Steel Surface Defect Recognition in Smart Manufacturing Using Deep Ensemble Transfer Learning-Based Techniques

Smart manufacturing and Industry 4.0 are transforming traditional manufacturing processes by utilizing innovative technologies such as the artificial intelligence(AI)and internet of things(IoT)to enhance efficiency,reduce costs,and ensure product quality.In light of the recent advancement of Industry 4.0,identifying defects has become important for ensuring the quality of products during the manufacturing process.In this research,we present an ensemble methodology for accurately classifying hot rolled steel surface defects by combining the strengths of four pre-trained convolutional neural network(CNN)architectures:VGG16,VGG19,Xception,and Mobile-Net V2,compensating for their individual weaknesses.We evaluated our methodology on the Xsteel surface defect dataset(XSDD),which comprises seven different classes.The ensemble methodology integrated the predictions of individual models through two methods:model averaging and weighted averaging.Our evaluation showed that the model averaging ensemble achieved an accuracy of 98.89%,a recall of 98.92%,a precision of 99.05%,and an F1-score of 98.97%,while the weighted averaging ensemble reached an accuracy of 99.72%,a recall of 99.74%,a precision of 99.67%,and an F1-score of 99.70%.The proposed weighted averaging ensemble model outperformed the model averaging method and the individual models in detecting defects in terms of accuracy,recall,precision,and F1-score.Comparative analysis with recent studies also showed the superior performance of our methodology.

Effect of Ti-Mo-V composite addition on microstructure andmechanical properties of marine 10Ni5CrMoV steel

The effects of Ti-Mo-V composite addition on the evolution of precipitates in marine 10Ni5CrMoV steel andthe corresponding strength and toughness mechanisms were systematically investigated.Ti-Mo-V composite addition canform the Ti_(x)Mo_(y)V_(z)C carbide with TiC as core and Mo-V as shell in the order of Ti(C)→V→Mo.The yield strength of thespecimens is increased from 815 MPa to 876 MPa due to the nanoscale precipitates enhancing the pinning effect on grainboundaries and dislocations,and the contribution of precipitation and dislocation strengthening is increased.The decrease of ductile-brittle transition temperature from−103 to−116℃is attributed to the decrease in equivalent grainsize and the increase of high-angle grain boundary misorientation,which hinders the initiation and propagation of cracks.When the mass fraction of Ti is 0.05%,the strength and cryogenic toughness can be improved synergistically,which alsoprovides a theoretical basis and experimental reference for exploring the more excellent combination of strength andcryogenic toughness of marine 10Ni5CrMoV steel.

Quickly obtaining densely dispersed coherent particles in steel matrix and its related mechanical property

Densely distributed coherent nanoparticles(DCN)in steel matrix can enhance the work-hardening ability and ductility of steel simultaneously.All the routes to this end can be generally classified into the liquid-solid route and the solid-solid route.However,the formation of DCN structures in steel requires long processes and complex steps.So far,obtaining steel with coherent particle enhancement in a short time remains a bottleneck,and some necessary steps remain unavoidable.Here,we show a high-efficiency liquid-phase refining process reinforced by a dynamic magnetic field.Ti-Y-Mn-O particles had an average size of around(3.53±1.21)nm and can be obtained in just around 180 s.These small nanoparticles were coherent with the matrix,implying no accumulated dislocations between the particles and the steel matrix.Our findings have a potential application for improving material machining capacity,creep resistance,and radiation resistance.

Novel Methodologies for Preventing Crack Propagation in Steel Gas Pipelines Considering the Temperature Effect

