Intelligent Modeling Combining Adaptive Neuro Fuzzy Inference System and Genetic Algorithm for Optimizing Welding Process Parameters

Modified 9Cr-1Mo ferritic steel is used as a structural material for steam generator components of power plants. Generally, tungsten inert gas (TIG) welding is preferred for welding of these steels in which the depth of penetration achievable during autogenous welding is limited. Therefore, activated flux TIG (A-TIG) welding, a novel welding technique, has been developed in-house to increase the depth of penetration. In modified 9Cr-1Mo steel joints produced by the A-TIG welding process, weld bead width, depth of penetration, and heat-affected zone (HAZ) width play an important role in determining the mechanical properties as well as the performance of the weld joints during service. To obtain the desired weld bead geometry and HAZ width, it becomes important to set the welding process parameters. In this work, adaptative neuro fuzzy inference system is used to develop independent models correlating the welding process parameters like current, voltage, and torch speed with weld bead shape parameters like depth of penetration, bead width, and HAZ width. Then a genetic algorithm is employed to determine the optimum A-TIG welding process parameters to obtain the desired weld bead shape parameters and HAZ width.     

Intelligent Modeling Using Adaptive Neuro Fuzzy Inference System (ANFIS) for Predicting Weld Bead Shape Parameters During A-TIG Welding of Reduced Activation Ferritic-Martensitic (RAFM) Steel

Reduced-activated ferritic-martensitic steels are considered to be the prime candidate for structural material of the fusion power plant reactor design. Tungsten inert gas (TIG) welding is preferred for welding of those structural materials. However, the depth of penetration achievable during autogenous TIG welding is very limited and hence productivity is poor. Therefore, activated-flux tungsten inert gas (A-TIG) welding, a new variant of TIG welding process has been developed in-house to increase the depth of penetration in single pass welding. In structural materials produced by A-TIG welding process, weld bead width, depth of penetration and HAZ width decide the mechanical properties and in turn the performance of the weld joints during service. To obtain the desired weld bead geometry, HAZ width and make a reliable quality weld, it becomes important to develop predictive tools using soft computing techniques. In this work, adaptive neuro fuzzy inference system is used to develop independent models correlating the welding parameters like current, voltage and torch speed with bead shape parameters like weld bead width, depth of penetration, and HAZ width. During ANFIS modeling, various membership functions were used. Triangular membership function provided the minimum RMS error for prediction and hence, ANFIS model with triangular membership functions were chosen for predicting for weld bead shape parameters as a function of welding process parameters.     

Process Parameter Optimization of AISI 316L(N) Weld Joints Produced using Flux-Cored Arc Welding Process

Bead on plate welds were carried out on AISI 316L(N) austenitic stainless steel using flux cored arc welding process. The bead on plates weld was conducted as per L₂₅ orthogonal array with statistical design of experiment technique. In this paper, the welding parameters will be optimized based on the weld bead geometry such as depth of penetration, bead width and weld reinforcement. Grey relational analysis and desirability approach are used to optimize the input parameters like wire feed rate, voltage, travel speed and torch angle while considering the multiple output variables simultaneously. Confirmation experiment has also been conducted to validate the optimized parameters.     

Finite Element Simulation of Laser Welding of 904L Super Austenitic Stainless Steel

Experiments on Laser butt welding of 904L super austenitic stainless steel was conducted using diffusion cooled 3.5 kW slab CO₂ laser welding system. The weld geometries such as bead width and depth of penetration were measured. The laser welding process has also been simulated using ANSYS a Finite Element Analysis tool. The effect of laser power, welding speed and focal point position on the bead geometry was investigated. The experimental plan was developed based on the Taguchi technique. The comparison of the results of the simulation indicates that Finite Element Method (FEM) can predict the responses adequately within the limits of welding parameters being used. It is suggested that FEM can be used as a tool for predicting the bead geometry at low values of heat input on laser welding.     

Application of GRA and TOPSIS Optimization Techniques in GTA Welding of 15CDV6 Aerospace Material

In this article, an attempt was made to optimize the welding parameters of gas tungsten arc welding of 15CDV6 steel. Experiments based on Taguchi’s L9 orthogonal array were carried out in this research paper. The input parameters such as current, voltage, travel speed were considered for joining 15CDV6 plates of thickness 3.7 mm. Aftermath, the welds were subjected to post weld heat treatment. The performance characteristics such as bead width, reinforcement, tensile strength, hardness and depth of penetration of the welds were also measured. Grey relational analysis (GRA) and technique for order preference by similarity to ideal solution method (TOPSIS) were used for identifying the optimised input parameters. Analysis of variance was used to identify the influence of each individual parameter on the multi-objective function. The metallurgical characterisations of the optimised weld were compared with the microstructures obtained using optical microscope. It was made clear that both GRA and TOPSIS produced different set of optimized parameters. But on experimentation, it was found that optimized parameters obtained from TOPSIS produced weld with better properties. At the initial stage, the base metal reflected inferior properties to weldments but there was a significant improvement in the properties of base metal after post weld heat treatment.     

