Modelling correlation between alloy composition and ferrite number in duplex stainless steel welds using artificial neural networks

Abstract

Weld metal composition is thought to be an important factor in influencing the austenite/ferrite ratio of duplex stainless steel microstructures. To produce the required balance in the austenite/ferrite ratio in the weld microstructure, the chemical composition of the welding consumables should be adjusted. In the present work, Bayesian neural network analysis has been employed to predict the ferrite number in duplex stainless steel welds as a function of composition. The technique accounts for modelling uncertainty, and automatically quantifies the significance of each input variable. In this paper, the influence of variations in the weld composition on the ferrite number have been quantified for two duplex stainless steels. Predictions are accurate compared to published methods. The role of Si and Ti in influencing the ferrite number in these alloys has been brought out clearly in this study while these elements are not given due considerations in the WRC–1992 diagram.     

Effect of heat treatment on creep and fracture behaviour of 1·25Cr–O·5Mo steel

Abstract

The effect of tempering treatments on the microstructure and creep behaviour of multipass 1·25Cr–0·5Mo steel weldments has been evaluated. While tempering invariably reduced the hardness, significant changes in microstructure were only found after heat treatment at 750°C. In this case ferrite bands developed adjacent to the fusion boundary. Tempering increased creep deformation and reduced failure lifetime for base metal specimens. For crossweld testpieces, the susceptibility to low ductility failures in the heat affected zone was found to be linked to the development of creep cavities and cracks. Thus, brittle failure modes were a function of stress, temperature, microstructure, post weld heat treatment, and to a lesser extent, specimen geometry.

MST/3069     

Creep of austenitic stainless steel welds

Austenitic stainless steels (SS) find extensive application in power, petrochemical and nuclear industries in view of their excellent elevated temperature mechanical properties, corrosion resistance, formability and weldability. However, they are susceptible to hot cracking during fusion welding. To avoid this problem, chemical composition of the welding consumable is generally adjusted to promote primary ferrite mode of solidification and retain about 3 to 10%δ-ferrite in the as-welded condition. The duplex microstructure of the weld metal undergoes transformation to carbides and a variety of intermetallic phases during elevated temperature service and causes deterioration in the mechanical properties. This paper presents a comprehensive review of the current understanding of the solidification microstructures, ageing processes and their influence on the creep behaviour of types 308 and 316 SS weld metals. The effects of varying chemical composition,δ-ferrite content, electrode coating and welding processes on creep strength and ductility are examined. Current trends in the design of welded components for creep application are also discussed.     

Creep deformation and failure of E911/E911 and P92/P92 similar weld-joints

This paper deals with characterisation of microstructure and creep behaviour of similar weld-joints of advanced 9% Cr ferritic steels, namely E911 and P92. The microstructures of the investigated weld-joints exhibit significant variability in different weld-joint regions such as weld metal (WM), heat-affected zone (HAZ), and base metal (BM). The cross-weld creep tests were carried out at 625 °C with initial applied stresses of 100 and 120 MPa. Both weld-joints ruptured by the “type IV cracking failure mode” in their fine-grained heat-affected zones (FG-HAZ). The creep fracture location with the smallest precipitation density corresponds well with its smallest measured cross-weld hardness. The welds of P92 steel exhibit better creep resistance than those of E911 steel. Whereas the microstructure of P92 weld after creep still contains laths, the microstructure of E911 weld is clearly recrystallized. The creep stress exponents are 14.5 and 8 for E911 and P92 weld-joints, respectively. These n-values indicate the “power-law creep” with dislocation-controlled deformation mechanism for both investigated weld-joints.     

Stabilisation of ferrite in hot rolled δ-TRIP steel

Abstract

A steel has recently been designed to benefit from the deformation induced transformation of retained austenite present in association with bainitic ferrite. It has as its major microstructural component, dendrites of δ-ferrite introduced during solidification. The δ-ferrite replaces the allotriomorphic ferrite present in conventional alloys of this kind. The authors examine here the stability of this δ-ferrite during heating into a temperature range typical of hot rolling conditions. It is found that contrary to expectations from calculated phase diagrams, the steel becomes fully austenitic under these conditions and that a better balance of ferrite promoting solutes is necessary in order to stabilise the dendritic structure. New alloys are designed for this purpose and are found suitable for hot rolling in the two-phase field over the temperature range 900–1200°C.     

