·Home ·Table of Contents ·Materials Characterization and testing

The Emission Acoustic Signals Registration for the Materials Brittleness Estimation D. Grabco, R. Jitaru, V. Rahvalov, A. Mihaes, D. Leu Institute of Applied Physics, Academy of Sciences, Academy str. 5, MD-2028, Chisinau, Moldova. E-mail: daria.grabco@phys.asm.md Contact

Abstract.

The device based on the registration of the acoustic emission signals under concentrated load action (microindentation and scratching) has been proposed. The possibilities of this apparatus for the non-destructive testing of the materials mechanical properties (strength and brittleness) and of the influence of various factors (preliminary deformation, temperature, aging, corrosion, etc.) on these parameters have been shown.

Introduction

Method of acoustic emission (AE) allows obtaining the information about the material strength properties, to estimate the surface layer brittleness, to study the various factors influence on the mechanical properties. The AE extensive application is based on the concept that the deformation and failure processes are the sources of the acoustic signals bearing the profound information about the internal structure transformation.

Fig 1: Schematic diagram of the apparatus used for the acoustic emission signals registration at the concentrated load action [4].


Earlier the method of the AE signals registration under concentrated load action (microindentation and scratching) was proposed in our laboratory by Boyarskaya, Kats and co-authors (Fig.1)[1]. In the case of static indentation the AE signals are registrated under both indentor penetration into material (N₁) and removal from it (N₂). The AE signals summary account, adequate to a certain length scratch, as a measure of brittleness was taken for the sclerometer testing. A special device based on this principle for the materials brittleness check was developed and fabricated in laboratory [2, 3]. This device represents a non-destructive testing method allowing the integral estimation of the mechanical properties of the large surface and volume regions in the different materials (metals, semiconductors, glasses, minerals, etc.) to obtain[4].
The enlargement of the device possibilities due to use of the more sensitive piezoelectric transducers and of a system loading modification permitted the number of investigated problems to be increase. It was studied the influence of the various factors (as a preliminary deformation, self-correction of the deformed zones during long rest or aging, etc.) on the some parameters characterized by the acoustic effects. The introduction of the K=N₂/N₁ parameter allowed a degree of the elastic deformation relaxation during sample unloading to estimate. This made possible the more deep study of the deformation mechanism under concentrated load action.
As known, the acoustic signals emission is connected with the dislocation movement, generation and annihilation of cracks, material failure[5-10]. Such a result was established for the silicium crystals: the AE signals were occurred on the samples, contained the dislocations and were absent on the samples free from dislocations [11, 12]. The AE signals occurrence in Si samples the authors have to link with the failure and high-rate movement of the dislocation ensembles under effect of both electrocurrent and thermoelastic stresses [12]. The AE signals origination under both the indentor loading and unloading was observed by a number of scientists and identified with the formation and further progress of cracks[4, 12].
So, the data from literature show the interrelation of the crystal defects and emission of the acoustic waves. Therefore it can be expected the external action on the crystal must be followed by the AE changes, resulting from the suitable dynamical structure rearrangement. However the same relation is insufficiently investigated yet.
With a view to show the possibilities of the mentioned AE method and to estimate the influence of the preliminary deformation degree, as well as, of others factors (temperature, aging, corrosion, etc.) on the intensity of the acoustic emission arising under indentation of a number of materials (MgO, InP, metals and alloys) were studied.

2. Experimental procedure

In the present paper the AE signals registration was carried out under indentation of the MgO (100) and InP (111) single crystals, as well as, of the water-pipe Fe alloys, pure and undergone to corrosion. These experiments were performed by means of the PMT-3 device, comprising the special AE installation (see Fig.1). The instrument allows to registrate the AE signals separately during the indentor penetration and standing in the sample (N₁) and under removal from it (N₂). The latter enabled us a N₂/N=K parameter to estimate, where N=N₁+N₂ is equal to number of AE signals during the all indentation process.
The Vickers diamond pyramid was used as an indentor. The indentor load (P) was varied in the range from 20 to 500g. Preliminary deformation of samples was performed by scratch method. The deformation degree and accordingly the internal mechanical stresses were varied by the change of the distance between scratches within the range 20 - 120 m m. So, depending on the distance between scratches the crystal regions with the various strength degree were formed.
In addition to this the different temperatures of preliminary deformation (T=293¸ 723K) as well as the long storage (aging) of the samples at 293K during the time (t=360¸ 720 hours) were also employed for the sake of the stress state change.

