FREE VOLUME CHANGE OF METALLIC GLASSES STUDIED BY THERMAL EXPANSION MEASUREMENTS

(1)

HAL Id: jpa-00220322

https://hal.archives-ouvertes.fr/jpa-00220322

Submitted on 1 Jan 1980

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

FREE VOLUME CHANGE OF METALLIC GLASSES STUDIED BY THERMAL EXPANSION

MEASUREMENTS

E. Girt, P. Tomić, A. Kuršumović, T. Mihać-Kosanović

To cite this version:

E. Girt, P. Tomić, A. Kuršumović, T. Mihać-Kosanović. FREE VOLUME CHANGE OF METAL- LIC GLASSES STUDIED BY THERMAL EXPANSION MEASUREMENTS. Journal de Physique Colloques, 1980, 41 (C8), pp.C8-875-C8-877. �10.1051/jphyscol:19808216�. �jpa-00220322�

(2)

JOURNAL DE PHYSInUE CoZZoque C8, suppZdment au n08, Tome 41, aoat 1980, page C8-875

FREE VOLUME CHANGE OF M E T A L L I C GLASSES S T U D I E D BY THERMAL EXPANSION MEASUREMEYTS

E. Girt, P. ~omi;, A. ~ur8umovid and T. ~iha:-~osanovi;

I n s t i 3 c t e o f Physics, Sarajevo, YugosZavia.

Abstract. - A dilatation set up was designed and used to investigate the low temperature relaxation of morphous Fe4gNi40L320 and Fe Ni P B through observed changes during isothermal annealing.

40 40 LG 6

In addition to t e an lcipated contrac lon on annealing, these amorphous alloys exhibited peaks in their dilatation curves brought about by a simultaneous but temporary expansion during structural relaxation. Although there was a considerable difference in the length change of Fe40NiGpBfQ a;d Fe Ni P B samples, the low temperature relaxation behaviour of both alloys was qua11 a lve y sit?la$OaAj the relaxation parameter could be described by an equation of the form :

Introduction. - During liquid quenching of a metallic glass the kinetic nature of the glass transition results in the departure of the fictive glass transition temperature, Tf, from its quasistatic counterpart, T as Tf is proportional to the quenching

g'

rate. Glassy material produced at a lower quenching rate naturally has a lower energy state and hence a lower molar volume. Glass quenched at a higher rate is metastable in two ways, firstly with respect to its lowest energy state and secondly with respect to its crystalline counterpart. During subsequent heating at temperatures insufficient for crys- talization this glass structure will relax towards more stable glassy state with lower energy and higher density. It was found that changes in many physical properties were brought about by structural relaxation and these could be used to monitor its kinetics. The large changes occuring in certain magnetic1 andmechanical properties may be in parti- 2 cular conveniently used for this purpose. Density or volume changes associated with such a relaxation, due to annealing-out of excess free volume3, are very

~ r n a 1 1 " ~ ' ~ ( 2 . 0,5 %) and difficult to measure. Since length changes are directly proportional to volume changes (~AI/I%Av/v), in this paper, we present an investigation of low temperature volume relaxation monitored by changes in length during annealing.

Experimental.- Glassy Fe40Ni40B20 and Fe40Ni40P14B6 samples were prepared in the ribbon form about 40 thick and 2 mm wide by melt spinning6. The dilatome- ter used for measuring the length changes during isothermal heating of these amorphous samples was a modification of the apparatus7 constructed for

thermal dilatation measurements of metallic ribbons.

The sample was subjected to a homogeneous temperature field with the variation of approximately

+ 0,04 K. Due to its low heat capacity the samples were able to reach furnace temperature in less than one second. The subsequent changes in the sample length were transmitted to the LVDT transducer connected to a nanovoltmeter (Keithly Y-180). This instrument, which enabled extremely accurate scan- ning of discrete dis~lacement values, was connected to the y-axis of a chart recorder. A time base of 25 s/cm was used in the present investigation. Con- sequently this isothermal annealing experiment provided direct observation of the complete process and data for further numerical analysis.

In order to keep the ribbons straight during measu- rements a tensile stress of about 0,3 N m was ap- -2 plied, which had no influence on low temoerature annealing process 5

.

The measurements were carried out at temperatures below 600 K in order to avoid the introduction of effects due to creep at higher temperatures.

Sesu1ts.- Figures 1 and 2 show relative length change, A 111, for Fe40Ni40B20 and Fe Ni P B metallic

40 40 14 6

glasses during isothermal annealing. The alloy containing phosphorus and boron showed faster annealing- out of excess volume than that containing only boron, indicating that the former is less stable towards low temperature relaxation. As shown in figure I the relative l e n ~ t h change were found to be an increasing function of time and temperature, the almost instan- taneous contraction at higher temperatures (580 K) being ti~ical of log(t) kinetics8. However the

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:19808216

(3)

JOURNAL D E PHYSIQUE

initial kinetics of length change at lower temperatures appeared to vary exponentially with time. The peaks P 1 and P observed in All1 isothermal curves

2

are possibly responsible for "minima" in thermal dilatation curves previously observed5 during isochronal annealing. There appeared to be no direct relationship between the relatively small amounts of expansion and the typical contraction occuring durin~

structural relaxation. Furthermore the peak widths and positions were found to be temperature dependent, decreasing and shifting towards shorter times with increasing temperature.

