Prediction of multiwire submerged arc weld bead shape using neural network modelling

Ridings, G. E.; Thomson, Rachel C.; Thewlis, G.

doi:10.1179/136217102225006822

Cited by 27 publications

(27 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Em aços de baixa liga e, também, em alguns aços inoxidáveis austeníticos, a fissuração no reaquecimento pode ocorrer nesses casos, em geral na região de crescimento de grãos da ZTA, mas, eventualmente, também na zona fundida. Para aços de baixa liga para trabalho a alta temperatura, a sensibilidade do material à fissuração no reaquecimento pode ser estimada por várias equações empíricas [25]: (28) válida para aços com %C £ 0,18 e %Cr < 1,5. Sensibilidade: DG > 0; (29) válida para aços com %C entre 0,1 e 0,25; %Cr < 1,5; %Mo < 0,2; %Cu < 1% e %V, %Nb e %Ti < 0,15.…”

Section: Fissuração No Reaquecimentounclassified

See 1 more Smart Citation

Algumas equações úteis em soldagem

Marques

Modenesi

2014

Soldag. insp.

View full text Add to dashboard Cite

Resumo Uma das principais características da soldagem é o grande número de parâmetros e variáveis envolvidas, além da complexidade de suas inter-relações. Isto dificulta a elaboração de modelos matemáticos adequados para descrever processos e/ou prever os resultados Palavras chave: Soldagem, Equações, ModelamentoAbstract: Most welding processes are characterized by a large number of variables, many of them related to each other by complex and mostly unknown relationships. This makes it rather difficult to develop mathematical models, to simulate different aspects of welding, and to preview the results of a welding procedure beforehand. As a result, a great number of empirical, or semi-empirical, equations have been developed by many authors since welding has been of some industrial importance. Many of those models can be useful to welding professionals but they are generally scattered in many different papers and in other publications. In this paper, an attempt is make to put together some equations related with different aspects of the welding technology. It is expected that this could assist welding professionals to resolve some problems in their daily work. The meaning of the different terms in those equations, their units, and values of some of the constants present are shown. Equations are grouped according with their application and, when it was possible, an attempt was made to unify the terminology and units used by the different authors. Keywords: Welding, Equations, Modeling1 Autor para o qual as correspondências devem ser enviadas. IntroduçãoA tecnologia da soldagem se caracteriza por um grande número de variáveis e parâmetros operacionais interrelacionados de maneira quase sempre complexa, o que torna difícil o desenvolvimento de modelos matemáticos teóricos ou empíricos para processos ou produtos. Muitas tentativas de modelamento têm sido feitas, mas com êxito limitado.Pode se encontrar na literatura um número bastante elevado de equações e ábacos bastante úteis na solução de problemas complexos ou mesmo no trabalho rotineiro dos profissionais da área de soldagem, que minimizam a necessidade de testes e ensaios, com economia de tempo e materiais.Entretanto, essas informações estão espalhadas em diferentes textos e publicações, o que dificulta sua utilização. Outra grande dificuldade é que muitos desses modelos apresentam severas restrições por serem empíricos e obtidos por regressão linear múltipla e, portanto, com aplicação limitada a situações diferentes daquelas utilizadas na sua obtenção. Outros modelos são soluções analíticas de equações diferenciais, ambas sujeitas a simplificações que as afastam consideravelmente do objeto real (por exemplo, a variação de temperatura durante uma operação de soldagem).Nesse trabalho foram reunidas diversas equações e ábacos obtidos na literatura, tanto em trabalhos originais como em outros que, posteriormente, utilizaram e desenvolveram os diferentes modelos, e na internet, tendo todo esse material um grande potencial para ser útil no trabalho rotineiro do prof...

show abstract

Section: Fissuração No Reaquecimentounclassified

“…Mais recentemente, modelos baseados em técnicas mais poderosas como, por exemplo, o uso de redes neurais, foram desenvolvidos [28]. Contudo, todos eles dependem criticamente dos dados usados em sua criação e devem ser usados com extremo cuidado.…”

Section: Formato Do Cordãounclassified

Algumas equações úteis em soldagem

Marques

Modenesi

2014

Soldag. insp.

