PSI - Issue 79

A. Bacco et al. / Procedia Structural Integrity 79 (2026) 335–341

336

1. Introduction Nowadays, welding processes are still the most widely used method for obtaining joints that are durable, resistant and more economical than those manufactured with more recent techniques. For this reason, research in this field is increasingly focused on optimising processes, to reduce their impact on the properties of the base materials and to expand their application range (Boccarusso et al., 2017; Meneghetti et al., 2017a-b). Hybrid joints currently represent a new frontier, as they would allow, with the right compromises, the advantages of two dissimilar materials to be combined in a single joint, such as exploiting the strength of steel combined with the anti-vibration capabilities of cast iron. Unfortunately, it is often particularly challenging to weld two dissimilar materials without one being overly compromised to joints made of the same material (Sepe et al., 2017). Even in homogeneous joints, the strong thermal gradients generated during welding, due to a succession of heating and cooling during the various passes, are a major issue due to the high residual stresses and plastic deformations that can be generated. These effects become more pronounced in case of dissimilar materials, which exhibit different properties, often dependent on temperature, thus responding differently to welding process. It is therefore essential to achieve a comprehensive understanding of these phenomena (Bhatti et al., 2015, Sepe et al., 2017, Sepe et al., 2020). However, experimental techniques provide only localized measurements of residual stresses, typically characterized by limited accuracy and relatively high costs. Consequently, numerical techniques, such as finite element modelling (FEM), are often used to simulate welding processes and to predict the distribution of residual stresses and deformations (Ferro et al., 2017, Sepe et al., 2021b-c). To accurately reproduce real phenomena, numerical models must incorporate both experimental process parameters and the specific properties of the involved materials. Hence, key variables values, such as the heat input during welding, the welding speed, and the number and depth of passes, must be properly implemented during the model development (Armentani et al., 2007b, Perić et al., 2014, Perić et al., 2019a-b). Moreover, for this type of application, the materials’ properties must also be realistically defined; therefore, both thermal and mechanical properties, along with their temperature-dependent variations, need to be provided, thus significantly increasing the computational load of the numerical analysis (Armentani et al., 2007a, Sepe et al., 2015, Sepe et al., 2021a).In this framework, this study presents a numerical model for the welding process of a hybrid cast iron-steel butt joint. The main objective is to evaluate the temperature distribution during successive passes and the residual stresses induced by the process, with particular attention to the thermo-mechanical responses of the involved materials.

Nomenclature I

Welding current Welding voltage Welding efficiency Welding speed

V η

v

Energy supplied to the weld bead Length of the weld bead Welding time for each pass

Q w

L bead t weld d pass

Depth for each pass