Dislocations Point Defects Interactions in Magnesium

The interactions between dislocations and point defects have been studied in magnesium containing 100 p.p.m. of impurity atoms. The experiments where conducted on one hand by measuring the ultrasonic attenuation changes due to a bias stress and on the other hand by measuring the internal friction on a torsion pendulum working at 2 Hz. INTRODUCTION Since few last years the studies of variation of the attenuation versus a bias stress Aa = f(a) show a certain interest. This advantage with respect to the classical internal friction studies is due to the fact that the stress responsible for the dislocation motion ( 0 ) can be separated from that of measurement ; moreover the amplitude of measurement is very small because it is in the ultrasonic range. These ultrasonic studiesleads to an interpretation usually more convincing for the obtained analastic phenomena. The experiments has been done, until now, mainly on the c.f.c. metals as copper [I], aluminium [2,3] and recently some results were reported on a b.c.c. metal (tungsten 141 ). The aim of this investigation is to present the first results obtained on a h.c.p. metal ; moreover the results of Aa = f(a) are compared with the classical internal friction spectrum 6 = f(T) obtained with the same material. EXPERIMENTAL RESULTS 99,99 % polycrystalline magnesium has been used. The samples were annealed for two hours at 300°c in argon atmosphere ; their dimensions were 5 X 5 X 70 mm3 for quasi static stress experiments and $ 1 mm X 50 mm for internal friction measurements. For Aa = f(a) measurements, the tensile quasi static stress has been applied along the same axis than that of the ultrasonic longitudinal wave (17 IfHz). The internal friction measurements has been performed with an inverted torsion pendulum (2 Hz). Dislocations were introduced in the sample by equivalent plastic deformation at 100 K (0.2 % for tensile test and 0.8 % on the external fibre of the sample in pendulum experiment). At first, isochronal annealings were performed between 108 K and 320 K ; figure 1 shows the attenuation changes Aa = f(a) measured at 108 K for three typical situations : (i) fifteen minutes after plastic deformation, the attenuation increases a little then decreases regularly with the stress increasing up to 6. 106 ~ . m ~ . (ii) By annealing 15 minutes at 200 K, the main decrease in the attenuation has disappeared. (iii) Finally after a treatment at 320 K, a continuous and important increase in Aa appears. Furthermore, it must be noted that after these annealings at increasing temperatures, the hysteresis loop observed in the curve attenuation versus bias stress a is at first negative (case a) and becomes positive (case c). On an other hand, the temperature effect on Aa = f(a) after annealing at 320K have been studied. Figure 2 shows two main features : firstly there is a rapid increase in the variations of Aa between IOPK and 240K (for example : Aa(ao = 5106 ~ . m ~ ) at 240K/Aa(oo) at 108K = 16) Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1981559 CS-394 JOURNAL DE PHYSIQUE secondly a new mechanism interferes above 240K, because, on one hand there is a crossing point in the loading-unloading cycle, and on the other hand the amplitude of the attenuation change seems to be reduced again above 240K. 0 ) a f t e r l 5 mn 108 K b) a f t e r 15 mn 200K