Improved approach for predicting weld creep strength factors of ferritic steels

Abstract

The cross-weld (CW) creep strength of ferritic steels is typically lower than that for parent metal (PM), and in the past the ratio of CW to PM creep strength (weld strength factor – WSF) was assumed to be limited to ～80%. For newer Cr steels WSF can be significantly lower for a typical design life of 100 000 h or more. The possibility of low WSF is also accommodated in the current design codes such as EN 13445, but no suggested WSF values are given for guidance. Assuming a too high WSF for such welds obviously results in an unsafe (too long) predicted creep life. Unfortunately, as a further complication the WSF of the newer Cr steels can decrease when the operating temperatures are increased for improved efficiency of future power plants. It is hence important that reliable and sufficiently high values of WSF can be guaranteed. However, there is often much less extensive data on the creep strength of welds than on parent steel, and also the extrapolation to long term values of WSF can add more relative uncertainty than what is expected in extrapolating the long term creep strength of parent steel. Here an improved approach is proposed to predict WSF using the Wilshire creep model to obtain the relationship between the CW creep strength and the corresponding parent material (PM) strength. The Wilshire model directly provides the WSF value for each CW data point, when the expected normalised stress is based on the CW time to rupture at stress and temperature. The corresponding master curve parameters are those for PM, when the PM hot tensile strength is also known. The WSF data points for each CW test can then be fitted for temperature and stress dependence. This approach avoids fitting distortion in WSF, unlike the traditional assessment where a master curve is first obtained for the CW creep strength. As an example, WSF of welded P91 steel at 100 000 h is here predicted in the temperature range of 550–650°C.     

Cr concentration dependence of overestimation of long term creep life in strength enhanced high Cr ferritic steels

Creep rupture data and microstructural degradation during aging of high Cr ferritic boiler steels with enhanced creep strength have been studied with special attention to prediction of long term creep rupture life. Tempered lath martensite structure in the high Cr ferritic steels remains unchanged during short term aging, whereas static recovery of the lath martensite structure proceeds when diffusion distance during aging becomes sufficiently long as is the case in long term creep. The static recovery brings about premature failure in long term creep and decreases in apparent activation energy for creep life. The decrease in activation energy is responsible for overestimation of rupture life reported in strength enhanced high Cr ferritic steels. The boundary from a short term region with high activation energy Q_H to a long term region with low activation energy Q_L moves towards longer time with decreasing Cr concentration. The difference in activation energy (Q_H − Q_L) primarily determines the extent of overestimation of rupture life predicted from short term data. In general, the extent of overestimation is less serious at 9%Cr as compared to 12%Cr.     

Heat-to-heat variation in long-term creep strength of some ferritic steels

Heat-to-heat variation in creep life has been investigated for some ferritic steels, mainly 12Cr steels, using long-term creep data in the NIMS Creep Data Sheets. The tempered martensitic plain 12Cr and 12Cr–1Mo–1W–0.3V steels exhibit large heat-to-heat variation in creep life, as shown by about one order of magnitude difference in time to rupture or more between the strongest and weakest heats. On the other hand, low-Cr steels of tempered bainitic 1Cr–1Mo–0.25V, ferritic–pearlitic 2.25Cr–1Mo and ferritic 9Cr–1Mo steels exhibit small heat-to-heat variation in creep life. The heat-to-heat variation in long-term creep strength is correlated with the degradation behaviour at long times, which depends on initial strength and concentrations of Al, nitrogen and Cr. The present results suggest that taking the mechanisms responsible for the heat-to-heat variation in creep life into account, quality of heat resistant steels as well as reliability of remaining life estimation can be further improved.     

Effect of specimen size and shape on creep rupture behavior of creep strength enhanced ferritic steel welds

The creep strength enhanced ferritic (CSEF) steels such as Grades 91, 92, 122, 911, 23 and 24 have become very important key materials for high efficiency fossil-fired power plants for last decades, however the long-term creep rupture strength and strength reduction in welds due to Type IV failure of these steels are serious problem to be urgently resolved. In order to use CSEF steel welds safely setting new weld strength reduction factors have internationally discussed. For instance ASME Boiler and Pressure Vessel Code Committee recently developed a new creep strength reduction factors for CSEF steels intercritically post-weld heat treated to be 0.5 based upon the creep rupture data obtained for standard creep specimens. However it is needed to make further consideration on the specimen size/shape effect on the creep strength of the welds to determine more appropriate weld strength reduction factors and joint influence factors. Present report provides comprehensive creep rupture test results of the specimens with various size and shape, and full size components dealing with creep rupture location/behavior and specimen size/shape effect on the creep rupture strength of CSEF steel welds.     