3. Results and discussion

The investigations showed the all N, N₁ and N₂ parameters increase with the load rise. For example, Fig. 2 illustrates the data for the summary account of AE signals N, obtained on (001) plane of MgO single crystals for two deformation orientations: d_ind. - indentation diagonal ï ï <100> and d_ind.ï ï <110>. The N₁ (P) and N₂ (P) dependencies are analogous to the N (P) ones, Fig. 2.

Fig 2: The dependence of AE signals summary account on the load for (001) plane of MgO crystals.

Fig 3: The K parameter dependence on load value for MgO single crystals. The indentation time, t, s.: curve 1- 2 and curve 2 - 1200.


So, the number of AE signals accompanying the indentor penetration and its unloading (removal from sample) is linearly increased with the external deformation force increase. This result is quite regular, as the more external load the greater deformation and accordingly the greater the amount of AE sources arise.
However the nature of the deformation processes is more clear from the data of the K=N₂/N parameter behaviour studying, Fig. 3. As mentioned the K parameter is the measure of the stress deformation relaxation, defined by AE signals registration. As it is seen from Fig. 3 the K (P) dependencies exhibit two-stage character: K sharply decreases with the P increase in the region of relatively small loads (P<100g) and weakly changes for P>100g, Fig. 3.
The obtained result allows concluding that the load dependence of the relaxation measure in indentor unloading is not unequivocal. But relaxation degree (the value of the K parameter) is closely connected with the relaxation power of the elastic stresses, arising during indentor penetration, before indentor unloading. And the more K values the more the relaxation under indentor lifting from sample takes place. However this fact in its turn is determined by the stress state, developing before indentor removal from sample.
So, the obtained K(P) dependencies are probably accounted for by the sharply increase of the gradient of elastic stresses, their specific contribution and deformation heterogeneity within the indentation with the load decrease. Therefore the state of the deformed material is strongly non-equilibrium and this tends to the accumulated energy relaxation during indentor unloading and the more stress gradient the greater the relaxation. This explanation, perhaps, well correlated with the K(P) dependencies in the P region from 20 to 100g, Fig. 3.
Under the high loads the volume of the plastic deformation is increased, the stress distribution becomes more homogeneous: that all is followed by the decrease of the relaxation process contribution into total deformation under microindentation. These considerations conform with the above given description of the K factors as a measure of relaxation under microindentation: the greater K the more significant are the relaxation processes.
It was established for MgO crystals that the measure of elastic stress relaxation during unloading decreases with the increase of plastic deformation volume appropriate with the indentor load rise, Fig. 3. This effect is pronounced in the region of relatively small loads (P<100g), Fig. 3.
So, the changes of the K magnitude allow roughly to judge about the deformation stress state of single zones of some materials: it is the elastic-stressed or more plastic ones.
The conclusion above can be made from the data of AE signals measurement under indentation time variation. The microindentation was proceeded during the t=2s (fast indentation) and during the 300, 1200 and 3600 s (continuous indentation).
The curves 2 in the Fig. 3 resemble the K values for continuous deformation during the t=1200s. It is seen that the curve 2 is mainly situated lower than curve 1. That means that the processes, occurred under continuous indentation, lead to the K decrease in comparison with the K values for the fast indentation. However, to establish the regularity of K change with t change failed. It will be noted that the number of AE signals (N, N₁) is mainly decreased too, but not for all loads. The creep processes, perhaps, follow the continuous indentation under constant external load, self-regulating relaxation of the stress state of crystal deformed regions. Therefore the partial relaxation of the elastic stresses during the indentor standing in the sample takes place. As a result the elastic strain degree in comparison with the fast indentation is decreased. This effect probably results in the ratio K decrease.
So, and in this case the decrease of the relaxation degree (K) during indentor unloading well correlated with the elastic stresses contribution decrease. The latter can be caused by the healing of the available cracks, generation of the new type ones, the start-up of the dislocation ensembles and high speed of their movement, the instantaneous rearrangement of some dislocation groups, etc. The obtained result shows that the K changes recognize the difference between the fast deformation and aged one during the material operation.
The established effect is most probably related to the change of AE sources nature in the last instance in comparison with the first one. The obtained data allow to conclude that the power of the relaxation processes during indentor unloading result from the elastic stress state, formed during indentor penetration. The deformation conditions as well as the material prehistory influence the latter, too. So, the proposed method of AE signals registration under microindentation permits somewhat to judge about the material elastic properties, about the its wear degree.
Thus, it was revealed that the acoustic emission is strong connected with the elastic-plastic deformation ratio, arising under microindentation. This ratio can be dependent not only on an external action force, but also on the crystal start stress (wear degree) of material. The study of the preliminary deformation influence on AE was made on the {111} planes of the pure and Fe-doped InP crystals. As mentioned the preliminary deformation of samples was performed by scratch method. The indentor load for the scratches was equal to 20g. Depending on the distance between scratches (d) the different stress degree of crystals were formed: AE is decreased with the d increase.

