Efficient welding conditions in magnetic pulse welding process

Abstract

This study investigates the experimental research of the appropriated conditions for the magnetic pulse welding of AA6060T6 tubular assembly. Some welding tests were performed with two process parameters: the charging voltage and the width of the air gap between the two parts to be welded. A torsion-shear test, associated with the material fracture surfaces observation, gives an insight about the appropriateness of the welding conditions. The failure mode of the destructive test gives a dimensional criterion of the weld that is used as weld quality. It appears that the voltage does not strongly affect the weld quality for a low gap. It is possible to find an optimal gap range giving a high weld length. When the gap is too small, it is necessary to increase the pressure on the flyer, and some cracks appear in the material. Similarly, when the gap is too large, the high impact energy damages the welded interface.

Introduction

Magnetic pulse welding (MPW) is a cold welding process which has quickly emerged owing to the rising requests in lightweight structures manufacturing. Nowadays, it can be considered as the ideal solution assembly in many applications as multi-material joining or tubular overlap parts [1]. It is a cost effective, fast and potentially reliable process which could favourably replace conventional welding process or be combined with other low cost ones. The bonding principle is to create a solid state weld by impact at the interface of the two parts to be welded [2], [3]. The feasibility of the MPW of a large variety of metal combination with dissimilar melting temperatures, such as copper/zirconium metallic glass, copper/brass, copper/aluminium, aluminium/magnesium, aluminium/steel, has been studied [4], [5], [6], [7]. The MPW process is particularly interesting for hollow tubes joining. In this case, the inner surface of the external hollow tube subjected to the magnetic field is fully welded on the outer surface of the inner tube. The MPW also allows welding plate shape workpiece by overlapping.

A typical magnetic pulse welding device includes two main units. The first one is an electrical pulse generator, consisting of a power supply and a bank of capacitors that are designed to store unusual high quantity of electrical energy. The typical magnitude is about several kJ. The second component of the MPW equipment is the work station, necessary to convert the provided electrical pulse into a pulsed magnetic field. The high current flows through a coil inside which the parts to weld are inserted. The produced magnetic field induces Eddy currents in the moving part, called flyer. The Eddy currents interact with the induced magnetic field according to the Lorentz force law, leading to a high velocity impact between the flyer part and the inner part, as illustrated in Fig. 1. The impact gradually occurring along the interface subjected to the magnetic field creates a weld at the atomic-level. The high pressure locally generated during the impact brings the two surfaces into intimate contact by plastic deformation so that the atoms are bonded. The high impact velocity is favourable to hyperplasticity and allows the local plastic deformation of the surface under high pressure. In addition, during the oblique collision, the impact generates a strong shearing at the interface, leading to the removal by jetting of the surfaces oxides and impurities. This phenomenon, called “jetting and bonding mechanism” cleans the surface and facilitates the atomic bonding [4]. The pressure occurring at the interface is in the range of some GPa and the average impact velocity is typically around some hundred m s⁻¹. The weld presents an unusual wavy shape caused by the impact intensity.

Several theories were presented to determine welding parameters related to the wave formation [8]. Ben-Artzy et al. found an empirical relationship between interfacial wavelength and the free moving distance of the shock waves in the welded tubular parts [7]. The collision angle and the collision velocity were originally found to be the basic welding parameters [4], [7]. Witmann suggested relationships giving a welding range [9]. Considering the wave formation as a consequence of a Kelvin-Helmoltz instability, the critical collision angle decreases with the increase of the collision velocity. Grignon et al. found a growing curve for the straight-wavy transition [10]. In several studies, it is noticed that finding a logical practical value of the collision angle is difficult [11], [12], [13], [14]. Watanabe and Kumai [14] investigated an in situ observation of the MPW using high-speed video camera and found that the collision angle is 0° at the initial collision point and increases continuously during the welding. The weld formation is able to occur at low impact angles as well as at high ones [11], [12], [13]. In recent studies, it is claimed that the weld formation involves several interrelated parameters [11], [12], [13]. The joint is formed following a deformation, temperature and temporal conditions determined by a combination of interfacial parameters. In practice, it is difficult to assess the accurate values of such parameter. In the present study, the determination of the welding conditions is carried out regarding the effect of the main monitorable process parameters on the weld quality. Generally, the wavy morphology of the welded interface is used as weld indicator. This Consideration is discussed in the literature [12], [15], [16], [17], [18]. The weld may have a straight shape and could be resistant as well as a wavy interface [15]. Göbel et al. claimed that the wavy interface is not a prerequisite for the weld formation [16]. Karhaman et al. compared the maximum tensile shear strength of several lap joints with different interfacial shape (straight, wavy and intermediate) and did not notice any significant discrepancy [17]. Aizawa et al. found that the weld joint is always stronger than the weaker metal whatever the feature of the welded joint [18]. The destructive tests are more efficient to reveal the weld quality if the stress concentrations due to structure effect are limited [19]. For instance, during a tensile-shear test or a peel test, the failure takes place in the non-welded zone. Hence the welded interface is not properly subjected to the mechanical stress because of stress concentration developed at the corner of the welded zone [19], [20]. For this reason, a torsion-shear test is suggested in the present study for the assessment of the weld quality. This test is more appropriated to minimise the structure effects.