Effect of inadequate heat treatment on creep strength of 12Cr–Mo–V steel

Abstract

Solution treatment of 12Cr–Mo–V steels below the specified temperature range leads to the development of spheroidized microstructures with dramatically reduced creep resistance. This is known to have resulted in the premature service failures of superheater tubing. Compositional and mechanical property checks currently specified in the relevant standards may not be sufficient to reveal deficiencies. Steels with Cr–Ni equivalents at the uppermost extreme of the range possible within the compositional limits of the tube steel standards show enhanced susceptibility to both δ– and α–ferrite formation. The significance of this is discussed in relation to creep strength, with particular reference to the differences between the effects of α- and δ-ferrite.

MST/147     

Microstructure evolution in AISI 304 stainless steel during near rapid directional solidification

Abstract

The microstructure evolution of near rapidly directionally solidified AISI 304 stainless steel was investigated in the present paper. It is found that the microstructure consists of δ ferrite dendrites with developed sidebranches and interdendritic austenite (γ) under the temperature gradient (G) of 20 K mm^–1 and growth rate (V) of 1·0 mm s^–1. Coupled growth microstructures of thin lamellar ferrite and austenite begin to form at a higher growth rate of 2·0 mm s^–1. The formation mechanism of the coupled microstructures is analysed based on the nucleation and constitutional undercooling criterion that the δ ferrite phase and austenite phase form alternately before the steady state growth of each phase is reached due to larger undercooling. With further increase of the growth rate up to 3·0 mm s^–1, the morphology of the δ ferrite transforms from lathy to cellular.     

Nature of inclusions in steel weld metals and their influence on formation of acicular ferrite

Abstract

The chemistry and structure of weld metal inclusions has been studied. Four submerged arc welds which utilized plate and consumables to cover a range of oxygen and deoxidant contents were examined. Analysis of the inclusions was carried out on carbon extraction replicas in a Philips 400T scanning transmission electron microscope, fitted with an energy dispersive analyser. Two major types of inclusion were found. With weld metal aluminium approaching the stoichiometric ratio with oxygen, the inclusions were crystalline and had a spinel structure at the centre with a discontinuous, polycrystalline, titanium-rich phase on the surface. With weld metal oxygen high compared with the stoichiometric ratio with aluminium, inclusions were glassy and essentially manganese silicate in composition, again with areas of a polycrystalline, titanium-rich phase on the surface. The interinclusion spacing varied little with weld metal oxygen content in the range 0·0268–0·0858 wt-%. The spacing was found to be of a similar order to the acicular ferrite grain size. The titanium-rich surface phase in all the welds was of fcc structure with a lattice parameter of 0·42 nm, which suggests a mixture of TiO and TiN, possibly rich in TiO. The spinel phase was also fcc and had a composition between galaxite (Al₂O₃MnO) and γ-alumina. Both these phases have a low lattice misfit with ferrite. A low lattice misfit of the inclusion surface layers with ferrite coupled with closely spaced inclusions would seem to be key factors in the development of an acicular ferrite weld metal microstructure.

