余应力的任何影响都被通常的实验分散所掩盖。STYLE项目对含有修补焊缝的Essete 1250管道进行了一次大型断裂试验，残余应力高到足以对裂纹萌生和撕裂的早期阶段产生可预见的影响，因为萌生发生在Lr<0.5时[3]。然而，这是一个单一的，复杂的，昂贵的测试，也许应该被视为一个验证基准，而不是一个科学研究的基础。雪莉和同事们已经通过使用模型材料来解决这个问题：低韧性铝合金，仍然会因微孔聚结而失效。这类材料允许使用小试样和标准试验机。引入残余应力的方法多种多样。对它们的大小和分布做出准确的估计通常并不简单，必须注意引入方法不会改变被测材料的性质（或者任何性质的变化都是很好理解的）。

EDF are interested in starting a PhD examining the impact of residual stresses on ductilefracture, specifically on ductile tearing resistance up to about 2mm of growth.This problem has been examined using a number of different approaches in the past, so it isworth reviewing the state of the art briefly before defining a new project.Experimental evidenceHere we have the problem of the limit load. If crack initiation occurs close to Lr=1then it is difficult to identify any impact of residual stress since the predicted effectsare both small and masked by experimental scatter.Brittle fracture occurs at low values of the crack driving force, and it is relativelystraightforward to design and perform small-scale tests where fracture occurs at lowLr where there is little plasticity-induced relaxation in the residual stress field (forexample see [1]). It is also possible to test over a range of Lr to examine the effectsof plasticity.Most ductile materials of direct relevance to the nuclear sector have high fracturetoughness and relatively modest yield strength, and the crack growth resistance risesas tearing proceeds. It is difficult to design small-scale tests that do not undergoinitiation and tearing at the limit load in engineering steels (for example, see [2]).Thus, any effects of residual stress are masked by the usual experimental scatter.The STYLE project performed a single large-scale fracture test on an Esshete 1250 pipecontaining a repair weld, where the residual stresses were high enough to have ameasurable effect on crack initiation and the early stages of tearing, since initiationoccurred at Lr<0.5 [3]. This however was a single, complex, expensive test, whichperhaps should be regarded as a validation benchmark rather than a basis for ascientific investigation.Sherry and co-workers have approached the problem by using a model material: alow toughness aluminium alloy that still fails by micro-void coalescence. This class ofmaterial allows the use of small specimens and standard testing machines.The methods used to introduce residual stresses are varied. Making accurateestimates of their magnitude and distribution is often not straightforward, and caremust be taken that the introduction method does not alter the properties of thematerial being tested (or that any property changes are well understood).