View full text Add to dashboard Cite

show abstract

“…The larger the voltage, the wider the weld (see Figure 2). Some sources state that generally speaking the penetration reduces a little when voltage is increased [2,8,[11][12][13][14], but when used for butt welds or bead-on-plate welds, the penetration remains the same. Penetration will reduce for V-shape weld preparations [1].…”

Section: Primary Process Parametersmentioning

confidence: 99%

“…The higher the current, the deeper the penetration (see Figure 3), causing a higher dilution percentage [1, 2,8,[11][12][13][14].…”

Section: Primary Process Parametersmentioning

confidence: 99%

Multi-wire submerged arc welding: a study of secondary process parameters

Boudry

Delboo

Faes

et al. 2014

SCAD

View full text Add to dashboard Cite

Abstract:The goal of this thesis is to investigate the effect of secondary process parameters on the geometry of the weld bead in multi-wire submerged arc welding. Bead-on-plate welding experiments will be conducted using 3, 4 and 5 wires. The parameters being varied will be the wire stick out, the angles of the torches and the spacing in between. Primary process parameters like voltage, current and welding speed will be kept constant. Penetration, bead width and height, percentage dilution, HAZ, weld penetration shape factor and weld reinforcement shape factor will be measured and evaluated. Fewer experiments will be needed because of the use of a design of experiments approach, namely a fractional factorial design. This will allow to determine the significance of each single parameter and their mutual interaction with respect to weld bead shape.Keywords: Multi-wire; Submerged arc welding; Design of experiments; Secondary parameters; Fractional factorial INTRODUCTIONMulti-wire submerged arc welding (SAW) is a high deposition rate process that is frequently used for piping and offshore applications. It is capable of joining thick steel plates or pipes in a fully automated process at high speed, which makes it economically very interesting to use for relatively large production series. Typically process parameters are set using rules of thumb and optimised through trial and error. Once good parameters are found, they are kept constant for the entire batch. In a previous master thesis the effect of primary process parameters (voltage, current, welding speed) on the geometrical quality of plate butt welds has been experimentally evaluated [3]. This study will focus on the effect of secondary process parameters on the geometrical weld quality. First, a short description of the multi-wire SAW process is given followed by a discussion of primary and secondary process parameters. Next, the experimental approach designed to evaluate the effects of secondary process parameters is elaborated. MULTI-WIRE SUBMERGED ARC WELDING Process descriptionSubmerged arc welding is a form of arc welding in which the weld pool is shielded by a mineral weld flux. This flux protects the weld pool from atmospheric influences, provides alloying elements to the weld pool and has several other properties that influence the weld [4]. The arc is formed within the flux and causes the continuously fed wire electrode to melt off and dilute with the molten base material. Because the arc is submerged, the heat loss is minimum resulting in a thermal efficiency as high as 95% [5]. No eye protection is needed [4][5][6][7][8], but this also implies that no visual monitoring can be done during welding [8]. This causes the need for automation [2,4]. To increase productivity, multiple wires can be fed into the same weld pool (see Figure 1). Every single wire can have different settings, which gives each of them a specific role. The first wire has specific settings to cause a deep penetration [6,8,9], the other wires have a more filling function. This makes...

show abstract

“…18 The inputs from three-wire submerged arc welds included plate thickness, preparation angle and depth, wire parameters (current, voltage angle, for three wires), speed, stickout and spacings between wires. The output was the entire weld bead profile.…”

Section: Category 3 Examplementioning

confidence: 99%

Performance of neural networks in materials science

Bhadeshia¹,

Dimitriu²,

Forsik³

et al. 2009

Materials Science and Technology

115

View full text Add to dashboard Cite

Neural networks are now a prominent feature of materials science with rapid progress in all sectors of the subject. It is premature, however, to claim that the method is established. There are genuine difficulties caused by the often incomplete exploration and publication of models. The assessment presented here is an attempt to compile a loose set of guidelines for maximising the impact of any models that are created, in order to encourage thoroughness in publication to a point where the work can be independently verified.

show abstract

Prediction of multiwire submerged arc weld bead shape using neural network modelling

Cited by 27 publications

References 18 publications

Algumas equações úteis em soldagem

Algumas equações úteis em soldagem

Multi-wire submerged arc welding: a study of secondary process parameters

Performance of neural networks in materials science

Contact Info

Product

Resources

About