MST/543     

Application of low Ms temperature consumable to dissimilar welded joint

Abstract

The welding of dissimilar joints is very common in systems used in oil exploration and production in deep sea waters. Commonly involves welding of low carbon steel pipes with low alloy steel forgings both with inner Inconel clad. The forged steel part undergoes a process of buttering with Inconel or carbon steel electrode before the weld of the joint. The buttering process is followed by a process of residual stresses relief. The conventional way of reducing the level of residual stresses in welded joints is to apply post welding heat treatments. Depending on the size and complexity of the parts to be joined, this can become a serious problem. An alternative technique for reducing residual stresses is to use an electrode that during the cooling process undergoes a displacive transformation at a relatively low temperature so that the deformation resulting from the transformation compensates the contraction during the cooling process, and, although many papers have been published in this direction using Fe–Cr–Ni alloys, most of them report a loss of toughness in the weld metal. Maraging steel is a family of materials with M_s temperature below 200°C and even without the final heat treatment of aging has superior mechanical properties to low alloy steels used in forgings. In this work, forged piece of AISI 4130 was buttered with Maraging 350 weld consumable and subsequently welded to ASTM A36 steel using Inconel 625 filler metal. In addition, the dissimilar base metal plates were welded together using Maraging 350 steel weld consumable. The levels of residual stress, and the toughness and microstructures of heat affected zone and weld metal were investigated.     

Development of transition metal joint for steam generator circuit of prototype fast breeder reactor

Abstract

A transition metal joint between type 304 stainless steel and 2·25Cr–1Mo steel, with Alloy 800 as the transition piece, is being developed for application in the steam generator circuit of the 500 MW prototype fast breeder reactor. As part of this programme, the hot cracking susceptibility of Inconel 82/182 and of 16–8–2 welding consumables were compared and the microstructure and mechanical properties of butt welds between type 304 stainless steel and Alloy 800, welded by the two consumables, were studied to select the appropriate welding consumables for this joint. It is recommended that the 16–8–2 consumable should be used for welding this joint because of its lower microfissuring tendency and reduced mismatch in the coefficient of thermal expansion across the joint, although this would mean a slight adverse effect on the elevated temperature mechanical properties. Further, to select the optimum post-weld heat treatment (PWHT) of the joint between Alloy 800 and 2·25Cr–1Mo steel, welded with Inconel 82/182 welding consumables, the effect of PWHT on the microstructure and mechanical properties was studied. Decreasing the PWHT temperature was found to improve the mechanical properties and the microstructural condition of this joint.

MST/842     

Multiple heat isothermal stress rupture correlations for type 316L(N) stainless steel and their use Part 2 – Applications for nuclear grade type 316L(N) stainless steel base metal and its weld metal

Abstract

The 'reference' multiple heat isothermal stress rupture correlations for stainless steel types 316 and 316L(N) base metals derived in Part 1 are used for establishing those for a specific 316L(N) stainless steel base metal and also its weld, both candidates for the forthcoming prototype fast breeder reactor at Kalpakkam. The phases that form in the weld metal during creep are the same as those in the base metal; however, the uniformly distributed δ ferrite ( ~ 7 ferrite number) in vermicular morphology present in the initial microstructure accelerates their formation and increases their quantities, resulting in poorer stress rupture properties. A simple modification allows for correlating and extrapolating the weld data to long rupture lives using the multiple heat isothermal correlations developed for the base metal.     

Microstructure and tensile properties of modified 9Cr–1Mo steel (grade 91)

Abstract

The influence of different soaking temperatures in the range 973–1623 K (below Ac ₁ to above Ac ₄) before oil quenching and tempering, on the microstructure, hardness, grain size, and tensile properties of modified 9Cr–1Mo steel has been studied. This was done in an effort to assess the tensile behaviour of the different microstructures likely to be encountered in the heat affected zone of a fusion welded joint of the steel. The steel developed predominantly martensitic structure after quenching. Soaking of steel in the intercritical temperature range (between Ac ₁ and Ac ₃) reduced the prior austenitic grain size and hardness. Soaking temperatures above Ac ₃ increased the grain size and hardness of the steel until the formation of δ ferrite at temperatures above Ac ₄. The δ ferrite formation at soaking temperatures above Ac ₄ reduced the grain size and hardness of the steel. The tensile strength of the steel exhibited a minimum for soaking in the intercritical temperature range where the ductility was highest. Strength increased and ductility decreased with further increases in soaking temperatures above Ac ₃. The formation of δ ferrite at soaking temperatures above Ac ₄ improved the ductility. The tensile properties have been correlated with the microstructures.     

