Metal transfer instability in gas metal arc welding

Abstract

This work presents a simplified model of metal transfer in gas metal arc welding. The model incorporates key features of metal transfer including the change in droplet diameters as welding moves from the globular into the spray metal transfer region, and the increase in welding voltage that is observed to occur as the pendant droplet grows. The model predicts that an instability arises in the globular metal transfer region, which leads to deterministic chaos and complex limit cycles with many droplet sizes. The instability also causes deterministic chaos with a characteristic gap in droplet diameters at the transition to spray mode metal transfer. The model explains observed features of metal transfer in some detail, including the existence and location of preferred bands of droplet sizes. Whether the instability is present or not defines the boundary between chaotic globular metal transfer and the stable drop spray transfer mode. The identification of deterministic chaos in gas metal arc welding metal transfer opens the way for new approaches to welding control.     

Statistics of signature images for arc welding fault detection

Abstract

The present work covers the statistical analysis of signature images for weld quality monitoring and fault detection. Welding signature images are two-dimensional histograms of voltage and current data. Signatures from production welding are compared with reference good welding using a statistical analysis of both the principal components, which are fitted with shifted normal distributions, and the remainders not explained by the principal components, which are fitted with a multidimensional normal distribution. A signature basis set is employed, allowing efficient real time computations. The statistical data fits are illustrated with data from overlap welds, and examples of fault detection with out of position welds are given.     

The humping phenomenon during high speed gas metal arc welding

Abstract

A commonly observed welding defect that characteristically occurs at high welding speeds is the periodic undulation of the weld bead profile, also known as humping. The occurrence of humping limits the range of usable welding speeds in most fusion welding processes and prevents further increases in productivity in a welding operation. At the present time, the physical mechanisms responsible for humping are not well understood. Thus, it is difficult to know how to suppress humping in order to achieve higher welding speeds. The objectives of this study were to identify and experimentally validate the physical mechanisms responsible for the humping phenomenon during high speed gas metal arc (GMA) welding of plain carbon steel. A LaserStrobe video imaging system was used to obtain video images of typical sequences of events during the formation of a hump. Based on these recorded video images, the strong momentum of the backward flow of molten metal in the weld pool that typically occurred during high speed welding was identified as the major factor responsible for the initiation of humping. Experiments with different process variables affecting the backward flow of molten weld metal were used to validate this hypothesis. These process variables included welding speed, welding position and shielding gas composition. The use of downhill welding positions and reactive shielding gases was found to suppress humping and to allow higher welding speeds by reducing the momentum of the backward flow of molten metal in the weld pool. This would suggest that any process variables or welding techniques that can dissipate or reduce the momentum of the backward flow of molten metal in the weld pool will facilitate higher welding speeds and productivity.     

Observations of the phenomenon of abnormal arc voltage occurring in pulsed metal inert gas welding of aluminium alloy

Abstract

At the time of arc reignition after short circuiting during electrode positive polarity, cathode spots are newly formed in the centre of the weld pool surface, where oxides scarcely exist. The work function of the cathode surface increases and the cathode spots concentrate because of the lack of sufficient oxides, leading to an increase in the potential gradient across the cathode fall space and the adjoining contraction space. Consequently, the arc voltage becomes abnormally high in spite of the short arc length. Moreover, when electrode polarity is switched from positive to negative immediately after a droplet has detached from the wire tip, cathode spots are newly formed on the surface of the molten metal remaining at the wire tip, where little oxide exists, leading to an abnormal increase in arc voltage as well. Therefore, the change in arc voltage does not necessarily indicate a fluctuation in the arc length.     

Signature images for arc welding fault detection

Abstract

The present work elucidates the signature image approach to welding fault detection, covering the calculation of signature image data objects from blocks of welding electrical data (voltage and current), the definition of appropriate vector operations, and the manipulation of the signatures to permit detection of welding faults. Detection of out of position welds in overlap joints is illustrated, and the relevant changes in signature features are related to the physics of the short arc welding process. Finally, it is shown how fault detection performance can be improved by optimising the way the signatures are calculated.     

