The industrial application of deep neural networks to automate the ultrasonic weldment flaw classification system has some limitations. The major problem that affects the classification performance ...of deep neural networks is the noise in the ultrasonic signals. So, in this article, a deep neural network, also known as autoencoder is investigated to remove noise from ultrasonic signals before feeding them to deep learning classifiers. A database was generated from specimens that were closely resembled with pipe weldment geometry having counterbore and weldment defects. Those signals were, later on, corrupted with noise to mimic industrial applicability. An autoencoder was then employed to remove noise from counterbore, planer and volumetric weldment defect signals. The classification performance of the convolutional neural network (CNN) was evaluated in three different ways. At first, without employing the autoencoder, secondly, on the denoised outputs of the autoencoder and on third CNN was trained with the noiseless signals but was tested on the denoised outputs of the autoencoder. The results demonstrate that the autoencoder can successfully remove noise from the ultrasonic weldment defect signals, which consequently improve the defect classification accuracy of the artificially intelligent deep learning classifiers.
•Investigation of denoising performance of the autoencoder for realistic ultrasonic weldment signals.•Classification of ultrasonic signals gathered from realistic weldment defects.•Time shifting of signals for data augmentation.•Performance evaluation of convolutional neural network with and without autoencoder for noisy signals.
Creep strength enhanced ferritic (CSEF) steels containing 9-12wt% chromium have been extensively used in fossil-fuel-fired power plants. Despite their excellent creep resistance at high temperatures, ...premature failures (especially Type IV cracking) are often found in the fine-grained heat affected zone (HAZ) or intercritical HAZ of the welded components. This failure mode is preceded by the strain localization in the HAZ, as measured by the Digital Image Correlation (DIC) technique. The present work aims to develop a finite-element based computational method to determine the micromechanical and microstructural origin of the strain localization phenomenon. We construct a two-dimensional digital microstructure based on the actual microstructure of ferritic steel weldments by using the Voronoi-tessellation method, to account for the effects of its large grain-size gradients. A mechanism-based finite element method is developed for modeling the high temperature deformation resulting from a synergy of thermally activated dislocation movements, diffusional flow and grain boundary sliding. The numerical results agree well with the strain measurements by our DIC technique, particularly revealing the effect of pre-welding tempering on the evolution of strain localization in HAZ of creep resistant steel weldments. It is found that the diffusional creep with dependence on grain sizes, dislocation creep with dependence on material strength, and more importantly, grain boundary sliding, contribute synergistically to the creep strain accumulation in the HAZ, and their relative degree of significance is quantified. The creep rupture life will be investigated in the companion paper.
The influences of the applied radial bending stresses on the corrosion behaviour of dissimilar weldments of two types of martensitic heat-resistant steel (F92/Co3W2) are investigated with NaCl–Na2SO4 ...deposited salts at 620 °C and 24 h employing a four-point bending jig. The applied external stress exhibits no obvious influence on corrosion products, which mainly consist of Fe2O3, Fe3O4, and FeCr2O4. However, the stress tends to aggravate the corrosion degree, and the weldments with stress addition yield more mass gain and greater corrosion scale thickness. Such phenomenon is more serious in the parent F92, implying F92 is more sensitive to the corrosion environment with stress. The corrosion scale of weldments without stress shows typical double-layer structure: outer Fe–O layer and inner Fe–Cr–O layer, while the weldments with bending stresses show interesting multi-layers (10 or more layers) with Fe-oxide and Fe, Cr-oxide alternately appearing. The mechanism for hot corrosion behaviour and the influence of hot corrosion on the local mechanical properties of different welding zones with and without stresses are further discussed.
•Hot corrosion behavior of F92/Co3W2 dissimilar weldments.•Corrosion layer changes from double-layer to multi-layer alternation after imposed stress.•Effect of stress on Co3W2 base material was weaker than that on F92.•Stress degrades microregions mechanical properties from nanoindentation results.
•Thermogravimetry technique has been used to identify the kinetics of oxidation.•Oxide scale of HAZs was found in more spalling and cracking than that of Weld metal.•The non-presence of Cr oxide in ...inner-scale of HAZs, leads to inner diffusion of oxygen to metal.•The parabolic-growth rate was investigated in oxidized regions of weldment.