Hardness–fracture toughness–creep rupture strength relationship for welds in P91 main steam pipe

Abstract

An experimental study on the impact toughness, hardness, fracture toughness and creep rupture strength of P91 steel weld metals under two conditions (R and D) has been carried out. The results show that weld metal having higher hardness, lower impact toughness and lower fracture toughness (sample R) has higher creep rupture strength. The creep rupture strength is closely related to creep life of component. Thus, assessment of the weld hardness of P91 steel is of value in assessing the creep life of pipes in service. The lower toughness of sample R is attributed mainly to the higher welding current and input line energy used. The lower hardness and strength of sample D, which received a second post-weld heat treatment, partially result from too high a secondary tempering temperature being employed.     

一种预测TP347H钢长时持久强度的方法

基于多区LMP法,利用TP347H钢的高温（700℃、750℃）短时（≤5×10^3h）试验数据预测其长时持久强度（600～750℃,5×10^3～2×10^5h）．结果表明：在应力-时间图中根据口／dT。的特定比值对应力分区,用低应力-长时区LMP参数中的C值计算长时持久强度,其预测值与真实值吻合良好;与传统的单区LMP法相比,多区LMP法的应用不仅显著降低了持久强度的过估倾向而且大大缩短了试验时间,为这类钢的长时持久强度评估提供了准确而有效的方法．     

3D creep cavitation characteristics and residual life assessment in high temperature steels: a critical review

Abstract

The need for a new paradigm to estimate remaining creep life of service exposed steels is critically assessed. New approaches to residual life assessment are proposed, in the light of a decade’s experience of the use of micro-tomography to characterise the three-dimensional (3D) nature of cavitation damage in structural materials. Imaging of conventional structural materials such as steels with high absorption to X-rays has been realised by synchrotron micro-tomography (SR-μCT), providing new insights into phenomena such as creep failure. The unique feature of SR-μCT studies is the direct imaging in 3D of cavities (hundreds of micrometres in size) present in the bulk, revealing the spatial characteristics and morphology of the creep voids. Quantitative analyses of the cavitation characteristics revealed by 3D datasets, when scaled with respect to time, stress and temperature, provide functional information suitable for developing constitutive equations for creep. The application of SR-μCT, a non-destructive technique providing high fidelity data, significantly reduces the ambiguity in developing functional relationships to predict creep failure. The explicit use of such constitutive equations to estimate the residual life of components in creep, and the consequent assessment of structural integrity, would prove invaluable. Micro-tomography studies related to creep in materials are reviewed, with special emphasis on a 10·86%Cr heat resistant steel, to demonstrate the type of data available for life assessment and design against creep failure. A brief discussion of current methods to estimate residual life in the light of recent 3D micro-tomography data follows. Finally, the possibility of new approaches, using micro-tomography data in conjunction with destructive 3D approaches such as serial sectioning, to formulate advanced residual life estimates, is briefly considered.     

Analysis of cross-weld creep rupture data on the ½Cr½Mo¼V type IV zone

For the ferritic steel ½Cr½Mo¼V, the operating experience of medium and long term creep failures in power station welds has been of cavitation and cracking in the heat affected zone (HAZ) adjacent to the parent material, in what is referred to as the Type IV zone. This experience has led to the generation of uniaxial cross-weld creep rupture data on ½CrMoV weldments and to the development of life assessment procedures such as R5 Volume 7, which uses the rupture strength of the Type IV zone to calculate the life of power station components. Recently, the ECCC Working Group 3A has conducted a collation of the UK and German cross-weld creep rupture data on ½CrMoV weldments. Creep failure in ½CrMoV cross-weld specimens occurs in a variety of weldment zones; typically in the parent, the Type IV zone or the weld metal. Post test examination of the specimens has enabled those tests that failed in the Type IV zone to be identified and a creep rupture data assessment has been performed to derive a new model for the rupture strength of the Type IV zone.     

Modelling and experimental studies of alternative heat treatments in Steel 92 to optimise long term stress rupture properties

Abstract

The desire for power plant to give increased generating efficiency and decreased CO₂ emission has led to considerable effort over the last 10–15 years, to develop ferritic–martensitic steels which can be used for steam temperatures up to about 650°C. Examples are the addition of boron and increasing chromium content to 10–12 wt-%. However, high chromium levels have led to problems with long term precipitate stability. One approach which has not been widely explored, is the use of novel heat treatments to optimise the preservice microstructure to give the best long term creep rupture strength. Increased austenitising temperatures and lower tempering temperatures have been examined in Steel 92 (9Cr–0·5Mo–2W) and have produced significant improvements in creep rupture strength at temperatures up to 650°C compared with material given a conventional heat treatment. This has been achieved without any loss in ductility compared with conventional heat treatments. Test data for durations in excess of 40 000 h are presented. Modelling of microstructure evolution based on Monte Carlo simulations has shown important differences especially in the stability of grain boundary M₂₃C₆ and intragranular MX particles, between material with conventional and modified heat treatments. The model predictions are in good agreement with metallographic observations made on material before and after stress rupture testing. Continuum creep damage mechanics modelling based on the microstructural evolution has also been applied to predict creep life of Steel 92 and satisfactory agreement with creep rupture tests has been obtained.     