Artificial neural networks and acoustic emission applied to stability analysis in gas metal arc welding

Abstract

The present paper describes the application of neural networks to obtain a model for estimating the stability of gas metal arc welding (GMAW) process. A neural network has been developed to obtain and model the relationships between the acoustic emission (AE) signal parameters and the stability of GMAW process. Statistical and temporal parameters of AE signals have been used as input of the neural networks; a multilayer feedforward neural network has been used, trained with back propagation method, and using Levenberg Marquardt's algorithm for different network architectures. Different welding conditions have been studied to analyse the incidence of the parameters of the process in acoustic signals. The AE signals have been processed by using the wavelet transform, and have been characterised statistically. Experimental results are provided to illustrate the proposed approach. Finally a statistical analysis for the validation of the experimental results obtained is presented. As a main result of the study, the effectiveness of the application of the artificial neural networks for modelling stability analysis in welding processes has been demonstrated. The regression analysis demonstrates the validity of neural networks to predict the stability of welding process using the statistical characterisation of the signal parameters of AE that have been calculated.     

Keyhole gas tungsten arc welding of commercially pure zirconium

Abstract

Keyhole gas tungsten arc welding (K-GTAW), a novel variant of GTAW, has been used to join commercially pure zirconium. The process enables single pass welding of 6˙35 mm thick zirconium using conventional GTAW equipment and a high current torch, without expensive filler metal addition or joint preparation. The mechanical properties and the microstructure of the resulting joints were characterised. It is concluded that the K-GTAW process, with its high productivity combined with low capital investment requirements, can be successfully used for welding relatively heavy section zirconium.     

Wire melting phenomena in gas metal arc welding

Abstract

The wire melting rate in gas metal arc welding using constant voltage power supplies is well described by a parabolic model that considers the effects of electrical resistance and arc heating. However, tests performed with constant current power supplies indicate that the melting rate can deviate from this model when welding current is close to the globular–spray transition value. Although the causes of this anomaly have not yet been well established, most authors consider them to be related to variations in liquid metal temperature close to the metal transfer mode transition current. The present work evaluates the effect of welding parameters on the wire melting rate when a constant current power supply is used. A numerical model that considers the contribution of both Joule and arc heating was developed to calculate the temperature distribution in the wire and its melting rate. The model was used to assess the changes in either arc heating or metal vaporisation that might explain the melting irregularities. High speed cinematography was used to analyse metal transfer in the current range associated with the melting irregularities.     

Arc power and efficiency in gas tungsten arc welding of aluminium

Abstract

A calorimetric study of gas tungsten arc welding of aluminium is described. The present study comprised experiments in which autogenous welding runs were each made on a block of electrical conductor grade aluminium. The blocks were all approximately cubic in shape which, when combined with the high thermal conductivity of aluminium, ensured that their temperature equalised soon after the completion of a run. Each sample was immersed in insulating material before welding so that heat losses to the surroundings were minimised. Thermocouples were attached to the block in each experiment and the bulk temperature rise was related to the energy input associated with the welding run. The effects of arc polarity, alternating current balance, shielding gas composition, arc length and welding current on the arc power and arc efficiency were investigated. The results obtained with alternating current are compared to those for direct current, and the differences are explained.     

Analysis of electrical characteristics for hybrid pulsed micro-tungsten inert gas welding arc

Abstract

The present paper experimentally investigates and analyses in detail the electrical characteristics of the arc for the hybrid pulsed micro-tungsten inert gas process. In particular, an average voltage–current (V–I) graph plotting approach is proposed to describe the dynamic characteristics of the hybrid arcs using the filtered voltage and current of the arc. It is shown that the average voltage and electrical resistance of the heavily hybrid pulsating (HHP) arc are lower than those of the slightly hybrid pulsating (SHP) arc, particularly during the background current stage. Furthermore, the analyses reveal that the average V–I characteristic graph for the arc forms a hysteresis loop, primarily because of the thermal inertia of the arc, during a pulse period. This closed loop is larger for the SHP arc than for the HHP arc, and shrinks longitudinally with increasing frequency and width of the current pulse. These results thus demonstrate that the HHP arc has a greater thermal inertia, particularly at higher current and pulse frequency. Experimental results show also that the mean voltage of the arc and the features of the hysteresis loop vary markedly with weld penetration in addition to pulsating frequency and pattern, and thus suggest the possibility of monitoring the hybrid pulsed weld penetration via arc sensing.     

