Indian Welding Journal

Open Access Open Access  Restricted Access Subscription Access
Open Access Open Access Open Access  Restricted Access Restricted Access Subscription Access

Artificial Neural Networks Based Prediction of Penetration in Activated Tungsten Inert Gas Welding


Affiliations
1 Mechanical Engineering Department, Kalyani Govt. Engineering College, Kalyani- 741235, West Bengal, India
     

   Subscribe/Renew Journal


Using GTAW, or tungsten inert gas (TIG) welding, weld penetration is usually lesser than the other arc welding processes. ATIG (Activated-flux TIG) welding can be a good alternative to provide deep penetration, and hence, improved productivity. In this work, 304L SS plate of 8 mm thickness was used as base plate, and a flux with a mixture of SiO2, MnO2 and MoO3 was used as a ternary flux in the ratio of 1:1:2. A 2-factor 3-level response surface methodology of central composite design was considered for designing experimental runs. Back Propagation (BP) type Artificial Neural Networks (ANN) model was developed to assess penetration in ATIG welding by using heat input and pulse frequency as the two process parameters. The ANN chosen has 2-10-1 network structure. Results show that the predicted values through ANN are conforming quite well to the experimentally obtained penetration, and hence, the applicability of ANN.

Keywords

Welding, Activated Tungsten Inert Gas Welding, ATIG, Artificial Neural Networks, ANN, NN, Prediction, Depth of Penetration.
User
Subscription Login to verify subscription
Notifications
Font Size

  • Howse DS and Lucas W (2000); Investigation into arc constriction by active fluxes for tungsten inert gas welding, Sc and Tech of Welding and Joining, 5, pp.189-193.

  • Gurevich SM, Zamkov VN and Kushnirenko NA (1965); Improving the penetration of titanium alloys when they are welded by argon tungsten arc process, Avtomatich Svarka, 9, pp. 1-4.

  • Lin HL and Wu TM. (2012); Effects of Activating Flux on Weld Bead Geometryof Inconel 718 Alloy TIG Welds, Materials and Manufacturing Processes, 27, pp.1457-1461.

  • Modenesi PJ, Âquio R, Ario, A. and Pereira IM (2000); TIG welding with single-component fluxes, Journal of Materials Processing Technology, 99, pp. 260-265.

  • Huang HY (2010); Effects of activating fluxzon the welded joint characteristics in gas metal arc welding, Materials and Design, 31, pp. 2488-2495.

  • Tseng KH and Hsu CY (2011); Performance of activated TIG process in austenitic stainless steel welds, Journal of Materials Processing Technology, 211, pp. 503-512.

  • Cai Y, Luoa Z, Huanga Z, and Zenga Y (2016) ; Effect of cerium oxide flux on active flux TIG welding of 800 Mpa super steel, Journal of Materials Processing Technology, 230, pp. 80-87.

  • Bhattacharya A (2016); Revisiting Arc, Metal Flow Behavior in Flux Activated Tungsten Inert Gas Welding, 31, pp. 343-351.

  • Roy S, Samaddar S, Uddin NM and Das S (2017); Effect of Activating Flux on Penetration in ATIG Welding of 316 Stainless Steel", Indian Welding Journal, Vol.50, pp.72-80.

  • Ahmad A and Alam S (2019); Parametric Optimization of TIG Welding using Response Surface Methodology, Materials today: Proceedings,18, pp.3071-3079.

  • Vora JJ, Abhishek K and Srinivasan S (2015) ; Attaining Optimized ATIG Welding Parameters for Carbon Steels by Advanced Parameter-less Optimization Techniques: with Experimental Validation, Journal of the Brazilian Society of Mechanical Sciences and Engineering, 41, pp. 260-280.

  • Sivakumar J, Vasudevan M and Korra N N (2020); Systematic Welding Process Parameter Optimization in ATIG Welding of Inconel 625, Transactions of the Indian Institute of Metals, 73, pp. 555-569.

  • Acharya S. and Das S (2020); Effect of Activating Flux in Gas Tungsten Arc Welding, Weld Fab TechTimes, 4, pp.12-21.

  • Chandrasekar G, Kannan R, Prabakaran MP and Ganesamoorthy R (2020); Effect of Activating Flux (Metal Oxide) on the Weld Bead Nomenclature of Tungsten Inert Gas Welding Process - A Review, IOP Conf. Series: Materials Science and Engineering, 988, 012084.

  • Unni KA and Vasudevan M (2020) ; Numerical Modelling Fluid Flow and Weld Penetration in Activated TIG Welding, Materials Today: Proceedings, 27, pp. 2768-2773.