Using the software ANSYS-19.2/Explicit Dynamics,this study performedfinite-element modeling of the large-diameter steel pipeline cross-section for the Beineu-Bozoy-Shymkent gas pipeline with a non-through straight crack,strengthened by steel wire wrapping.The effects of the thread tensile force of the steel winding in the form of single rings at the crack edges and the wires with different winding diameters and pitches were also studied.The results showed that the strengthening was preferably executed at a minimum value of the thread tensile force,which was 6.4%more effective than that at its maximum value.The analysis of the influence of the winding dia-meters showed that the equivalent stresses increased by 32%from the beginning of the crack growth until the wire broke.The increment in winding diameter decelerated the disclosure of the edge crack and reduced its length by 8.2%.The analysis of the influence of the winding pitch showed that decreasing the distance between the winding turns also led to a 33.6%reduction in the length of the straight crack and a 7.9%reduction in the maximum stres-ses on the strengthened pipeline cross-section.The analysis of the temperature effect on the pipeline material,within a range from-40℃ to+50℃,resulted in a crack length change of up to 5.8%.As the temperature dropped,the crack length decreased.Within such a temperature range,the maximum stresses were observed along the cen-tral area of the crack,which were equal to 413 MPa at+50℃ and 440 MPa at-40℃.The results also showed that the presence of the steel winding in the pipeline significantly reduced the length of crack propagation up to 8.4 times,depending on the temperature effect and design parameters of prestressing.This work integrated the existing methods for crack localization along steel gas pipelines.

晶粒形态对Ti-Mo钢耐蚀性能的影响

采用周期浸润腐蚀实验机、光学显微镜(OM)、扫描电镜(SEM)和电子背散射衍射(EBSD)技术等研究了晶粒形态对Ti-Mo微合金钢耐蚀性能的影响规律。结果表明:Ti-Mo微合金钢经600~750℃等温热处理后的显微组织为单相铁素体,耐蚀性能受晶粒尺寸、大角度晶界比例、再结晶程度的影响;晶界和变形组织作为阳极,晶粒作为阴极,不均匀的电化学腐蚀使实验钢的腐蚀电流升高,促进了电偶腐蚀的发生,这是影响Ti-Mo微合金钢腐蚀行为的主要机制;在不损害力学性能的前提下,可通过采取适当增大晶粒尺寸、降低大角度晶界比例、提高再结晶程度的措施,提高Ti-Mo钢的耐蚀性能;在本实验条件下,当等温温度为600℃时,Ti-Mo钢的晶粒尺寸为9.66μm、大角度晶界比例为80.3%、平均错配角为1.99°、腐蚀质量损失率为1.98 g/(m^(2)·h),其耐蚀性能较高。

等温温度对Ti-Mo微合金高强双相钢相变及组织性能的影响

采用热模拟实验、光学显微镜、显微硬度仪和透射电镜等研究了等温温度对Ti-Mo微合金高强双相钢相变、组织及性能的影响。结果表明:当等温温度在750~725℃时,实验钢组织主要为马氏体+少量贝氏体或铁素体,随着等温温度降低(≤700℃),铁素体含量增多,排碳行为增加了过冷奥氏体的稳定性,抑制了中低温区贝氏体相变,在随后的冷却阶段,过冷奥氏体更倾向转变为马氏体。铁素体与马氏体的显微硬度均随等温温度降低而增加,但进一步降低温度,马氏体硬度基本保持稳定(361 HV0.1);650℃时铁素体显微硬度最高(242 HV0.1),这主要是由于铁素体内部产生了大量细小的相间析出,同时析出强化作用减小了铁素体与马氏体的硬度差异,降低了双相钢强度对马氏体体积分数的依赖性,等温温度较高时,相间析出形成被抑制。

Effect of Mo content on the thermal stability of Ti-Mo-bearing ferritic steel

The effects of tempering holding time at 700℃on the morphology,mechanical properties,and behavior of nanoparticles in Ti-Mo ferritic steel with different Mo contents were analyzed using scanning electron microscopy and transmission electron microscopy.The equilibrium solid solution amounts of Mo,Ti,and C in ferritic steel at various temperatures were calculated,and changes in the sizes of nanoparticles over time at different Mo contents were analyzed.The experimental results and theoretical calculations were in good agreement with each other and showed that the size of nanoparticles in middle Mo content nano-ferrite(MNF)steel changed the least during aging.High Mo contents inhibited the maturation and growth of nanoparticles,but no obvious inhibitory effect was observed when the Mo content exceeded 0.37wt%.The tensile strength and yield strength continuously decreased with the tempering time.Analysis of the strengthening and toughening mechanisms showed that the different mechanical properties among the three different Mo content experiment steels were mainly determined by grain refinement strengthening(the difference range was 30-40 MPa)and precipitation strengthening(the difference range was 78-127 MPa).MNF steel displayed an ideal chemical ratio and the highest thermodynamic stability,whereas low Mo content nano-ferrite(LNF)steel and high Mo content nano-ferrite(HNF)steel displayed relatively similar thermodynamic stabilities.