Neural network model of heat and fluid flow in gas metal arc fillet welding based on genetic algorithm and conjugate gradient optimisation

Abstract

Although numerical calculations of heat transfer and fluid flow can provide detailed insights into welding processes and welded materials, these calculations are complex and unsuitable in situations where rapid calculations are needed. A recourse is to train and validate a neural network, using results from a well tested heat and fluid flow model to significantly expedite calculations and ensure that the computed results conform to the basic laws of conservation of mass, momentum and energy. Seven feedforward neural networks were developed for gas metal arc (GMA) fillet welding, one each for predicting penetration, leg length, throat, weld pool length, cooling time between 800°C and 500°C, maximum velocity and peak temperature in the weld pool. Each model considered 22 inputs that included all the welding variables, such as current, voltage, welding speed, wire radius, wire feed rate, arc efficiency, arc radius, power distribution, and material properties such as thermal conductivity, specific heat and temperature coefficient of surface tension. The weights in the neural network models were calculated using the conjugate gradient (CG) method and by a hybrid optimisation scheme involving the CG method and a genetic algorithm (GA). The neural network produced by the hybrid optimisation model produced better results than the networks based on the CG method with various sets of randomised initial weights. The CG method alone was unable to find the best optimal weights for achieving low errors. The hybrid optimisation scheme helped in finding optimal weights through a global search, as evidenced by good agreement between all the outputs from the neural networks and the corresponding results from the heat and fluid flow model.     

Microstructures of gas metal arc weld metal of 950 MPa class steel

Abstract

The effects of shielding gas composition on the properties and microstructure of single pass weld metals produced by GMA (gas metal arc) groove welding of 950 MPa class steel plates have been investigated. The shielding gas employed was a mixture of argon (Ar) and carbon dioxide (CO₂) (0–25%), and the weld heat input was ～3 kJ mm. With increasing CO₂ content, the hardness of the weld metal decreased from 380 HV to 280 HV, and the absorbed energy of the Charpy impact test decreased from 130 J to 90 J. The microstructures of the weld metal, consisting primarily of low carbon martensite and carbide free bainite, became more bainitic as the CO₂ content of the shielding gas was increased. It was also found that the MA constituent, embrittling microstructure, was formed in the granular bainitic area, the volume fraction of which increased with the CO₂ content of the shielding gas.     

Derivation of current density distribution by arc pressure measurement in GTA welding

Abstract

Arc pressure is an important factor in understanding physical arc phenomena, especially its effects on the penetration, size and shape of GTA welding. The purpose of the present study is to derive the current density distribution on the anode base metal from the measured arc pressure distribution in GTA welding using the results from previous investigators. Using the measured arc pressure distribution on the anode base metal in GTA welding from the central part to the circumference and the equations of Lin et al. and Maecker, the current density distribution was derived. Applying the derived equation from the present work, the current density distribution was calculated from the low current region to the high current region by means of shield gas mixing ratio. It is compared and discussed with the practical welding current and the derived current by numerical integration of the current density distribution from the central part to the circumference region.     

Coating composition,weld parameter and consumable conditioning effects on weld metal composition in shielded metal arc welding

Abstract

Pipeline construction has been moving towards the use of higher strength low carbon steels as these become available. In order to achieve the required high strength and toughness to match those of the parent pipe, girth weld metals have to rely mostly on their composition. The aim of the present study was to characterise the influence of the consumable chemistry and moisture content, and the welding parameters on the composition of weld metals produced by E8010 and E9010 shielded metal arc (SMA) cellulosic consumables. The alloying compounds in the consumable coating were isolated using a caustic soda solution, analysed with electron backscattered diffraction and identified as a variety of oxides. Making use of a mechanised SMA welding machine, the effect on weld metal alloying element contents of the welding conditions was evaluated. Smaller welding arc voltages and lower consumable coating moisture content produced weld metals with richer chemistry. The element transfer mechanisms from the consumable coating to the weld metal are also discussed.     

