556 IEEE TRANSACTIONS ON MAGNETICS, VOL. 46, NO. 2, FEBRUARY 2010 Detection of Fatigue Limit Thanks to Piezomagnetic Measurements Said Lazreg and Olivier Hubert ENS Cachan/CNRS UMR8535/UPMC/PRES UniverSud Paris, LMT-Cachan, F-94235 Cachan cedex, France Many works have been attempted to propose fast methods to estimate the fatigue limit of materials. Fatigue limit is a high-cycle fatigue property that can be interpreted as a stress level above which the material is subjected to an accumulation of irreversible changes cur- rently associated to the micro-plasticity activation. When a ferromagnetic material is concerned, the sensitivity of magnetic properties to microstructure makes magnetic techniques an interesting way to estimate this stress level. In this work, piezomagnetic measurements have been performed for that purpose. Results are compared to those obtained thanks to a self-heating method. Index Terms—Fatigue limit, magento-elastic coupling, piezomagnetism. I. INTRODUCTION M ANY components of engineering structures are sub- jected to cyclic mechanical loading that can bring their final failure [1]. The phenomenon is called fatigue. The fatigue behavior of materials and structures depends on many parameters and has been extensively studied by many authors in order to understand the phenomenon and clear out its basic properties. Fatigue data are usually scattered so that the iden- tification of the fatigue properties requires the use of many specimens. The fatigue limit is one of these interesting properties. It is defined as the stress level below which a fatigue test will not lead to the failure of the material. When applying a stress level exceeding this limit, micro-plasticity occurs and initiates cracks. This high-cycle fatigue (HCF) property is usually deduced from the so-called Wöhler diagram [1], which plots in a semilogarithmic stress/number of cycles graph the number of cycles to failure as a function of stress amplitude. This destructive technique requires a large number of speci- mens and several days of experiments. An alternative technique for a rapid estimation of the fatigue limit is therefore relevant. Temperature measurement is one of these techniques. The so-called “self-heating” method involves mechanically cycling the sample and determining the steady-state temperature for different stress levels. Beyond a stress limit, the temperature starts to increase significantly. The change of temperature denotes that fatigue limit is exceeded: The intrinsic dissipation is associated to micro-plasticity [2]. When a magnetic mate- rial is concerned, several magnetic methods are available to measure the fatigue limit: measurement of classical magnetic quantities such as coercive field or permeability [3], or less conventional measurement of Barkhausen Noise Emission [4]. Theses methods demonstrate a high sensitivity to the changes of the microstructure. Nevertheless they require an interruption of the fatigue test, and online monitoring is usually not possible. In the case of cyclic loading, the dependence of the magneti- zation on the stress defines the so-called piezo-mag- netic behavior [5], [6]. is usually nonlinear, non monot- onous and hysteretic [7], [8]. The idea detailed in the paper Manuscript received June 21, 2009; revised September 19, 2009; accepted September 20, 2009. Current version published January 20, 2010. Corre- sponding author: O. Hubert (e-mail: hubert@lmt.ens-cachan.fr). Digital Object Identifier 10.1109/TMAG.2009.2033126 is to perform a direct piezomagnetic monitoring of a real fa- tigue test without any interruption of the test. The evolution of self-heating and piezomagnetic loops area as function of the number of cycles and stress level are compared. A new crite- rion to detect the fatigue limit is proposed. II. EXPERIMENTAL PROCEDURE A. Test Apparatus The experimental device used is presented in Fig. 1. It allows the monitoring of the magnetization and the changes of temper- ature of a specimen submitted to a sinusoidal tensile-compres- sive mechanical loading. The mechanical system consists of a mechanical testing machine (MTS high-dynamic hydraulic ma- chine-5T). Two thermocouples are positioned respectively on the specimen and on the inferior jaw of the machine. The differ- ence of temperature between the specimen and the inferior jaw is stored. The magnetic device is composed of a primary winding (P-coil), an H-coil for the measurement of the mag- netic field , and a pickup coil (B-coil). The magnetizing current delivered by a current amplifier to the P-coil creates the magnetic field. The B-coil ensures the measurement of the emf signal . The magnetic induction and magneti- zation are deduced after time integration. The last stored signal is the signal delivered by the force cell and propor- tional to the uniaxial applied force and stress . signal is the expression of a change in magnetization following Lenz’s law (1), where and respectively denote the number of turns of the B-coil and its area ( is the permeability of air). Magnetization is shown to be sensitive to variations of stress or magnetic field that can be experimentally observed by applying to the material a variable magnetic field with a fixed mechanical stress or applying a constant magnetic field and changing the mechanical loading. The former corre- sponds to the magnetic behavior under static stress ; the later is the piezomagnetic behavior . Dynamic (cyclic) measurements provide the hysteretic behavior. The anhysteretic behavior can be deduced from the reversible (anhysteretic) measurements (see [5] for instance). When magnetic field remains constant, the magnetization can be estimated from a direct time integration of signal (2). (1) 0018-9464/$26.00 © 2010 IEEE