Development and prospects of molten steel deoxidation in steelmaking process

In the long traditional process of steelmaking,excess oxygen is blown into the converter,and alloying elements are used for deoxidation.This inevitably results in excessive deoxidation of products remaining within the steel liquid,affecting the cleanliness of the steel.With the increasing requirements for steel performance,reducing the oxygen content in the steel liquid and ensuring its high cleanliness is necessary.After more than a hundred years of development,the total oxygen content in steel has been reduced from approximately 100×10^(-6)to approximately 10×10^(-6),and it can be controlled below 5×10^(-6)in some steel grades.A relatively stable and mature deoxidation technology has been formed,but further reducing the oxygen content in steel is no longer significant for improving steel quality.Our research team developed a deoxidation technology for bearing steel by optimizing the entire conventional process.The technology combines silicon–manganese predeoxidation,ladle furnace diffusion deoxidation,and vacuum final deoxidation.We successfully conducted industrial experiments and produced interstitial-free steel with natural decarbonization predeoxidation.Non-aluminum deoxidation was found to control the oxygen content in bearing steel to between 4×10^(-6) and 8×10^(-6),altering the type of inclusions,eliminating large particle Ds-type inclusions,improving the flowability of the steel liquid,and deriving a higher fatigue life.The natural decarbonization predeoxidation of interstitial-free steel reduced aluminum consumption and production costs and significantly improved the quality of cast billets.

Prediction model for corrosion rate of low-alloy steels under atmospheric conditions using machine learning algorithms

This work constructed a machine learning(ML)model to predict the atmospheric corrosion rate of low-alloy steels(LAS).The material properties of LAS,environmental factors,and exposure time were used as the input,while the corrosion rate as the output.6 dif-ferent ML algorithms were used to construct the proposed model.Through optimization and filtering,the eXtreme gradient boosting(XG-Boost)model exhibited good corrosion rate prediction accuracy.The features of material properties were then transformed into atomic and physical features using the proposed property transformation approach,and the dominant descriptors that affected the corrosion rate were filtered using the recursive feature elimination(RFE)as well as XGBoost methods.The established ML models exhibited better predic-tion performance and generalization ability via property transformation descriptors.In addition,the SHapley additive exPlanations(SHAP)method was applied to analyze the relationship between the descriptors and corrosion rate.The results showed that the property transformation model could effectively help with analyzing the corrosion behavior,thereby significantly improving the generalization ability of corrosion rate prediction models.

Effects of iron oxide on crystallization behavior and spatial distribution of spinel in stainless steel slag

Chromium plays a vital role in stainless steel due to its ability to improve the corrosion resistance of the latter.However,the re-lease of chromium from stainless steel slag(SSS)during SSS stockpiling causes detrimental environmental issues.To prevent chromium pollution,the effects of iron oxide on crystallization behavior and spatial distribution of spinel were investigated in this work.The results revealed that FeO was more conducive to the growth of spinels compared with Fe2O3 and Fe3O4.Spinels were found to be mainly distrib-uted at the top and bottom of slag.The amount of spinel phase at the bottom decreased with the increasing FeO content,while that at the top increased.The average particle size of spinel in the slag with 18wt%FeO content was 12.8μm.Meanwhile,no notable structural changes were observed with a further increase in FeO content.In other words,the spatial distribution of spinel changed when the content of iron oxide varied in the range of 8wt%to 18wt%.Finally,less spinel was found at the bottom of slag with a FeO content of 23wt%.

An Integrated Analysis on the Synergistic Reduction of Carbon and Pollution Emissions from China’s Iron and Steel Industry

