Proceedings of the 17 th ISME Conference ISME17 October 3-4, 2015, IIT Delhi, New Delhi 1 ISME-M-014 OPTIMISATION OF OUTPUT PARAMETER OF SUBMERGED ARC WELDING BY RESPONSE SURFACE METHODOLOGY Siddharth Tamang, N. K. Singh Mechanical Engineering Department ISM Dhanbad Dhanbad, Jharkhand, India sid.khiungba@yahoo.com, nks_221@yahoo.co.in Abstract—Submerged arc welding is a widely used welding technology which relies on the selection of the optimum combination of input variables for achieving the required qualities of weld. Response surface methodology (RSM) was applied to develop a model for predicting the responses: Penetration (P), Bead Width (W) and Reinforcement (H). The four input variable i.e arc voltage (V), arc current (A), travel speed (S) and electrode extension (N) were selected for assessing the desired quality and process optimization. The main and interaction effects of these input variables on important bead geometry parameters were determined and their influence is reported. I. INTRODUCTION The development of submerged arc welding (SAW) process evolved from the need of shielding the deposited weld metal from the contaminating effect of the atmosphere. This also resulted in the minimization of heat loss as the flux also creates an insulating layer. Hence, the welding process of SAW usually involves the arc submerged inside a layer of granular flux due to which it is devoid of sparks, spatter and visible arc [1]. The process may be controlled either automatically or manually with the former method being used widely [2]. The SAW process is normally limited to the flat or horizontal-fillet welding positions; however R. Shakamoto et al [3] have invented a SAW welding method in the vertical or upward direction. The tedious trial and error methods where the welds are examined and determined if they are as required have been simplified by researchers who proposed mathematical models. These models help in determining the relations between input and output parameters in welding. A widely used method to determine the welding process model is to use experimental design and regression analysis [4]. In the previous years the various design of experiments techniques have been used to carry out optimization. Evolutionary algorithm and computer network have been adapted for many applications in different areas. One the widely used methodology for this purpose is the Response surface methodology (RSM) [5–10]. The response surface methodology (RSM) is a model formation procedure to find how different factors affect the response of an experiment[4]. Different polynomial models are developed of first and second order. These parameters include the parameters under consideration and their statistical significance useful for developing, improving and optimizing process (geometry)[11]. Statistical softwares like MINITAB, Design Expect and MATLAB can be employed to develop the experimental plan and analyse the data. ANOVA analysis can help to determine the adequacy of the model and the significance of the parameters. The main objective of the research paper is to develop mathematical models using RSM with the aid of MINITAB statistical software to predict the bead width, reinforcement and penetration responses. The effect of welding variables namely arc voltage (V), arc current (A), travel speed (S) and electrode extension (N) and their interactions on the above mentioned responses are elucidated by the 3D surface plots generated by the model. II. MATERIAL The experiments were conducted as per the design matrix randomly to avoid errors due to noise factors. The mild steel work piece (150 × 60 × 10 mm – 2 pieces) is cut and V groove of angle 45° as per the standards is prepared. The experiments were conducted with rectifier type semi-automatic SAW equipment with AC power source. Copper coated electrodes AWS ER70S-6, 3.15 mm diameter, of coil form (ESAB brand) and basic-fluoride-type (equivalent to DIN 8557) granular flux was used. The job was firmly fixed to a base plate by means of fixture and then the submerged arc welding was finally carried out. The welding parameters were recorded during actual welding to determine their fluctuations, if any. The slag was removed and the job was allowed to cool down. Welding is carried out for the square butt joint configuration. The job is cut at three sections for similar welding conditions. The material used was a hot rolled mild steel plate of grade IS2062. Square butt joints of 10mm thickness were used to