Shear wave phased array EMATs have a range of properties that make them a desirable solution compared to their piezoelectric counterparts: They do not require direct contact with the specimen and they can operate at elevated temperatures. Shear wave phased array EMATs do not need a Perspex wedge or couplant, making them easy to deploy across industrial applications even at elevated temperatures. Nonetheless, they are not commonly used because of concerns over the strength of signal excitation and lack of knowledge about potential changes in the emitted wavefields as a result of changes in the magnetic properties of the inspected material. In this work, a shear wave phased array EMAT designed to induce shear waves into paramagnetic, diamagnetic, ferromagnetic and ferromagnetic and magnetostrictive metals without compromising its angle coverage or steerability is proposed. The phased array EMAT was designed based on a dipole element wave generation source. The proposed dipole element source is formed by the interaction of the bias magnetic field of the PA EMAT and the electromagnetic field induced by the race-track coils of the PA EMAT. For each race-track coil this results in a combined source that consists of two parallel and slightly separated shear line sources of opposite polarity. A detailed study of the wave generation based on the surface loads induced by the dipole element is presented. The surface loads of the three excitation mechanisms involved in the EMAT transduction -Lorentz force, magnetisation force and magnetostriction force-were estimated in a 2D finite element model. The waves generated by the induced surface loads were estimated in a second 2D finite element model. A shear wave steerability analysis of the phased array EMAT on different metallic materials based on the 2D finite element results is presented in combination with an experimental evaluation on Aluminium 6082 (diamagnetic), mild steel EN1A (ferromagnetic), stainless steel 316 (paramagnetic), stainless steel 304 (paramagnetic) and nickel 201 (ferromagnetic and magnetostrictive). The experimental results showed good agreement with the simulation study. The simulation and experimental results concluded that the angle coverage of the phased array EMAT is almost insensitive to the inspected metallic material.