Finite element analysis and experimental research on notched strengthening effect of P92 steel

Abstract

The notched strengthening effect during creep of P92 steel has been studied by finite element analysis and experimental research. It was found that there was a transforming tendency from ductile to brittle at the root of the notch and the extent of the transforming intensified with the increment of the nominal stress. It was the transforming tendency that increased the value of creep life enhancement factor. With the help of finite element software, Kachanov–Rabotnov creep damage constitutive model was embedded into the interface program and the notched specimens creep was simulated. The result has shown the Kachanov–Rabotnov model can be used to simulate the notched strengthening effect of P92 steel accurately when the material constant α?=?0·73.     

Effects of Mn and Ti on microstructure and inclusions in weld metal of high strength low alloy steel

Abstract

The influences of alloying elements on chemical composition of non-metallic inclusions, impact toughness and microstructure in weld metals of high strength low alloy steels have been studied. Results indicated that microstructure had changed from a mixture of acicular ferrite, proeutectoid ferrite, ferrite side plates and microphases to a mixture of acicular ferrite, bainite and microphases due to the addition of Mn and Ti. The impact toughness of weld metal was improved correspondingly. The volume fraction and composition of inclusions both influenced the proportion of acicular ferrite. Mn and Si based oxide globular inclusions located at the boundary of acicular ferrite plates in the weld metal produced using C–Mn–Si–Cu wire. When Mn and Ti were added to welding wires, the inclusions within acicular ferrite plates permitted fewer primary acicular ferrite plates to grow into relatively larger dimensions. Secondary acicular ferrites nucleating on pre-existing ferrite plates refined microstructure effectively.     

Comparison of creep behaviour of 2.25Cr–1Mo/9Cr–1Mo dissimilar weld joint with its base and weld metals

Abstract

Evaluation of the creep behaviour of 2.25Cr–1Mo and 9Cr–1Mo ferritic steel base metals, 9Cr–1Mo steel weld metal, and 2.25Cr–1Mo/9Cr–1Mo ferritic–ferritic dissimilar weld joints has been carried out at 823 K in the stress range 100–260 MPa. The weld joint was fabricated by shielded metal arc welding using basic coated 9Cr–1Mo electrodes. Investigations of the microstructure and hardness variations across the joint in the as welded, post-weld heat treated (973 K/1 h), and creep tested conditions were performed. The heat affected zone (HAZ) in both the steels consisted of a coarse prior austenitic grain region, a fine prior austenitic grain region, and an intercritical structure. In the post-weld heat treated condition, a white etched soft decarburised zone in 2.25Cr–1Mo steel base metal and a black etched hard carburised zone in 9Cr–1Mo steel weld metal around the weld fusion line developed. Hardness troughs also developed in the intercritical HAZ regions of both the steels. The width of the carburised and decarburised zones and hardness differences of these zones were found to increase with creep exposure. The 9Cr–1Mo steel weld metal showed higher creep strength compared to both the base metals. The 9Cr–1Mo steel base metal exhibited better creep resistance than the 2.25Cr–1Mo steel base metal at lower applied stresses. The dissimilar joint revealed lower creep rupture strength than both the base metals and weld metal. The creep strain was found to concentrate in the decarburised zone of 2.25Cr–1Mo steel and in the intercritical HAZ regions of both the steels. Creep failure in the stress range examined occurred in the intercritical HAZ of 2.25Cr–1Mo steel even though this region showed higher hardness than the decarburised zone. Extensive creep cavitation and cracks were observed in the decarburised zone.     

Creep rupture properties of nickel-base transition joints after long-term service

Abstract

Three superheater transition joints, between 2·25Cr–1 Mo and 316 stainless steel, welded with nickel–base weld metal, removed from service after 72337 h, have been examined using optical and scanning electron microscopy. In addition, microhardness measurements have been made and local chemical compositions have been analysed using the energy dispersive X-ray attachment on a scanning electron microscope. Temperature accelerated creep rupture tests have been carried out between 590 and 625°C at stresses of 31–62 MN m^?2 on cross–weld tensile specimens machined longitudinally from the walls of the joints. Detailed metallographic examinations showed the same failure mode as that found in long–term service failures. Therefore, the use of post-exposure temperature accelerated testing of uniaxial cross-weld specimens appears to be a viable method of assessing the remanent life of nickel-base transition joints operating at elevated temperatures. The applicability of various multiaxial stress rupture equations to transition joint failures is considered. The present rupture data are compared with previous data generated from initially as-welded specimens to provide upper and lower estimates of the long-term failure lives.