Creep oxidation behaviour and creep strength prediction for Alloy 617

Creep experimental data was obtained by a series of creep tests with different stress levels at 950 °C for Alloy 617. Oxidation behaviour was investigated by observing the microstructures of fractured specimens after the creep tests. Oxidation thickness was measured quantitatively with the creep rupture times, and the oxidation microstructures were represented by a SEM image. In addition, the long-term creep strength for Alloy 617 was predicted by using a multi-constant method with two C instead of the conventional one with a unique C in the Larson-Miller (LM) parameter. For 10⁵ h at 950 °C, the creep strength for the conventional method was 7.2 MPa, but for the multi-constant method it was reduced to 4.7 MPa. The conventional method did not thoroughly match with the creep rupture data, and revealed an overestimation for the prediction of the long-term creep strength. On the other hand, the multi-constant method revealed a good agreement with the creep rupture data, and its method was thus more accurate than the conventional one. This multi-constant analysis can be used to accurately predict the long-term creep rupture of Alloy 617.     

Finite-element creep damage analyses of P91 pipes

In this paper, uniaxial and notched bar creep test data are used to establish the material behaviour models for two P91 steels of differing strength. The two steels are denoted here as Bar 257 steel, tested at 650 °C and A-369 steel, tested at 625 °C. Single-state variable and three-state variable creep damage constitutive models were used in the investigation.Methods for determining the material properties in the two sets of equations are briefly described. Finite-element analyses are performed using these material properties for a P91 pipe, subjected to internal pressure and end loading. The failure lives of the pipe were obtained, and on this basis, a preliminary assessment of using the two different sets of constitutive equations for failure predictions of high-temperature components under creep damage conditions can be made.     

Development of low thermal expansion nickel base superalloy for steam turbine applications

Abstract

The target operating temperature of ultrasupercritical power plants is increasing and is planned to reach 700°C. Austenitic superalloys are promising materials for these applications to replace ferritic heat resistant steels, because of their high strength at 650–700°C. In general, austenitic nickel base superalloys show higher creep rupture strength than ferritic heat resistant steels; however, they have higher coefficients of thermal expansion, lower creep rupture ductilities, and higher costs. The effect of the Mo and Co content, amount of γ' phase, and Al/Ti ratio in the γ' phase on the thermal expansion behaviour of a Mo containing superalloy has been investigated by use of the conventional Mo containing Alloy 252 as a reference. Tensile and creep rupture properties were also measured. Following a modified heat treatment, the Co free superalloy developed on the basis of these tests showed higher creep rupture ductility than Alloy 252, while retaining comparable low thermal expansion and high creep rupture strength. Creep rupture properties at 700°C for up to 20 000 h were satisfactory, suggesting that the alloy is suitable for long term applications. Initial assessments of the weldability and mechanical properties of weld joints at 750°C are encouraging for boiler tube applications.     

New ECCC assessment of creep rupture strength for steel grade X10CrMoVNb9-1 (Grade 91)

A first assessment of creep rupture strength for steel grade X10CrMoVNb9-1 (Grade 91) was performed by ECCC in 1995. The results were included in the European standard EN 10216. Due to a significant increase of test data and test duration it was decided in 2005 to make a re-assessment of the extended database. Different procedures have been used independently by different assessors. The method with the best overall fit of the data set has found to be the ISO CRD method. This is characterized by a two steps procedure: in the first step the mean isotherms are evaluated from the test data, afterwards the evaluated isotherms are used for averaging by a Manson–Haferd master-curve. The results have been chosen as the basis to specify long term creep rupture strength values in a new ECCC data sheet for X10CrMoVNb9-1 (Grade 91).     