Decarbonization and decontamination of the iron and steel industry(ISI),which contributes up to 15%to anthropogenic CO_(2) emissions(or carbon emissions)and significant proportions of air and water pollutant emissions in China,are challenged by the huge demand for steel.Carbon and pollutants often share common emission sources,indicating that emission reduction could be achieved synergistically.Here,we explored the inherent potential of measures to adjust feedstock composition and technological structure and to control the size of the ISI to achieve carbon emission reduction(CER)and pollution emission reduction(PER).We investigated five typical pollutants in this study,namely,petroleum hydrocarbon pollutants and chemical oxygen demand in wastewater,particulate matter,SO_(2),and NO_(x) in off gases,and examined synergies between CER and PER by employing cross elasticity for the period between 2022 and 2035.The results suggest that a reduction of 8.7%-11.7%in carbon emissions and 20%-31%in pollution emissions(except for particulate matter emissions)could be achieved by 2025 under a high steel scrap ratio(SSR)scenario.Here,the SSR and electric arc furnace(EAF)ratio serve critical roles in enhancing synergies between CER and PER(which vary with the type of pollutant).However,subject to a limited volume of steel scrap,a focused increase in the EAF ratio with neglection of the available supply of steel scrap to EAF facilities would lead to an increase carbon and pollution emissions.Although CER can be achieved through SSR and EAF ratio optimization,only when the crude steel production growth rate remains below 2.2%can these optimization measures maintain the emissions in 2030 at a similar level to that in 2021.Therefore,the synergistic effects between PER and CER should be considered when formulating a development route for the ISI in the future.

钼含量和均匀化处理冷却方式对β型Ti-Mo合金中点缺陷弛豫的影响

采用真空电弧熔炼制备β型Ti-xMo(x为质量分数/%,3~24)合金,并进行950℃×20 min均匀化处理,随后分别水冷和炉冷至室温,通过内耗测试结合物相和微观结构分析,研究了钼含量和冷却方式对该合金点缺陷弛豫的影响。结果表明:炉冷和水冷Ti-xMo合金中均出现2个Snoek型驰豫峰,炉冷Ti-xMo合金的弛豫峰随钼含量的变化特征与水冷合金相似。2种冷却方式下Ti-xMo合金在加热时的低温峰峰高随钼含量的增加而增大,但峰温几乎不变,峰高的增大与β相和间隙氧原子含量的增多有关;除Ti-3Mo合金外,其余合金高温峰的峰高和峰温都随钼含量增加而增大,峰高和峰温的增大与钼和氧的相互作用增强有关。

Stress-assisted corrosion mechanism of 3Ni steel by using gradient boosting decision tree machining learning method

Traditional 3Ni weathering steel cannot completely meet the requirements for offshore engineering development,resulting in the design of novel 3Ni steel with the addition of microalloy elements such as Mn or Nb for strength enhancement becoming a trend.The stress-assisted corrosion behavior of a novel designed high-strength 3Ni steel was investigated in the current study using the corrosion big data method.The information on the corrosion process was recorded using the galvanic corrosion current monitoring method.The gradi-ent boosting decision tree(GBDT)machine learning method was used to mine the corrosion mechanism,and the importance of the struc-ture factor was investigated.Field exposure tests were conducted to verify the calculated results using the GBDT method.Results indic-ated that the GBDT method can be effectively used to study the influence of structural factors on the corrosion process of 3Ni steel.Dif-ferent mechanisms for the addition of Mn and Cu to the stress-assisted corrosion of 3Ni steel suggested that Mn and Cu have no obvious effect on the corrosion rate of non-stressed 3Ni steel during the early stage of corrosion.When the corrosion reached a stable state,the in-crease in Mn element content increased the corrosion rate of 3Ni steel,while Cu reduced this rate.In the presence of stress,the increase in Mn element content and Cu addition can inhibit the corrosion process.The corrosion law of outdoor-exposed 3Ni steel is consistent with the law based on corrosion big data technology,verifying the reliability of the big data evaluation method and data prediction model selection.

Interfacial reaction between AZ91D magnesium alloy melt and mild steel under high temperature

