[en] An Electro-Magnetic Acoustic Transducer (EMAT) is a non-contact source used in Ultrasonic Testing (UT) which generates three types of dynamic excitations into a ferromagnetic part: Lorentz force, magnetisation force, and magnetostrictive effect. This latter excitation is a strain resulting from a magnetoelastic interaction between the external magnetic field and the mechanical part. Here, a tensor model is developed to transform this effect into an equivalent body force. It assumes weak magnetoelastic coupling and a dynamic magnetic field much smaller than the static one. This approach rigorously formulates the longitudinal Joule’s magnetostriction, and makes it possible to deal with arbitrary material geometries and EMAT configurations. Transduction processes induced by an EMAT in ferromagnetic media are then modelled as equivalent body forces. But many models developed for efficiently predicting ultrasonic field radiation in solids assume source terms given as surface distributions of stress. To use these models, a mathematical method able to accurately transform these body forces into equivalent surface stresses has been developed. By combining these formalisms, the magnetostrictive strain is transformed into equivalent surface stresses, and the ultrasonic field radiated by magnetostrictive effects induced by an EMAT can be both accurately and efficiently predicted. Numerical examples are given for illustration. (paper)

[en] Several ultrasonic and electromagnetic methods have been developed to meet the need for quality control of parts following thermal and thermochemical treatments. In this article, we present the methods developed and illustrate the results obtained

[en] ElectroMagnetic Acoustic Transducers (EMATs) are contactless transducers generating ultrasonic waves in conductive media, notably shear horizontal and torsional waves (in plates and pipes, respectively), possibly in hostile environments. In a ferromagnetic part, the elastic strain and the magnetic field couple through magnetostriction phenomena, so that a magnetostriction and magnetization forces add up to the Lorentz force created in any conductive medium. Here, a model is proposed to predict these forces for an arbitrary bias field due to the EMAT permanent magnet and whatever the current intensity in its electric circuit, whereas the usual assumption of high bias field and low intensity current leads to important model simplifications. To handle the nonlinear behavior of all the three forces when the usual assumption cannot be made, forces are expressed in the time domain. In particular, magnetostriction force generates waves at several harmonic frequencies of the driving current frequency. Forces are then transformed into equivalent surface stresses readily usable as source terms in existing models of ultrasonic radiation, under the assumption that ultrasonic wavelengths are much longer than force penetration depths, (which is generally true in NDT applications of EMATs). Force spectra computed in various EMAT configurations are compared for illustration

[en] Laser welding of aluminium generally creates embedded welding defects, such as porosities or cracks. Non Destructive Inspection (NDI) after processing may ensure an acceptable weld quality by defect detection. Nowadays, NDI techniques used to control the inside of a weld are mainly limited to X-Rays or ultrasonics. The current paper describes the use of a Laser Ultrasound (LU) technique to inspect porosities in 2 and 4-mm thick sheet lap welds. First experimentations resulted in the detection of 0.5-mm drilled holes in bulk aluminium sheets. The measurement of the depth of these defects is demonstrated too. Further experimentations shows the applicability of the LU technique to detect porosities in aluminium laser welds. However, as the interpretation of raw measures is limiting the detection capacity of this technique, we developed a signal processing using Time-Reversal capabilities to enhance detection capacities. Furthermore, the signal processing output is a geometrical image of the material's inner state, increasing the ease of interpretation. It is based on a mass-spring simulation which enables the back-propagation of the acquired ultrasound signal. The spring-mass simulation allows the natural generation of all the different sound waves and thus enables the back-propagation of a raw signal without any need of filtering or wave identification and extraction. Therefore the signal processing uses the information contained in the compression wave as well as in the shear wave