基于EBSD技术的P91钢蠕变过程中小角度晶界演化行为表征

在620℃、145 MPa条件下对给定的P91钢进行高温蠕变持久与间断试验,采用电子背散射衍射(EBSD)技术研究其在蠕变过程中小角度晶界的演化行为。通过引入EBSD图像中的取向差分布来表征小角度晶界处(0.5~5°)的边界位错密度,分析了边界位错密度在蠕变过程中与小角度边界的数量、塑性应变以及内部微观组织演化之间的关系。此外,通过改变EBSD像素点与像素点之间的计算步长,探讨了步长选择对边界位错密度计算结果的影响。结果表明,小角度晶界处的位错密度在蠕变过程中先迅速上升,在最小蠕变率处达到极值后缓慢下降,直到最后基本保持不变;同时,EBSD的计算步长越小,得到的位错密度值越准确。     

一种含Re/Ru镍基单晶合金的蠕变行为及其机理

针对含Re/Ru单晶合金共晶组织中的高熔点元素和高推重比发动机材料的高承温能力等问题,研究了固溶温度和组织演化对其高温蠕变行为的影响。结果表明,在1328℃固溶处理的合金中仍存在粒状残余共晶,在1332℃固溶处理可将其消除,使合金在1100℃/140 MPa的蠕变寿命由321 h延长到476 h。在蠕变初期的合金中γ′相转变为N-型筏状组织,在蠕变后期位错剪切进入γ′相形成的位错锁,可抑制位错运动,使合金的蠕变抗力提高。但是高应变速率下位错的交替滑移使筏状γ′相扭折和形成亚晶,可降低合金的蠕变抗力,特别是筏状γ′相转变成“类块状”形态的逆向组织演化提高了应变速率,是合金高温蠕变较后期间的变形和损伤特征。其中,合金在1140℃寿命的大幅度降低,归因于γ′相的溶解使其尺寸减小和体积分数降低。     

超超临界机组主蒸汽取样管Ⅳ型开裂分析

通过宏观和微观分析、硬度测试和管系应力计算分析了某1000 MW超超临界机组主蒸汽取样管泄漏的原因。结果表明:取样管泄漏性质为焊接热影响区Ⅳ型蠕变开裂。其早期失效的主要原因是管系布置的柔性不足导致二次应力过高。提出对T/P91等马氏体钢管道,需重视支吊架的合理设置和维护检修,以防管道发生早期失效。建议对高合金钢焊接接头进行寿命评估时除考虑等效蠕变应力之外,还需考虑应力三轴因子,以提高焊接接头蠕变寿命评估的精度。     

一种含4.5%Re/3.0%Ru的单晶镍基合金的高温蠕变行为

通过对含4.5%Re/3.0%Ru单晶镍基合金进行高温蠕变性能测试,并采用扫描电镜(SEM)、透射电镜(TEM)对不同蠕变期间的试样进行组织形貌观察,研究了该合金的高温蠕变行为。结果表明,本实验所选用的单晶合金在高温蠕变期间具有良好的蠕变抗力,在1040℃/160MPa的蠕变寿命达到725h。高温蠕变初期,合金中γ′相沿垂直于应力轴方向转变成筏状结构,其稳态蠕变期间的变形机制是位错在基体中滑移和攀移越过筏状γ′相。高温蠕变后期,合金的变形机制是位错在基体中滑移和剪切筏状γ′相。位错的交替滑移使筏形γ′相扭曲,并在γ/γ′两相界面发生裂纹的萌生与扩展直至断裂,是合金在高温蠕变后期的断裂机制。     

添加气膜孔对镍基单晶合金DD6蠕变寿命的影响

本研究在外加应力为200 MPa、温度为1 000℃的条件下,对镍基单晶合金DD6带气膜孔试样(飞秒激光环形扫描与旋切扫描相结合加工而成)和不带气膜孔试样进行了蠕变实验。实验结果显示,带气膜孔试样的蠕变寿命显著短于不带气膜孔试样的蠕变寿命。借助扫描电镜和透射电镜对试样蠕变前、后的微观结构进行观测发现:不带气膜孔试样的变形机理为位错在塑性比较好的基体相内(γ相)均匀滑移;而带气膜孔试样由于变形主要集中在孔周围,位错大量增殖交叉,从而在γ/γ'两相界面处产生极大的应力集中,直接将强化相即脆硬相(γ'相)剪切贯穿,导致裂纹快速扩展,合金的蠕变寿命显著缩短。     

一种高W单晶镍基合金的组织演化与蠕变行为

通过蠕变曲线测定及组织形貌观察,研究了一种9％W单晶镍基合金的组织演化与蠕变行为.结果表明,当在1040℃施加应力大于160MPa时,合金表现出明显的施加应力敏感性,在试验的温度范围内,合金具有较低的应变速率和较长的蠕变寿命,测定出合金的表观蠕变激活能为465kJ/mol.在蠕变初期,合金中γ'相沿垂直于应力轴方向,转变成N-型筏状结构,在稳态蠕变期间合金的变形机制是位错攀移越过筏状γ'相,而蠕变后期合金的变形机制是位错剪切进入筏状γ'相.蠕变断裂后,在试样不同区域筏状γ'相具有不同的形貌,在远离断口区域,筏状γ'相与应力轴方向垂直,而在近断口区域筏状γ'相尺寸及扭曲程度增加的原因是承受的应力及变形程度增大所致.     