Multi-objective Optimization of Welding Parameters in MMAW for Nano-structured Hardfacing Material Using GRA Coupled with PCA

Hardfacing is one of the most productive and practical approaches to cut down operating expense on the maintenance front and at the same time to improve performance and reliability of the equipment. Presence of nano-particles in hard facing materials significantly enhances surface area to volume ratio and accordingly it improves conductivity, hardness, heat and wear resistant properties. The main objective of this paper is to efficiently apply manual metal arc welding process for hardfacing of nano-structure based electrode. The most important process variables that have been considered in conducting the experiments are welding current, arc voltage and welding speed; while the response parameters include weld bead width, reinforcement and bead hardness, respectively. Taguchi’s (L25) orthogonal array has been used to perform the experimental runs. A combination of grey relational analysis coupled with principal component analysis has been applied to identify optimal settings of the input process parameters. Moreover, the exact input and output welding parameters have been examined with the help of genetic algorithm. Finally, confirmation test has also been carried out with the optimal welding process parameters to validate the experiment result.     

Comparative Study on Processing Property Between CWW CO2 Gas Shielded Welding and SAW

 Cable welding wire (CWW) CO2 gas shielded welding is an innovative process arc welding with high efficiency, high quality and low consumption, in which cable wire is used as consumable electrode. CWW CO2 gas shielded welding and submerged arc welding (SAW) are used for contrast studies on processing property of high strength steel A36 used in ship structure. The results show that the shapes of weld seam, using CWW CO2 gas shielded welding and SAW, are good and no weld defect such as air hole, flaw, slag inclusion, incomplete fusion, lack of penetration and so on are found in the weld seam. Because the rotating of arc during CWW CO2 gas shielded welding process has a strong stirring effect on molten pool, the grain in the heat affected zone (HAZ) of the joints, using CWW CO2 gas shielded welding, is small. Tensile failure positions of joints by CWW CO2 gas shielded welding and SAW are all in the base metal, but tensile strength of CWW CO2 gas shielded welding joint is higher than that of SAW joint by an average of 2. 3%. The average impact energy of HAZ, using CWW CO2 gas shielded welding and SAW, is almost equal, but the average impact energy of the weld seam using CWW CO2 gas shielded welding is increased by 6%, and the average impact energy of the fusion line is increased by 7%. The 180° bending tests for the joints of CWW CO2 gas shielded welding and SAW are all qualified, and the joints hardness is all less than HV 355, but hardness of CWW CO2 gas shielded welding wire welding joint near the fusion line is obviously lower. It can be concluded that the properties of CWW CO2 gas shielded welding are better than those of the SAW joint, and CWW CO2 gas shielded welding is suitable for welding high strength steel A36 used in ship structure.     

Effect of Oxide Fluxes on Depth of Penetration in Flux Bounded Tungsten Inert Gas Welding of AISI 304L Stainless Steel

Flux Bounded Tungsten Inert Gas (FBTIG) welding is a modified TIG welding process in which increased depth of penetration (DoP) can be achieved by laying thin flux coatings on either side of the weld centerline. The effect of three single component fluxes viz., SiO₂, TiO₂ and Cr₂O₃ on bead geometry of autogenous melt runs in AISI 304L stainless steel for the gap between the flux layers varying from 2 to 7 mm, is studied. Results show that DoP can be improved significantly in FBTIG process using single component fluxes. Nature of the flux and the gap between the flux layers influence the weld bead geometry. Among the three fluxes used, SiO₂ is more efficient in improving the DoP. Arc constriction is the predominant mechanism operative in improving the DoP in FBTIG welding. Possibility of change in solidification mode in FBTIG weld metals of stainless steels is highlighted.     