MST/403     

Crack growth behaviour of P92 steel under creep and creep–fatigue conditions

Abstract

High temperature creep and creep–fatigue crack growth tests were carried out on standard compact specimens machined from ASME P92 steel pipe. The effects of various loading conditions on crack growth behaviours were investigated. Crack initiation time was found to decrease with the increasing initial stress intensity factor under creep condition and further to decrease by the introduction of fatigue condition. For creep test, the crack growth rate can be well characterised by the facture mechanics parameter C*. For creep–fatigue test, the crack growth behaviour is dominated by the cycle dependent fatigue process when the hold time is shorter, but it becomes dominated by the time dependent creep process when the hold time becomes longer.     

Texture evolution in hot band and annealed hot bands of low alloyed ferritic stainless steel

Abstract

The texture evolution in hot band and annealed hot bands of low alloyed ferritic stainless steel with about 11 wt-%Cr was experimentally investigated using quantitative texture analysis. While the hot band texture was composed of components of α fibre and in particular δ fibre, its microstructure was a banded structure of mostly relaxed martensite and retained ferrite. Both the texture and microstructure of the hot band was derived from partially recrystallised austenite. During single phase hot band annealing, there was a strong sharpening in the strength of the texture components of δ fibre by strain induced boundary migration of the retained ferrite and formation of fine carbide sheets leading to the persistence of ferrite banding. In contrast, two phase hot band annealing resulted in the formation of a nearly equiaxed duplex ferrite grain structure with an aggregate of precipitated carbides within the transformed ferrite grains and complete elimination of microstructural banding of the hot band, and also led to the occurrence of a texture memory phenomenon.     

An analysis of the creep fracture characteristics of austeniticferritic steel transition welds

A detailed analysis has been made of the creep failure characteristics of crossweld testpieces taken from a transition weld between $\text{2}\text{1}\text{4}$ Cr1Mo ferritic steel and AlSl Type 316 austenitic steel plates joined using a 17Cr 8Ni 2Mo austenitic weld metal. Creep conditions were selected which led to fracture of the crossweld samples in a manner which simulated the low-ductility premature failures encountered with this type of transition weld during service in electricity generating plants, i.e. fracture occurred through the development of intergranular cracks in the heat-affected zone of the $\text{2}\text{1}\text{4}$ Cr1 Mo steel at a distance of some 75–150 ~tm from the weld interface. The numerical modelling procedure described is shown to predict this typical failure pattern, and the associated localized strain accumulation in the region where cracking occurs, by taking into account the known creep properties of the weld metal, the parent $\text{2}\text{1}\text{4}$Cr1Mo steel and the various regions of the heat-affected zone.     

Creep resistance of similar and dissimilar weld joints of P91 steel

Abstract

Two experimental weld joints, a similar weld joint of 9Cr–1Mo steel and a dissimilar weld joint of 9Cr–1Mo and 2.25Cr–1Mo steels, were fabricated by the TIG+E method and post-weld heating was applied. Creep testing was carried out at temperatures ranging from 525 to 625°C in the stress range 40–240 MPa. Creep rupture strength was evaluated using the Larson–Miller parameter. Extended metallography including transmission electron microscopy was performed and critical zones were indicated where fractures were concentrated during the creep exposure. At high temperatures rupture of the dissimilar weldment occurred in the heat affected zone (HAZ) of the weld metal while rupture of the similar weldment was located in the HAZ of the parent material. The processes of recovery seem to be the main causes of decrease in creep rupture strength of both weld joints in comparison to the parent materials.