The metallurgical quality control of magnesium(Mg)and Mg alloys in melting process is required to ensure a satisfied mechanical and corrosion performance,while the typical used steel crucible introduces impurities and interfacial interaction during melting process.Therefore,a systematic study about impurities diffusion and interfacial interaction between molten Mg and steel is necessary.In the present study,the interfacial reaction between molten AZ91D Mg alloy and mild steel during melting process was investigated with the melting temperatures of 700℃,750℃ and 800℃.The results show that Al(Fe,Mn)intermetallic layer is the intermetallic primarily formed at the interfaces of AZ91D melt and mild steel.Meanwhile,Al_(8)(Mn,Fe)5is indexed between Al(Fe,Mn)and AZ91D.AlFe_(3)C appears between the mild steel and Al(Fe,Mn)at 700℃ and 750℃,but absent at 800℃ due to the increased solubility of carbon in Mg matrix.It is found that the growth of the intermetallic layer is controlled by diffusion mechanism,and Al and Mn are the dominant diffusing species in the whole interfacial reaction process.By measuring the thickness of different layers,the growth constant was calculated.It increases from 1.89(±0.03)×10^(-12)m^(2)·s^(-1)at 700℃ to 3.05(±0.05)×10^(-12)m^(2)·s^(-1)at 750℃,and 5.18(±0.05)×10^(-12)m^(2)·s^(-1)at 800℃.Meanwhile,the content of Fe is linearly increased in AZ91D with the increase of holding time at 700℃ and 750℃,while it shows a significantly increment after holding for 8 h at 800℃,indicating holding temperature is more crucial to determine the Fe content of AZ91D than holding time.

Flow characteristics and hot workability of a typical low-alloy high-strength steel during multi-pass deformation

Heavy components of low-alloy high-strength(LAHS) steels are generally formed by multi-pass forging. It is necessary to explore the flow characteristics and hot workability of LAHS steels during the multi-pass forging process, which is beneficial to the formulation of actual processing parameters. In the study, the multi-pass hot compression experiments of a typical LAHS steel are carried out at a wide range of deformation temperatures and strain rates. It is found that the work hardening rate of the experimental material depends on deformation parameters and deformation passes, which is ascribed to the impacts of static and dynamic softening behaviors. A new model is established to describe the flow characteristics at various deformation passes. Compared to the classical Arrhenius model and modified Zerilli and Armstrong model, the newly proposed model shows higher prediction accuracy with a confidence level of 0.98565. Furthermore, the connection between power dissipation efficiency(PDE) and deformation parameters is revealed by analyzing the microstructures. The PDE cannot be utilized to reflect the efficiency of energy dissipation for microstructure evolution during the entire deformation process, but only to assess the efficiency of energy dissipation for microstructure evolution in a specific deformation parameter state.As a result, an integrated processing map is proposed to better study the hot workability of the LAHS steel, which considers the effects of instability factor(IF), PDE, and distribution and size of grains. The optimized processing parameters for the multi-pass deformation process are the deformation parameters of 1223–1318 K and 0.01–0.08 s^(-1). Complete dynamic recrystallization occurs within the optimized processing parameters with an average grain size of 18.36–42.3 μm. This study will guide the optimization of the forging process of heavy components.

Improving creep strength of the fine-grained heat-affected zone of novel 9Cr martensitic heat-resistant steel via modified thermo-mechanical treatment

The infamous type Ⅳ failure within the fine-grained heat-affected zone (FGHAZ) in G115 steel weldments seriously threatens the safe operation of ultra-supercritical (USC) power plants.In this work,the traditional thermo-mechanical treatment was modified via the replacement of hot-rolling with cold rolling,i.e.,normalizing,cold rolling,and tempering (NCT),which was developed to improve the creep strength of the FGHAZ in G115 steel weldments.The NCT treatment effectively promoted the dissolution of preformed M_(23)C_(6)particles and relieved the boundary segregation of C and Cr during welding thermal cycling,which accelerated the dispersed reprecipitation of M_(23)C_(6) particles within the fresh reaustenitized grains during post-weld heat treatment.In addition,the precipitation of Cu-rich phases and MX particles was promoted evidently due to the deformation-induced dislocations.As a result,the interacting actions between precipitates,dislocations,and boundaries during creep were reinforced considerably.Following this strategy,the creep rupture life of the FGHAZ in G115 steel weldments can be prolonged by 18.6%,which can further push the application of G115 steel in USC power plants.