物理冶金信息指导机器学习的镍基单晶高温合金蠕变寿命预测

蠕变寿命是影响镍基单晶高温合金材料服役寿命和力学性能的关键材料参数。因此,如何准确有效地预测合金的蠕变寿命具有重要现实意义。尽管多年来许多研究学者已经建立起多种蠕变寿命的预测模型,但是由于不同温度应力下的蠕变机制复杂且蠕变过程涉及长时间的显微组织演化,已有模型尚难以实现有效预测。对此,采用物理冶金原理指导下的数据挖掘结合机器学习这一研究策略,通过文献调研建立起了高温低应力下的镍基单晶合金的高质量蠕变数据集,在物理冶金原理指导下对原始数据进行挖掘,提高了原始数据的内在质量,并基于Pearson系数和随机森林平均精确度降低值分别对原始数据特征进行了相关性分析和重要性评估,表明所建立的数据集符合基本的物理冶金学机制,同时阐明了引入的三维物理冶金信息对于蠕变寿命预测的重要意义。随后,基于机器学习方法在数据挖掘后的数据集上对合金的蠕变寿命进行了预测,并根据平方相关系数(R²)、平均绝对误差(MAE)和过拟合程度评估了不同的机器学习模型。结果表明,支持向量回归(SVR)模型在本研究中具有较好的泛化能力且不容易过拟合,同时结合了物理冶金信息的机器学习模型拥有更好的预测准确性和泛...     

[110]取向DD3镍基单晶合金的组织结构及蠕变行为

通过蠕变曲线的测定及微观组织观察,研究了[110]取向镍基单晶合金的组织结构与蠕变行为。结果表明,经完全热处理后,合金中立方γ′相沿〈100〉取向规则排列;蠕变期间合金中形成筏状γ′相的取向与应力轴方向成45°角,蠕变后期在近断口区域筏状γ′相发生扭折。在1040℃、137MPa条件下,合金在稳态蠕变期间具有较高的应变速率和较短的蠕变寿命,而蠕变期间的变形特征是位错在γ基体通道中滑移和剪切筏状γ′相;其中,γ′相形成筏状结构后,沿与应力轴成45°角的基体通道承受最大剪切应力,使蠕变位错易于在基体通道中滑移,是使合金具有较大应变速率的主要原因。     

一种GH4169镍基合金的组织结构与蠕变性能

通过对一种等温锻造GH4169镍基合金进行直接时效处理,蠕变性能测试及组织形貌观察,研究了该合金的组织结构与蠕变行为。结果表明,GH4169合金的组织结构由γ基体,γ′相、γ″相和δ相组成,且各相之间保持共格界面。测定出合金在660℃/700MPa条件下的蠕变寿命为123h。合金在680℃/700MPa的蠕变寿命为39h,在实验温度和应力范围内,计算出直接时效合金的蠕变激活能为588.0kJ/mol。合金在蠕变期间的变形机制是位错滑移和孪晶变形,其中,沿晶界析出的粒状碳化物,可抑制晶界滑移,是使合金具有较好蠕变抗力的主要原因。随蠕变进行,开动的滑移系中位错运动至晶界受阻,并塞积于该区域引起应力集中,当应力集中值大于晶界的结合强度时,可促使其在与应力轴垂直的晶界处发生裂纹的萌生与扩展,直至断裂,是合金在蠕变期间的断裂机制。     

基于Norton-Bailey模型的P92钢初期蠕变过程分析

以Norton-Bailey时间硬化模型和Kachanov损伤模型为基础,提出一种能够全面描述蠕变全寿命期蠕变变形发展的本构模型。与Norton-Bailey模型相比,该模型可以更准确地反映蠕变第一阶段的变形累积,以Kachanov损伤演化方程描述蠕变第三阶段的变化;推导模型第一阶段和第三阶段的参数确定方法;根据实验数据,确定P92钢的模型参数。将该模型嵌入到在ANSYS接口程序中,并用于高温合金钢P92的蠕变寿命计算,结果表明模型与实验数据基本吻合。     