Performance Evaluation of Arc Welding Process Using Weld Data Analysis

Shielded metal arc welding (SMAW) and metal inert gas (GMAW) welding process are the two most widely used welding processes. These processes are widely used for the construction and fabrication purpose in almost all type of industries. Some of the important factors which govern the weld quality in these welding processes are welding power sources, role of shielding gas (for GMAW process), welding consumables and skill of the welders. Currently, effects of these factors are evaluated by examining the quality of the weld produced and not by monitoring how welding process is affected by change in these factors. This is an indirect method because actual contribution made by individual parameter in physical process is effectively ignored. Further, this is expensive and time-consuming as the assessment can be carried out only after the weld is completed. Hence, a procedure to assess the quality of welding process using the data acquired while welding is in progress is preferred to testing of the weld for this purpose. In both SMAW and GMAW processes, welding speed, voltage and current are important parameters that affect the quality of the welds. Among these, monitoring of welding speed is relatively easy; but monitoring voltage and current is not. This is because, welding is a stochastic process in which wide variation in voltage and current occurs and duration of these variations is so short that they are not observed in the voltage and current displayed in the power source. However, with the help of a high-speed data acquisition system, voltage and current variations during actual welding process can be recorded and subsequently analysed to reveal very useful information on the welding process, and subsequently quality analysis of individual welding parameters can also be done. In the present study, the voltage and current signals acquired using a digital storage oscilloscope have been used to study SMAW and GMAW processes. Data was acquired for duration of 20 s at a sampling rate of 100,000 samples/s while welding is in progress. In the case of SMAW process, welding data was acquired for welds made using different welding power sources, but with same welder and same type of electrode. In the case of GMAW process, welds were made using same wire and same welder but with different gases for shielding and at different set currents. Dynamic variation in the voltage and current signals were carefully studied using time domain and statistical analyses. Results showed that differences in the characteristics of the different power sources used for SMAW process and effect of shielding gases and arc current on GMAW process could be easily revealed by such analysis. For SMAW process, results obtained could also be correlated with the appearance of the weld beads. Hence, a procedure involving high-speed data acquisition of voltage and current signal while welding is in progress and the statistical analysis of the acquired data have been proposed for monitoring of these two arc welding processes.     

Adaptive Neuro-Fuzzy Inference System (ANFIS)-Based Models for Predicting the Weld Bead Width and Depth of Penetration from the Infrared Thermal Image of the Weld Pool

Type 316 LN stainless steel is the major structural material used in the construction of nuclear reactors. Activated flux tungsten inert gas (A-TIG) welding has been developed to increase the depth of penetration because the depth of penetration achievable in single-pass TIG welding is limited. Real-time monitoring and control of weld processes is gaining importance because of the requirement of remoter welding process technologies. Hence, it is essential to develop computational methodologies based on an adaptive neuro fuzzy inference system (ANFIS) or artificial neural network (ANN) for predicting and controlling the depth of penetration and weld bead width during A-TIG welding of type 316 LN stainless steel. In the current work, A-TIG welding experiments have been carried out on 6-mm-thick plates of 316 LN stainless steel by varying the welding current. During welding, infrared (IR) thermal images of the weld pool have been acquired in real time, and the features have been extracted from the IR thermal images of the weld pool. The welding current values, along with the extracted features such as length, width of the hot spot, thermal area determined from the Gaussian fit, and thermal bead width computed from the first derivative curve were used as inputs, whereas the measured depth of penetration and weld bead width were used as output of the respective models. Accurate ANFIS models have been developed for predicting the depth of penetration and the weld bead width during TIG welding of 6-mm-thick 316 LN stainless steel plates. A good correlation between the measured and predicted values of weld bead width and depth of penetration were observed in the developed models. The performance of the ANFIS models are compared with that of the ANN models.     

Effect of Heat Input on Macro,Micro and Tensile Properties of Flux Cored Arc Welded Ferritic Stainless Steel Joints

This paper discusses the influence of Flux Cored Arc Welding Process parameters such as welding current, travel speed, voltage on bead profile, metallurgical and mechanical properties of welds of 2 mm thick 409M ferritic stainless steel sheets. The study reveals that, grain coarsening, volume fraction of martensite, hardness of heat affected zone and % of delta ferrite in ER 309 weld metal increases with increase in heat input. However, the results show that variation of heat input does not make any significant effect on tensile strength of the joint. Hence, welding parameters that provide uniform bead profile for the weld are recommended for fabrication.     

Mathematical modeling of three-dimensional heat and fluid flow in a moving gas metal arc weld pool

Mathematical models capable of accurate prediction of the weld bead and weld pool geometry in gas metal arc (GMA) welding processes would be valuable for rapid development of welding procedures and empirical equations for control algorithms in automated welding applications. This article introduces a three-dimensional (3-D) model for heat and fluid flow in a moving GMA weld pool. The model takes the mass, momentum, and heat transfer of filler metal droplets into consideration and quantitatively analyzes their effects on the weld bead shape and weld pool geometry. The algorithm for calculating the weld reinforcement and weld pool surface deformation has been proved to be effective. Difficulties associated with the irregular shape of the weld bead and weld pool surface have been successfully overcome by adopting a boundary-fitted nonorthogonal coordinate system. It is found that the size and profile of the weld pool are strongly influenced by the volume of molten wire, impact of droplets, and heat content of droplets. Good agreement is demonstrated between predicted weld dimensions and experimently measured ones for bead-on-plate GMA welds on mild steel plate.     