A review on the multi-scaled structures and mechanical/thermal properties of tool steels fabricated by laser powder bed fusion additive manufacturing

The laser powder bed fusion(LPBF) process can integrally form geometrically complex and high-performance metallic parts that have attracted much interest,especially in the molds industry.The appearance of the LPBF makes it possible to design and produce complex conformal cooling channel systems in molds.Thus,LPBF-processed tool steels have attracted more and more attention.The complex thermal history in the LPBF process makes the microstructural characteristics and properties different from those of conventional manufactured tool steels.This paper provides an overview of LPBF-processed tool steels by describing the physical phenomena,the microstructural characteristics,and the mechanical/thermal properties,including tensile properties,wear resistance,and thermal properties.The microstructural characteristics are presented through a multiscale perspective,ranging from densification,meso-structure,microstructure,substructure in grains,to nanoprecipitates.Finally,a summary of tool steels and their challenges and outlooks are introduced.

Value of Charpy impact test results with sub-and half-size specimens extracted from overage steel bridge members

For the maintenance of steel bridges,the mechanical properties of steel used in the bridges must be elucidated.When enough dimensions of specimens cannot be extracted from the actual members,miniaturized specimens are used for evaluation.In the case of the Charpy impact test,sub-and half-size specimens are specified instead of full-size specimens of which the thickness is 10 mm.The value of absorbed energy and energy transient temperature obtained by Charpy impact tests with sub-size and half-size specimens were investigated from the viewpoint of maintenance of bridges in this study.The absorbed energy was not in proportion to the thickness of specimens of steel used in the actual overage bridges.The tendency of energy transient temperature obtained by thin specimens of the overage steel differed from that of the present steel.A method for evaluating the performance against brittle fracture occurrence based on the WES3003 criterion was examined.The results show the significance of evaluation based on the energy transient temperature for reasonable maintenance of bridges.

Recent progress in visualization and digitization of coherent transformation structures and application in high-strength steel

High-strength steels are mainly composed of medium-or low-temperature microstructures,such as bainite or martensite,with coherent transformation characteristics.This type of microstructure has a high density of dislocations and fine crystallographic structural units,which ease the coordinated matching of high strength,toughness,and plasticity.Meanwhile,given its excellent welding perform-ance,high-strength steel has been widely used in major engineering constructions,such as pipelines,ships,and bridges.However,visual-ization and digitization of the effective units of these coherent transformation structures using traditional methods(optical microscopy and scanning electron microscopy)is difficult due to their complex morphology.Moreover,the establishment of quantitative relationships with macroscopic mechanical properties and key process parameters presents additional difficulty.This article reviews the latest progress in microstructural visualization and digitization of high-strength steel,with a focus on the application of crystallographic methods in the development of high-strength steel plates and welding.We obtained the crystallographic data(Euler angle)of the transformed microstruc-tures through electron back-scattering diffraction and combined them with the calculation of inverse transformation from bainite or martensite to austenite to determine the reconstruction of high-temperature parent austenite and orientation relationship(OR)during con-tinuous cooling transformation.Furthermore,visualization of crystallographic packets,blocks,and variants based on actual OR and digit-ization of various grain boundaries can be effectively completed to establish quantitative relationships with alloy composition and key process parameters,thereby providing reverse design guidance for the development of high-strength steel.

Effect of temperature and time on the precipitation ofκ-carbides in Fe-28Mn-10Al-0.8C low-density steels:Aging mechanism and its impact on material properties

In low-density steel,κ-carbides primarily precipitate in the form of nanoscale particles within austenite grains.However,their precipitation within ferrite matrix grains has not been comprehensively explored,and the second-phase evolution mechanism during aging remains unclear.In this study,the crystallographic characteristics and morphological evolution ofκ-carbides in Fe-28Mn-10Al-0.8C(wt%)low-density steel at different aging temperatures and times and the impacts of these changes on the steels’microhardness and properties were comprehensively analyzed.Under different heat treatment conditions,intragranularκ-carbides exhibited various morpho-logical and crystallographic characteristics,such as acicular,spherical,and short rod-like shapes.At the initial stage of aging,acicularκ-carbides primarily precipitated,accompanied by a few spherical carbides.κ-Carbides grew and coarsened with aging time,the spherical carbides were considerably reduced,and rod-like carbides coarsened.Vickers hardness testing demonstrated that the material’s hardness was affected by the volume fraction,morphology,and size ofκ-carbides.Extended aging at higher temperatures led to an increase in carbide size and volume fraction,resulting in a gradual rise in hardness.During deformation,the primary mechanisms for strengthening were dislocation strengthening and second-phase strengthening.Based on these findings,potential strategies for improving material strength are proposed.