Microstructure Evolution of a 10Cr Heat-Resistant Steel during High Temperature Creep

The microstructure evolution of a 10Cr ferritic/martensitic heat-resistant steel during creep at 600℃ was investigated in this work.Creep tests demonstrated that the 10Cr steel had higher creep strength than conventional ASME-P92 steel at 600℃.The microstructure after creep was studied by transmission electron microscopy,scanning electron microscopy and electron probe microanalysis.It was revealed that the martensitic laths were coarsened with time and eventually developed into subgrains after 8354 h.Laves phase was observed to grow and cluster along the prior austenite grain boundaries during creep and caused the fluctuation of solution and precipitation strengthening effects,which was responsible for the two slope changes on the creep rupture strength vs rupture time curve.It was also revealed that the microstructure evolution could be accelerated by stress,which resulted in the lower hardness in the deformed part of the creep specimen,compared with the aging part.     

Influence of aluminium on creep strength of 9–12% Cr steels

The influence of aluminium on creep strength of 9% Cr steels is predicted by a fundamental model for creep. Through thermodynamic modelling the particle structure is determined for a temperature and composition range. This shows how AlN is formed at the expense of MX carbonitrides of VN character when the aluminium content is increased. The remaining MX particles are of NbC type and have approximately one fifth of the original phase fraction. The evolution in microstructure such as particle coarsening is included in the model as well as the recovery. Rupture time is predicted using a modified Norton equation including back-stresses calculated from microstructure. The predictions show correspondence to some of the creep data for the steel P91 over a temperature and stress range. Furthermore, simulation with high Al content verifies the observed early failure of Al rich components. Overall, the simulations show a decrease in rupture time by a factor 6 due to Al additions of up to 0.2%.     

Evaluation of long-term creep damage in high Cr ferritic steel welds

Abstract

The creep strength of high Cr ferritic heat-resistant steel welds decreases due to the formation of Type IV creep damage in the heat affected zone (HAZ) during long-term service at high temperatures. In order to elucidate the processes of Type IV creep damage, creep rupture and creep interruption tests using ASME Gr.91 and Gr.122 steel welds were conducted. It was found that creep voids formed at an early stage (0.2 of life) and coalesced to form a macro crack at 0.8 of life for the Gr.91 steel weld. On the other hand, for the Gr.122 steel weld, a small number of Type IV creep voids formed at 0.5 of life, increased slightly until 0.9 of life and rapid crack growth occurred after that. Differences of creep damage behaviour between the Gr.91 and Gr.122 steel welds are discussed. The Type IV creep damage distributions obtained were compared with analytical results using the finite element method and damage mechanics.     

Study on Laves phase in an advanced heat-resistant steel

The Laves phase is one of the most significant precipitates in ferritic/martensitic heat-resistant steels. Laves phase precipitates in the creep rupture specimens with different rupture life were studied on a 10 wt.% Cr heat-resistant steel. JMatPro thermodynamic and kinetic calculations were carried out to simulate and predict the precipitation behavior of the Laves phase in the steel at the equilibrium state. The morphologies of the Laves phase developed with creep time were characterized under both scanning electron microscope (SEM) and transmission electron microscope (TEM). Effects of Co on the growth behavior of Laves phase and the corresponding fracture mode were analyzed. It was found that the Laves phase in the steel grew to 200 nm in size after only 1598 h at 600°C, indicating that the addition of Co in the steel could accelerate the growth of Laves phase, and the coalescence of large Laves phase would lead to the brittle intergranular fracture.     

Study on Laves phase in an advanced heat-resistant steel

The Laves phase is one of the most significant precipitates in ferritic/martensitic heat-resistant steels. Laves phase precipitates in the creep rupture specimens with different rupture life were studied on a 10 wt.% Cr heat-resistant steel. JMatPro thermodynamic and kinetic calculations were carried out to simulate and predict the precipitation behavior of the Laves phase in the steel at the equilibrium state. The morphologies of the Laves phase developed with creep time were characterized under both scanning electron microscope (SEM) and transmission electron microscope (TEM). Effects of Co on the growth behavior of Laves phase and the corresponding fracture mode were analyzed. It was found that the Laves phase in the steel grew to 200 nm in size after only 1598 h at 600°C, indicating that the addition of Co in the steel could accelerate the growth of Laves phase, and the coalescence of large Laves phase would lead to the brittle intergranular fracture.     

