Effect of Impurities on the Internal Friction Spectrum of Magnesium after Plastic Deformation

We have studied the impurity effect on the internal friction spectrum, with two different samples magnesium 4N and magnesium 6N containing zinc and iron : for the sample 4N we observe a broad peak depending on the oscillation strain amplitude : this peak shifts towards low temperature when the amplitude increases. for the sample 6N containing zinc and iron, the internal friction spectrum is different form the one of the 6N sample previously mentioned. After annealing at 250 K we observe, at low strain amplitude, besides the B1 and B2 peaks, txo other peaks situated around 100 and 150 K. But at a high strain amplitude we find the two peaks B1 and B2 as in the 6N sample. After annealing at 500 K the B1 and B2 peak are not observed but a large peak appears whdch shifts towards low temperaturewith increasing€ This peak is specific on thermomechanical unpinning of dislocations. m 1.Introduction.The internal friction spectrum of a high purity magnesium after ~tasticdeformation is very similar to the one observed in f.c.c. metals. Indeed, the hroad peak observed for the less pure materials [I-2 ] is replaced by several peaks in the case of pure magnesium [3-41 . There are : a)two peaks situated around 40 and 80 K labelled B1 and B2 respectively. They are stable during annealing. b) two peaks situated around 105 and 220 K, called P1 and Pg respectively and removed during temperature cycling. C) a peak at 15 K or an important decrease of internal friction between 10 and 30 K after annealing up to 300 K d) a stable peak near 420 K We have shown elsewhere that the two peaks B1 and B2 have the characteristics of the Bordoni relaxation [5,61 i.e. : they are relatively stable during annealing their maximum temperature depend weakly on the microstructural state and obeys the Arrhenius relation. their relaxation strength presents a maximum as a function of stress behaviour. In this paper we have studied the impurity effect on the internal friction Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1981522 C5-146 JOURNAL DE PHYSIQUE 2. Experimental results.Three differents polycrystalline samples have been used : a high purity magnesium noted flgl with one at.PPm of impurity (99,9999 %). a less pure magnesium noted Mg2 with C 100 at.PPm of impurity (99,99 %). a high purity magnesium noted Ng3 containing zinc and iron (20 at ppm zinc +25f 13 at ppm iron). After drawing, the samples were annealed for two hours at 300°C under argon atmosphere (grain size = 1 mm). The internal friction have been performed with an inverted torsion pendulum oscillating at a frequency of about 1 Hz, with a maximum strain amplitude equal to 410-~. The specimens have been deformed in situ by torsion of 2 % at 10 K then the internal friction was recorded between 10 K and 500during linear heating at a rate of 60 K/h. Fig.1, 2 and 3 show the internal friction spectra as a function of temperatrue between 10 and 250 K, for the three samples Mgl and Mg2 and %3. Immediately, after plastic deformation at 10 K (fig.l), we observe : for the Mgl sample (curve 1) : two small peaks B1 and B2 situated around 40 and 80 K, and an important peak PI situated around 105 K. Furthermore we observe a peak P2 situated about 220 K. for the Mg3 sample (curve 3) : there are a relatively large peak at about 40 K and a peak (may be PI) at about 105 K. for the Mg2 sample (curve 2) : there is a large maximum situated between 30 and 120 K. It must be noted that the internal friction background is important for Hg2 and Mg3 samples. After heating up to 250 K (fig.2) we observe essentially : for Mgl sample, two peaks B1 and B2 at about 40 and 80 K and a small peak 111 at about 15 K (curve 1). for Mg3 sample (curve 3) two small peaks (may be B1 and B2) at about 40 and 65 K, but the internal friction background is relatively strons. for Mg2 sample (curve 2) always a large maximum is present between 40 and 120 K and also the internal friction background is very important (we can compare the internal friction at 150 K for the three samples). Fig. 3 presents the internal friction spectra after annealing at 400 K for the three samples ; one can note : -forthe Mg,sample (curve l), the two peaks B1 and B2 have disappeared and an important decrease of the internal friction is noticed between 10 K and 30 K, the pic II, being annealed. for the Mg2 and Mg3 samples, we observe a relatively large maximum at about 40 K (for Mg3 ; curve 3) and 25 K (for Mg2, curve 2). The internal friction background is more elevated for the llgg sample than Mg3and especially an increase of the internal friction background is observed at about 150 K for Mg2 sample. We have shown elsewhere [61 that the internal friction background is very sensitive to the oscillating strain amplitude E, .In order to eliminate the internal background effect we have measured the internal friction at a small amplitude (some -6 10 ). The fig. 4 shows the internal friction spectrum of the Mg3 sample after plastic deformation at 10 K (2 2) and annealing at 250 K. On this figure one can note that besides the B:1 and B2 peaks, two other peaks situated around 100 K and 145 K appear. One peak seems to shift to lower temperature with increasing the strain amplitude. After heating upto 500 K (the B1 and B2 peaks have been already disappeared) we have studied the variation of the internal friction as a function of the strain amplitude (E ) for different temperatures (Fig. 5). The curves 6= f ( ~ ) shift m rn towards low temperature with increasing the measurement temperature (curve 1 to 4) and the internal friction maximum has about the same value. 3. Discussion.Wa have shown elsewhere that the B i and B2 peaks observed in a high purity magnesium after plastic deformation have the same features as the Bordoni relaxation in the f.c.c. metals. For a less pure magnesium, the internal friction spectrum is constituted of a large maximum between 40 and 120 K (fig. 1 and 2). This maximum has been observed by several authors in a less pure magnesium [1,2]. In a magnesium 6N containing zinc and iron, at a low measurement amplitude, we observe besides the Bordoni peakssome other peaks(Fig.41.We can consider then, the large maximum observed in a less pure magnesium can be the superposition of two or more peaks. This idea has been already proposed by TsuiandSack. This maximum shifts towards lower temperature when measurement amplitude E increases [ 6 ] . Thus, this maximum can be attributed to a m thermally activated breakaway process and we try to analyse the results by the T.G.L.