Figure 3 shows the results from figures 1 and 2 plot- ted on logaritmic time scale, neglecting the effects of these small expansions, from which it may be de- duced that the densification during low temperature structural relaxation for these metallic glasses obeys logaritmic time kinetics. These results are in agreement with the kinetic behaviour observed by both ~ ~ a m i ~ and chen1° in their investigations of structural short range ordering during low temperature relaxation.

Our results could be explained on the basis of kinetics of stuctural relaxation proposed by ~ ~ a m i ~ and chen1° where they showed that the observed structura3 relaxation parameter (in our case: ARv= 3A1/1) could

where :

Rvo is the amount of relative volume change that occurred during quenching resulting from finite quenching rate.

R is the actual relative volume change during relaxation. k is Boltzmann constant and a and c are cons- tants.

T is the annealing temperature.

The term aR in equation (2) represents the activa- vo

tion energy for start of volume relaxation during annealing of as-quenched samples. This was calculated from ln(to/T) vs. 1/T plot in figure 4 to be of the order of 100 KJ/mole. Further stages of volume relaxation were found to require higher activation energy proportional to the amount of the actual relaxation, Rv, i.e. increasing with amount of densification.

Assuming that the observed expansions were not directly affected by the extent of contraction their activation energies of the order of 50 KJ/mole was determined from

h(h)

^VS.1/T plots (figure 4) to be constant and therefore independent of the extent of

the relaxation.

Discussion.- The preliminary measurements of length changes during the isothermal annealing of Fe Ni B

40 40 20 and Fe Ni P B amorphous samples are consistent

40 40 14 6

with our earlier results obtained on Fe Ni B 40 40 20 loy by isochronal annealing. A clear contraction of the samples was observed together with a certain amount of unexpected expansion. Although our present measurements do not extend to very long times and high temperatures the kinetics found for annealing-out of excess (free) volume is clearly logaritmic in time, which agrees with other r e ~ u l t s ~ ' ~ ~ for structural short range ordering in metallic glasses.

It is worth notong that similar kinetic behaviour was determined for annealing out of voids introduced in aluminum during neutron irradiation 8

.

We also wish to emphasise the occurence of small peaks ( an increase in the length ) in All1 isotherms of our samples at longer times of annealing, which are not closely connected with densification. These peaks occurred at longer times for lover annealing temperatures.and they also showed rather a complex nature of the low temperature structural relaxation in metal- licglasses- A c o m p l e t e p h y s i c a l e x p l a n a t i o n f o r this behaviour will be presented in paper to follow.

REFERENCES

El] H.H. Liebermann, C.D. Graham, Jr., and P.J.

Flanders, IEEE Trans. Mag. MAG-13 (1977) 1541.

[d H.S. Chen, J. Appl. Phys. 2 (1978) 3289.

[3] M.H. Cohen and G.S. Grest, Phys. Rev. B (1979) 1077.

[d Y. Waseda and T. Egami, J. Mat. Sci. fi ⁽¹⁹⁷⁹⁾

1249.

[d A. Kuriumovi6, E. Girt, E. ~ a b i 6 , N. ~juhovi.6 and B. ~eontii, J. Non-Cryst.So1. submitted for pu- blication.

161 H.H. Liebermann and C.D. Graham, Jr., IEEE Trans.

Mag. MAG-12 (1976) 921.

[7] E. Girt, A. ~urxumovic' and T. ~ihaf-Kosanovi6, J. Phys. E : Sci. Instr. - 13 (1980) in press

E. Girt, P. ~omi: and A. Kur~umovic', to be published ( J. Phys. E : Sci. Instr. ) .

R.A. Vandermeer and J.C. Ogle, Acta Met. 2

(1980) 151.

[9] T. Egami, J. Appl. Phys. 50 (1979) 1564

[ld

H.S. Chen, J. Appl. Phys. - 49 (1978) 4595.

(4)

Figure 1.- Relative length change of as-quenched Fe40Ni40B20 for isothermal annealing.

lo'

Figure 3.- Relative length change vs. log(tJ of as- quenched Fe40Ni40B20 and Fe Ni P B during iso- thermal annealmg. 40 40 14 6

Figure 2.- Relative length change of as-quenched Fe Ni P B for isothermal annealing.

40 40 14 6

Figure 4.- Plots of ln(t /T) vs. 1/T for contraction and ln(t ) for expansionoyielding corresponding activatign energies.