Long-term creep strength predictions from short-term creep test data for high Cr creep-resistant steels and microstructural evolution origin of over-predictions

A new tensile creep model that integrates the tensile strength at creep temperature is investigated for its generic applicability in predicting the long-term creep strengths from short-term creep test data for high Cr creep-resistant steels using creep and tensile strength data measured for a grade of 11Cr steel. The results show that, when the long-term creep strengths are specified by stresses producing the required minimum creep rate, they can be accurately predicted using short-term creep test data. However, when they are specified by stresses giving the required creep rupture time, using only short-term creep test data will lead to over-predictions. The microstructure evolution origin of such over-predictions is traced to the Z-phase precipitation during creep in creep-resistant steels with more than 9 wt.% Cr. The conventional concept on the relationship between creep test stress and creep mechanisms is also re-evaluated in light of the new results.     

Microstructural Evolution on the T91 Dissimilar Metal Joints during Creep Rupture Tests

T91 steel is one of the new materials presently employed in power plant pipe components. The creep rupture strength and microstructure of the T91+10CrMo910 and T91+13CrMo44 welded joints were analyzed during creep rupture tests. Creep transgranular ductile rupture occurred at the 10CrMo910 matrix in the T91+10CrMo910 welded joints and creep intergranular brittle rupture occurred at the 13CrMo44 HAZ in the T91+13CrMo44 joints. Microhardness measurements showed high hardness at the heat affected zone (HAZ) of T91 and a sharply drop at the 13CrMo44 HAZ during creep rupture. The metallographic tests showed that no obvious microstructure degradation was observed in the 10CrMo910 HAZ and matrix, while creep cracks appeared at the 13CrMo44 HAZ. T91 steel had relatively high creep resistant strength in the welded joints tested. Recovery occurred in the T91 HAZ with the growth of subgrain size and the decrease of dislocation density during creep. It was concluded that the dissimilar joints of T91 and low alloy h     

Indentation creep behavior of AE42 and Ca-containing AE41 alloys

The indentation creep behavior of AE42 and 0.4–1.2 wt.% Ca-containing AE41 alloys was studied. The microstructure was analyzed by an optical microscope, XRD and SEM equipped EDS before and after indentation creep. The results indicate that the creep resistance of AE41 alloys is improved with the addition of Ca. The indentation creep resistance of Ca-containing AE41 alloys is better than that of AE42 at 150 °C and 175 °C. The microstructure analysis shows that the Al₁₁Nd₃ phase in AE42 is prone to decompose above 150 °C, which deteriorates its creep resistance behavior. Ca-containing AE41 alloys have good indentation creep resistance because of the formation of heat-resistant phase Al₂Ca in the alloys.     

Creep rupture behaviour of modified 9Cr-1Mo heat-resistant steel strengthened with different mechanisms

The creep rupture behaviours and microstructural changes of a modified 9Cr-1Mo heat-resistant steel were investigated at 853 K. Analysis of creep results suggests that dislocation climb is the dominant deformation mechanism with true stress exponent of 5 under the present conditions. Based on the microstructural analysis, strengthening contributions from M₂₃C₆ carbides and MX carbonitrides were clarified. The M₂₃C₆ carbides can promote grain boundary strengthening by exerting Zener pinning forces, whereas MX carbonitrides can enhance the creep strength by interacting with mobile dislocations to induce threshold stress. Besides, softening of the steel is related not only to the decrease of dislocations, but also the coarsening of precipitates and substructures. The value of creep damage tolerance factor is close to 6.6, which further confirms that the creep damage is mainly attributed to the microstructural degradations, such as the coarsening of precipitates and substructures and decrease of dislocations.     

A model based creep equation for 9Cr–1Mo–0·2V (P91 type) steel

Abstract

The isothermal constant stress creep tests data for a 9Cr–1Mo–0·2V (P91 type) steel were submitted for a phenomenological analysis in order to obtain the relevant creep equation for such steel. Namely, the minimum creep strain rate of P91 type steel cannot be described by the simple Arrhenius type power law constitutive model. The incorporation of the threshold stress concept in the analysis of creep data leads to a modified power law, which satisfactorily describes the creep behaviour of the examined P91 steel. However, the threshold stress is not a good material parameter, as it often varies with temperature and/or applied stress. This adds uncertainty to the extrapolation of the creep rates into ranges where experimental data are not available. Besides the fact that the physical foundation for a threshold stress is questionable from a scientific point of view, this is a serious practical limitation of the modified power law creep equation. The second creep equation proposed in the present paper is the improved stress dependent energy barrier model. The improvement of the standard model is based on two assumptions: first, on the hypothesis that the application of a stress also affects the energy barrier to be overcome when a local region transitions from the initial to the final state, and second, by applying a simple power function of stress instead of a hyperbolic sin function in the model based equation. The obtained value of stress exponent, n=5·5, is too high for entirely climb controlled creep. The apparent activation energy of approximately 510 to 545 kJ mol^?1, which is considerably higher than the activation energy for lattice diffusion, is the stress dependent activation energy of the slowest, dominant rate controlling process of the supposed multiple creep mechanisms.