Overheating is the main concern in steam generating system because of the deposition of oxalates on the water side resultant of high-temperature which creates localized spallation/cracking in normally protective oxide scale, leads to catastrophic failures. This research article, therefore, focused on oxidation kinetics, oxide cracking of HAZs and weld-metal of Shielded Metal Arc (SMA) weldment in T22 steel after exposure to thermal-cyclic conditions at 900 °C in air. Thermogravimetry method was applied to examine the kinetics-of oxidation. The scale thickness on HAZs was merely higher than weld-metal because the formation of higher extent of surface cracks, and non-availability of Cr-in inner-scale as confirmed by EDX, provides the easy diffusion path for ionic-transport. Oxide-scale of these regions of weldment obeys the parabolic-kinetics.
A three-dimensional coupled model in a Eulerian framework has been developed in COMSOL Multiphysics software and used to study the complex phenomena of thermal and material flow during the friction ...stir welding (FSW) process. The moving heat source (tool) effect is modelled using a coordinate transformation. The frictional heat as a function of temperature-dependent yield strength of AA2219-T87 material and the deformation energy of plasticized material flow are considered. Further, the plasticized material flow around the rotating tool is modelled as non-Newtonian fluid using partial-sticking/sliding boundary condition with a computed slip factor (
δ
) at the workpiece-tool material interfaces. The coupled Eulerian model prediction accuracy has been validated against the experimental weldment zones and found a good agreement in terms of the shape and size. Subsequently, the effects of tool-pin profiles (cylindrical and conical) on thermal distribution, material flow, shear strain rates, thermal histories, and weldment zones were studied. It is found that the maximum temperatures, material flow velocities, and shear strain rates are low with the conical tool pin in contrast to the cylindrical one, and it is partly attributed to increased mixing of shoulder and pin-driven material flow around the rotating tool, which in turn decreased the size of weldment zones. Also, the maximum temperatures, material flow velocities, and shear strain rates on the advancing side are higher than those of the retreating side. Therefore, it is suggested to use the CFD model to design the FSW process and tool parameters in a cost-effective way in contrast to the tedious experimental route.
•Mechanical properties and microstructure of weldments were quantitatively evaluated.•The ICHAZ and adjacent regions exhibit the most sensitivity to fatigue loading.•The correlation between the ...micro-region properties and microstructure was clarified.
Understanding the evolution of microstructure and micro-region properties of weldments during cyclic loading is a challenge for the reliability assessment of high-temperature components. This work is devoted to quantitatively evaluating the variation of micro-region properties and corresponding responsible microstructural features of P92 steel weldments under interrupted fatigue tests and subsequent creep fracture. To achieve this target, high-resolution characterization techniques, including electron backscatter diffraction (EBSD) and nanoindentation tester were used. The results revealed that the microstructures in the inter-critical heat affected zone (ICHAZ) and adjacent regions, fine-grained HAZ (FGHAZ) and parent metal (PM), show significant sensitivity to fatigue loading. Notably, the grain size in the ICHAZ reaches saturation after fatigue cycles of 10% fatigue lifetime, and the weld metal (WM) keeps unchanged throughout the fatigue loading. After subsequent creep fracture, only the WM presents an increasing trend in the average grain size. Nanoindentation tests uncover that the reductions of the elastic modulus and microhardness in each region also present three cyclic softening behaviour during the fatigue process due to the evolution of martensite lath structure. However, due to nucleation and growth of cavities accelerated by carbide precipitation, relatively low elastic modulus and microhardness were observed after subsequent creep loading.
•Investigation of CNN for classification of noisy ultrasonic flaw signals.•Time shifting of signals for data augmentation.•Performance comparison of fully connected deep neural network and CNN.
...Ultrasonic flaw classification in weldment is an active area of research and many artificial intelligence approaches have been applied to automate this process. However, in the industrial applications, the ultrasonic flaw signals are not noise free and automatic intelligent defect classification algorithms show relatively low classification performance. In addition, most of the algorithms require some statistical or signal processing techniques to extract some features from signals in order to make classification easier. In this article, the convolutional neural network (CNN) is applied to noisy ultrasonic signatures to improve classification performance of weldment defects and applicability. The result shows that CNN is robust, does not require specific feature extraction methods and give considerable high defect classification accuracies even for noisy signals.