  • Niagaj J (2021), Influence of Activated Fluxes on the Bead Shape of ATIG Welds on Carbon and Low Alloy Steels in Comparison with Stainless Steel AISI 304L, Metals, 11, pp. 530-543.

  • Vora J, Patel VK, Srinivasan S, Chaudhari R, Pimenov DY, Giasin K and Sharma S (2021); Optimization of Activated Tungsten Inert Gas Welding Process Parameters Using Heat Transfer Search Algorithm: With Experimental Validation Using Case Studies, Metals, 11, pp.1-16, 2021.

  • Acharya S, Gonda D and Das S (2022); Achieving Favourabe Depth of Penetration and Productivity of ATIG Welds Utilising the AHP, Indian Science Cruiser, 36, pp. 17-23.

  • Ates H (2007); Prediction of gas metal arc welding parameters based on artificial neural networks, Materials & Design, 28, pp. 2015-2023.

  • Pal S, Pal SK and Samantaray AK (2008); Artificial neural network modelling of weld joint strength prediction of a pulsed metal inert gas welding process using arc signals, Journal of Materials Processing Technology, 202, pp. 464-474.

  • Chandrasekhar N & Vasudevan, M (2010); Intelligent Modeling for Optimization of A-TIG Welding Process, Materials and Manufacturing Processes, 25, pp.1341-1350.

  • Vasudevan M, Bhaduril AK, Baldev Raj and Rao KP (2007); Genetic-Algorithm based computational models for optimizing the process parameters of A-TIG welding to achieve target bead geometry in type 304 L (N) and 316 L(N) stainless steels, Materials and Manufacturing processes, 22, pp. 641-649.

  • Vasudevan M, Bhaduri AK, Baldev Raj and Rao KP (2007); Genetic-algorithm-based computational models for optimizing the process parameters of A-TIG welding to achieve target bead geometry in type 304L(N) and 316 L(N) stainless steels, Materials and Manufacturing Processes, 22, pp.641-649.

  • Saha S and Das S (2018); Investigation on the effect of activating flux on tungsten inert gas welding of austenitic stainless steel using AC polarity, Indian Welding Journal, 51(2), pp. 84-92.

  • Saha S and Das S (2019); Application of activated tungsten inert gas (A-TIG) welding towards improved weld bead morphology in stainless steel specimens, Annual Technical Volume of Production Division Board, The Institution of Engineers (India), IV, pp. 13-23.

  • Saha S and Das S (2020); Effect of polarity and oxide fluxes on weld-bead geometry in activated tungsten inert gas (A-TIG) welding, Journal of Welding and Joining, 38(4), pp. 380-388.

  • Saha S, Paul BC and Das S (2021); Productivity improvement in butt joining of thick stainless steel plates through the usage of activated TIG welding, SN Applied Sciences, 3(416), pp. 416/1-10.


Abstract Views: 69

PDF Views: 2




  • Artificial Neural Networks Based Prediction of Penetration in Activated Tungsten Inert Gas Welding

Abstract Views: 69  |  PDF Views: 2

Authors

Samarendra Acharya
Mechanical Engineering Department, Kalyani Govt. Engineering College, Kalyani- 741235, West Bengal, India
Debasish Gonda
Mechanical Engineering Department, Kalyani Govt. Engineering College, Kalyani- 741235, West Bengal, India
Santanu Das
Mechanical Engineering Department, Kalyani Govt. Engineering College, Kalyani- 741235, West Bengal, India

Abstract


Using GTAW, or tungsten inert gas (TIG) welding, weld penetration is usually lesser than the other arc welding processes. ATIG (Activated-flux TIG) welding can be a good alternative to provide deep penetration, and hence, improved productivity. In this work, 304L SS plate of 8 mm thickness was used as base plate, and a flux with a mixture of SiO2, MnO2 and MoO3 was used as a ternary flux in the ratio of 1:1:2. A 2-factor 3-level response surface methodology of central composite design was considered for designing experimental runs. Back Propagation (BP) type Artificial Neural Networks (ANN) model was developed to assess penetration in ATIG welding by using heat input and pulse frequency as the two process parameters. The ANN chosen has 2-10-1 network structure. Results show that the predicted values through ANN are conforming quite well to the experimentally obtained penetration, and hence, the applicability of ANN.

Keywords


Welding, Activated Tungsten Inert Gas Welding, ATIG, Artificial Neural Networks, ANN, NN, Prediction, Depth of Penetration.

References





DOI: https://doi.org/10.22486/iwj.v57i1.223729