Creep property of carbon and nitrogen free high strength new alloys

The carbon and nitrogen free new alloys which were composed of supersaturated martensitic microstructure with high dislocation density before the creep test have been investigated systematically. These alloys were produced from the new approach which raised creep strength by the utilization of the reverse transformed austenite phase as a matrix and intermetallic compounds such as Laves phase and mu-phase as precipitates during heating before the creep test. It is important that these alloys are independent of any carbides and nitrides as strengthening factors. The high temperature creep test over 700 °C exceeds 50,000 h, and the test is continuous. Creep behavior of the alloys is found to be different from that of the conventional high-Cr ferritic steels. The addition of boron to the alloy pulled the recrystallization temperature up in the high temperature, and it became a creep test in the un-recrystallization condition, and the creep property of high temperature over 700 °C was drastically improved. The minimum creep rates of Fe–Ni alloys at 700 °C are found to be much lower than those of the conventional high Cr ferritic heat resistant steels, which is due to fine dispersion strengthening useful even at 700 °C in these alloys. As a result it became clear that the value for 100,000 h was exceeded at 700 °C and 100 MPa calculated from the Larson-Miller parameter at C = 20.     

Creep properties and microstructural evolution of austenitic TEMPALOY steels

The steam parameters in the new high efficiency fossil fuel power plants are continuously increasing, requiring new advanced materials with enhanced creep strength able to operate on the most severe temperature and pressure conditions. For super-heater and re-heater applications, TEMPALOY AA-1 steel, an evolution of 18Cr10NiNbTi alloy, has been developed through the addition of 3%Cu and B, significantly enhancing the creep resistance, while offering typical corrosion properties of 18%Cr steels. This paper describes Tenaris’ tubular products in the field of austenitic grades for applications in Ultra Super Critical power plants: the production route and the main microstructural and mechanical properties of TEMPALOY AA-1 and TEMPALOY A-3 steels, including the effect of shot blasting on steam-oxidation resistance, their creep–rupture properties and their microstructural evolution during temperature exposure are presented.     

Remaining creep life estimation by strain assessment on plant

By utilising a generalised damage parameter first introduced in the Rabotnov-Kachanov equations for tertiary creep a model is proposed for remanent creep life prediction based on in situ strain assessment. It is shown that for relatively ductile materials the rupture life can be accurately related to a single strain or strain rate measurement without a knowledge of the rupture strain. Materials data are required only in the form of the minimum creep rate-rupture life product. Considerable evidence suggests that the latter is approximately constant for relatively low stresses in a range of materials so that the specific creep response of the component material is not required. It is also demonstrated that an assessment of strain at more than one stage of the life negates the need for materials data. Consideration is given to the effect of multiaxial stressing and the model is applied to the life prediction of low alloy ferritic steel tubes and pipes.     

The high-temperature ageing of a low-alloy steel in the presence and absence of strain

The effect of high-temperature ageing in the presence and absence of strain was investigated for a high-temperature hot-rolled low-alloy steel. Mechanical tests were conducted on artificially aged material and material removed from operating plant. The steel investigated was developed for less critical elevated-temperature applications, which do not require heat-treated pressure-vessel-quality alloy materials, such as the Cr-Mo steels. Wide scatter in the high-temperature creep properties was observed. Due to a very high phosphorous content, reversible temper embrittlement at temperatures above 400°C results in Chapry V-notch values of less than 4 J cm⁻² after exposure times of only a small fraction of the component's design life. There is a difference in creep mechanism for short-time creep-rupture tests and those executed beyond 10 000 h. Confidence can only be obtained in creep-rupture data if the test field is extended to temperatures and stresses below 480°C and 150 MPa, respectively, and also to yield rupture data for up to 30 000 h. The very low fracture toughness values at room temperature and the high ductile-to-brittle transition temperature of aged material necessitate that special design features be considered. From a utility point of view, the suitability and economics of this steel for large, high-temperature structural applications—where normal design, operating and maintenance philosophies are expected to be followed—are questioned.     

The evaluation of weldment creep strength reduction factors by experimental and numerical simulations

The current design rules for welds are usually based upon the uniaxial creep rupture strength data. The effects of the stress multiaxility and the corresponding stress redistribution process of welded components are relatively ignored. As the present high-temperature testing techniques require large resources when testing welded components in full scale, the simulation of the effects will rely more on the numerical modelling. To evaluate the weldment joint efficiency this paper has proposed a general procedure in which the spatial distribution of constitutive parameters is determined by uniaxial testing while the creep process of components is simulated by numerical methods. Finite element methods are employed in the creep analysis of an AISI 316 butt-welded joint in pressurised tubes with a creep soft weld. To interpret the rupture behaviour of the tubes, different criteria are used to predict the rupture life. On the basis of the predicted structural rupture performance equations, the weldment creep reduction factors are evaluated for different design lifetimes. The reduction factors defined by the ASME code principle are found to be non-conservative in this case.