This study covers the review of the degradation of ferritic stainless-steel weldments between 2015 and 2022. The industrial and automotive sectors make extensive use of ferritic stainless steel (FSS) ...due to its superior oxidation and corrosion resistance, low price, high thermal conductivity, and low thermal expansion. However, it has been reported that ferritic stainless steel is harder to weld than austenitic stainless steel and that doing so would probably result in a weaker welded joint owing to the coarsening of grains high welding temperatures. According to past research, the amount of heat applied during the welding procedure affected how soon the FSS (409 M) weldment degraded after being exposed to NaCl (3.5%) medium. The coarsening of the grains was considered to be the cause of this. When the shielding gas' CO2 content increased, the intergranular corrosion of the FSS weld metal was found to increase. Welds made with the ER430LNb filler metal had significantly lower intergranular corrosion of FSS (AISI 441) than those made with the ER430Ti filler metal. It was discovered that boiling Cu-CuSO4 - 50% H2SO4 solution increased the corrosion rate for the FSS (AISI 430) weldment more than boiling 40% HNO3 Solution. Weldments made of FSS (AISI 430) were found to be negatively affected by the CuCuSO4 - 50% H2SO4 environment in terms of intergranular corrosion attack.
•Novel approach named Similar Weldment Case Selection (SWCS), which predicts welding results of a new material.•Nugget-size weld-current series (NWS) describes the shape of the relation between ...weldcurrent and nugget size.•Similarity between two NWSs of different materials calculated (quantified) with the dynamic time warping (DTW) method.•SWCS yields superior accuracy than the twelve algorithms do when the two materials are similar or different.
Resistance spot welding (RSW) is a critical joining method in sheet-metal industries. The machine-learning technique fueled by the historical experimental data of the existing materials has been used to build the data-driven model (DDM). The DDM is expected to be a promising tool to investigate a new material and its welding behavior because DDM can narrow the range of the test matrix and can thus reduce the number of necessary physical experiments and the cost. However, one of crucial data quality problems with machine learning is that training data sets’ lack of descriptability for test sets causes poor prediction. This research starts by indicating that such data quality problems that exist in the context of weldment design. To resolve this problem, the presented study introduces a novel approach named Similar Weldment Case Selection (SWCS), which predicts the key parameter, the nugget size, of spot welding results of a new material by selecting the most similar one among the existing welding cases and then constructing a prediction model to generate the results. In order to overcome the difficulties with defining the selection criteria only with the material properties and geometric features, this study has come up with another factor, nugget-size weld-current series (NWS), to consider; the NWS is a factor that describes the shape of the relation between weld-current and nugget size. The similarity between two NWSs of different materials is calculated (quantified) with the dynamic time warping (DTW) method. Initially, the twelve conventional algorithms are tested for varying degrees of descriptability between the two weldment designs for test and train datasets; the prediction accuracies are found to be proportional to the train set’s descriptability on the test set. The results are then compared with those from the SWCS. The SWCS yields superior accuracy than the twelve algorithms do when the two materials are similar or different. However, the superiority disappears when the two are the same.
In this work, an insight into the premature Type IV cracking was undertaken to clarify its mechanisms in Grade 91 steel pipe weldments. High-resolution microscopy observations of the as-welded ...heat-affected zone (HAZ) reveal that the commonly recognized fine-grained region susceptible of cracking on the edge of HAZ belongs to the inter-critical HAZ (ICHAZ), rather than the fine-grained HAZ (FGHAZ). Instrumented indentation tests uncover that the ICHAZ is the weakest region across the weld, exhibiting the largest displacement and the lowest hardness in three thermal stages. Localized deformation of matrix grains and high stress triaxiality in the ICHAZ promoted nucleation of creep cavities along grain boundaries. This localized deformation was induced by the creep strength mismatch of matrix grains with different Cr concentrations. Cavity-free regions exhibit a relatively homogenous Cr distribution, whereas, an inhomogeneous Cr distribution is observed in the cavity-containing regions. It is believed that this local Cr inhomogeneity in the ICHAZ is caused by the partial dissolution of Cr-rich M23C6 carbides and an insufficient homogenization during rapid welding thermal cycles.
Display omitted
•Heterogenous fine-grained strucuture belonging to the intercritical heat-affected zone degraded faster than the other regions.•Localized C and Cr inhomogeneities in the matrix weakened intercritical heat-affected zone.•Localized high stress triaxiality around coarsened precipitates induced cavity nucleation at the triple juctions of grain boundaries.