Influence of FCA Welding Process Parameters on Distortion of 409M Stainless Steel for Rail Coach Building

The influence of flux cored arc welding (FCAW) process parameters such as welding current, travel speed, voltage and CO2 shielding gas flow rate on bowing distortion of 409M ferritic stainless steel sheets of 2 mm in thickness was discussed. The bowing distortions of the welded plates were measured using a simple device called profile tracer. An experimental regression equation was developed to predict the bowing distortion and with this equation, it is easy to select optimized process parameters to achieve minimum bowing distortion. It is revealed that the FCAW process parameters have significant influence on bead profile and the bowing distortion.     

Microstructural and Mechanical Characterization of as Weld and Aged Conditions of AA2219 Aluminium Alloy by Gas Tungsten Arc Welding Process

In this article, Welding of AA2219 aluminium alloy using Gas tungsten arc welding process (GTAW) and evaluation of metallurgical, mechanical and corrosion properties of the joints are discussed. The weld samples were subjected to ageing process at the temperature range of 195°C for a period of 5 h to improve the properties. AA2219 aluminium plates of thickness of 25 mm were welded using gas tungsten arc welding (GTAW) process in double V butt joint configuration. The input parameters considered in this work are welding current, voltage and welding speed. Tensile strength and hardness were measured as performance characteristics. The variation in the properties were justified with the help of microstructures. The same procedures were repeated for post weld heat treated samples and a comparison was made between as weld condition and age treated conditions. The post weld heat samples had better tensile strength and hardness values on comparing with the as weld samples. Fracture surface obtained from the tensile tested specimen revealed ductile mode of failure.     

Application of Artificial Neural Network Modelling for Optimization of Yb: YAG Laser Welding of Nitinol

This article presents the investigation of Ytterbium: Yttrium aluminium garnet (Yb: YAG) laser welding of NiTinol sheet of thickness 1 mm. Welding speed, shielding gas blown distance, focal position, laser power were selected as input parameters and depth of penetration, bead width, hardness, corrosion current density were measured as performance characteristics. Experiment was designed based on L9 Taguchi design with 4 factors and 3 levels in each factor. Modelling was done using artificial neural network in four learning algorithms namely batch back propagation, quick propagation, incremental back propagation and Legvenberg-Marquardt back propagation. A comparison was made between these learning algorithms and it was found that based on least error, Legvenberg-Marquardt model was the best learning algorithm. Genetic algorithm was implemented for predicting the optimised laser welding parameters in joining NiTinol and the confirmation test results were in good agreement with the predicted results.     

高锰钢辙叉焊修技术在天铁的应用

叙述了天铁集团运输部应用KD-286焊条进行焊修高锰钢辙叉的工艺过程。针对该技术应用中出现的夹碳、裂纹、未融合及未焊透等问题,从焊接工艺方面进行了分析探讨。通过待焊补辙叉的焊修表面处理、调控刨削速度,改进堆焊顺序、控制冷却速度、调整引弧方位等措施消除了该问题,提高了辙叉使用寿命,创造了经济效益。     

Modeling of laser keyhole welding: Part II. simulation of keyhole evolution,velocity, temperature profile,and experimental verification

This article presents the simulation results of a three-dimensional mathematical model using the level set method for laser-keyhole welding. The details of the model are presented in Part I.^[4] The effects of keyhole formation on the liquid melt pool and, in turn, on the weld bead are investigated in detail. The influence of process parameters, such as laser power and scanning speed is analyzed. This simulation shows very interesting features in the weld pool, such as intrinsic instability of keyholes, role of recoil pressure, and effect of beam scanning. For verification purposes, visualization experiments have been performed to measure melt-pool geometry and surface velocity. The theoretical predictions show a reasonable agreement with the experimental observations.     

模具的合金堆焊研究

用合金堆焊模具，是在普通碳结构钢的基体上堆焊合金钢或硬质合金作为模具工作部分，以提高模具寿命及耐用度，降低模具成本.介绍了模具合金堆焊的工艺、焊接设备、堆焊材料的选用以及堆焊产生的同题及其防止措施．     

钽钢复合板钽覆层的焊接工艺研究

对钽钢复合板的钽覆层采用敷设盖板的方式焊接，研究了直流氩弧焊与交流脉冲氩弧焊两种焊接工艺对钽钢复合板焊接质量的影响。结果表明，直流氩弧焊的焊接热影响区较宽，焊接熔深为1～1．5mm，在钢与过渡金属层之间形成了中间夹层；交流脉冲氩弧焊的焊接热影响区较窄，焊接熔深为0．5～1mm，复合板焊接质量较好。与直流氩弧焊相比，交流脉冲氩弧焊焊接参数范围较宽，对焊工技能的要求相对较低，可实现连续化生产，因此更适合用于钽钢复合板钽覆层的焊接。