Laser stabilisation of arc cathode spots in titanium welding

Abstract

Cathode spot formation is very pronounced during arc welding of titanium and titanium alloys. The dynamic behaviour of these spots was observed to interfere with metal transfer during welding, this interference being a fundamental cause of poor weld quality in these alloys. In the present work, stabilisation of the arc cathode spot with a focused Nd–YAG laser beam during pulsed gas metal arc welding of titanium was investigated. The laser beam was focused near the leading edge of the weld pool and the laser power and focus spot size were varied to determine the values required to confine the cathode spot to the laser focus position. The results showed that, for fixed welding conditions, the laser power required to prevent cathode spot motion varied as a function of focus spot size. The required laser power was minimised at 200 W for a spot size of 0.6 mm. The laser stabilised arcs had lower voltage but approximately the same current density as stabilised arcs. Increased welding speeds required marginally higher laser powers to stabilise the spot, but the minimum power was still attained with a 0.6 mm focus spot diameter. The laser power density required for stabilisation decreased as spot size was increased, varying from almost 10⁶ W cm^?2 at the smallest spot size to approximately 10⁴ W cm^?2 at the largest. Cathode spot stabilisation improved weld quality by reducing spatter generation and weld bead irregularity.     

Novel rotation arc system for narrow gap MAG welding

Abstract

The present paper develops a hollow axis motor driven high speed rotation arc system for narrow gap welding (NGW), and introduces the features of this system. Some welding experiments were then carried out to investigate the characteristics of welding wire melting and weld formation for this new process. Experimental results show that the melting rate of wire increases and the residual melting ball diameter of wire tip decreases respectively with an increase in rotation speed, and this melting rate is higher in pulsed welding and NGW respectively than in dc and flat plate welding. Furthermore, the arc rotation can obviously improve the penetration into NGW groove sidewalls and bead shape, and thus the system has been used to weld practical NGW joints successfully.     

Parametric approach to positional fault detection in short arc welding

Abstract

This work investigates a parametric technique for short arc welding fault detection. The independent component analysis algorithm is applied to split welding voltage and current signals into arcing and short circuiting regions. A series of parameters including means, medians and normalised inner products are calculated by processing the different sections of welding electrical data. Out of position welds in overlap joints are studied with investigation of the connection between parameters and short arc welding features. The effect of changing shielding gas is also illustrated. Figures of merit to quantify the change in various parameters between faulty and good welds are compared and discussed. It is shown that the data separation into two regions with independent component analysis is useful in providing various parameters that can assist in short arc welding fault detection under different conditions.     

Development of low spatter CO2 arc welding process with high frequency pulse current

Abstract

Metal transfer phenomena and spatter generation in CO₂ arc welding with a solid wire were investigated, and a low spatter welding process using a high frequency pulse rectangular current was developed. The optimal conditions of high frequency pulse CO₂ arc welding were determined to be a peak current of 450–550 A and pulse frequency of 450–750 Hz. These high frequency pulse currents influenced the droplet oscillation due to resonance between the applied pulse frequency and the natural frequency of the droplet. A droplet was regularly transferred by 9–11 pulses, and the average interval of metal transfer was ～16 ms, which was half of that in conventional CO₂ arc welding. The average droplet weight is 34 mg, showing a large reduction in comparison with that of the conventional method. As a result, the total spatter weight was reduced by 70% in comparison with the conventional method, and particularly large spatters more than 0·5 mm in diameter were reduced from 25 to 3 mg s^?1.     

Control of nitrogen content and porosity in gas tungsten arc welding of high nitrogen steel

Abstract

In welding of high nitrogen steel (HNS), it is essential to control the nitrogen content and porosity in the weld metal. In this paper, the influence of shielding gas composition and heat input on the nitrogen content and porosity in the weld metal of HNS was investigated by gas tungsten arc welding. The experimental results indicate that the weld nitrogen content increases as N₂ in the shielding gas is increased in the same heat input of welding. The weld nitrogen content decreases with increasing the heat input for pure argon used as a shielding gas, whereas it increases with increasing the heat input for the shielding gas including some nitrogen. The nitrogen pore can be avoided when the nitrogen content in the shielding gas is <4% in the heat input range of 528–2340 J mm^–1.     

Arc heating hot wire assisted arc welding technique for low resistance welding wire

Abstract

An assistant arc was used to preheat the wire for hot wire tungsten inert gas (TIG) welding which was independent on the resistance of the welding wires and avoided the drawbacks of the traditional hot wire preheating method. The new method is suitable for low resistance wire such as copper and aluminium. The experimental results showed that the wire temperature varied linearly with preheating current and hyperbolically with wire feeding speed. The temperature of wires achieved 60% of their melting points when LF6, H90, HS201 and H08Mn2Si welding wires were used at a current less than 50 A. With arc assisted hot wire, the welding deposition efficiency of the HS201 wire was increased by 96% compared with the traditional TIG welding method, while its microstructure was similar to that of the cold wire welding.