Fig 4: The summary account of AE signals (N) dependence on the preliminary deformation degree for InP crystals during all indentation cycle (1), indentor penetration into sample (2) and indentor removal from InP (3). Predeformation temperature T, K: 293. Pind.= 100g


From Fig. 4 it is seen that all N, N₁ and N₂ parameters are decreased with the d increase; that means that AE signals number fall with the reduction of the degree of the preliminary deformation at 293K of InP crystals. However the course of the N(d), N₁(d) and N₂(d) dependencies is different, Fig. 4. N(d) dependence shows the three-stage character: the AE signals number at the first stage (40-60mm) weakly changes; sharply falls in the region from 60 to 100mm (second stage) and increases significantly within the range of 100-120mm, Fig. 4, curve 1. This result evidences that not only the InP surface layer stress state influences the acoustic waves initiation and propagation, but also other factors take place. The N₁(d) dependence shows other behaviour: the N₁(d) curve practically smoothly falls down, Fig. 4, curve 2.
Such dependence course shows that the number of AE signals, emitted under indentor penetration, and the change of the preliminary deformation degree of InP crystals correlate well: decrease of the predeformation degree leads to the N₁ fall. The shape of the curve 3, Fig. 4, is analogous to the curve 1: the weak increase on the I and III stages and the sharp fall on the II one occurred. The established N and N₂ changes evidence that the acoustic waves emission is dependent not only on the stress state degree of the InP crystals preliminary deformation, as N₁ (curve 2, Fig. 4), but also on the type and structure of the deformation defects. The latter is especially important for the AE signals, arising during indentor unloading, while the impulse relaxation of the deformation stresses occurs.
The state of defects, generated under preliminary deformation, was varied by the scratch temperature change in region from 293 to 723K. The N (d) dependencies at 293; 423; 573 and 723K are presented on the Fig. 5. As it is seen the N (d) curves are situated quite well: the higher temperature the lower N (d) curve is placed. Therefore the common regularity for whole predeformation interval is revealed: the number of AE signals is decreased with the preliminary deformation temperature increase, Fig. 5. The established effect is pronounced in the region of small d (40-80mm), when the preliminary deformation degree of InP crystals in our experiments is maximum; and weak one for low predeformation degree (d=80-120mm), Fig. 5.

Fig 5: Dependence of the summary account of acoustic emission signals on the preliminary deformation degree for all indentation process. The deformation temperature, T, K: 1-29


From the Fig. 5 it can be also seen that number N is decreased with the d increase for all investigated predeformation temperatures. However this effect is reduced with the T increase and for T=723K is practically absent, cf. the run of curves 1 and 4, Fig. 5.
The last shows that the stress state, developed by the preliminary deformation at high temperatures (T³ 573K), slight influences the acoustic emission signals, investigated under indentation, while the stress state at 293K is significant in such processes.
The regularities, analogous to the mentioned above for N (Fig. 5), were obtained for AE signals number N₂, arising under indentor removal from sample: the discrepancy in this case was not found.
The assumption above that emission of the acoustic waves, arising under indentation of InP crystals, is dependent on the both deformation defects state and mechanical stresses is completely confirmed by the obtained results.
It is known that the type and structure state of deformation defects, internal stresses, homogeneity of their distribution are changed with the temperature change for all crystals, including InP ones [5-12].
Especially the surface layer of InP crystals is sensitive to the test temperature: the surface is oxidised and InP crystals are decomposed at T>700K [12].The type, number and structure state of the deformation defects with the test temperature change are affected too. So, under bending of InP crystals in the temperature interval 573-973K the authors of the paper did not observed the microtwins, while the twins formation in the mechanical testing is typical of InP crystals. With the temperature rise the dislocations number and their role in the deformation processes under microindentation of the InP crystals are significantly increased, i.e. the dislocation plasticity becomes more appreciable[13]. The obtained result the ever-increasing role of the dislocation mechanism under InP indentor deformation with the temperature increase from 293K to 693K evidences and it is well agreed with the conclusion of the article [13]. Alternatively, the plasticity enhancement is followed by the elastic stresses relaxation. Therefore, it can be regarded that the elastic stresses of the deformed InP crystals are reduced and also the deformation defects spectrum is changed in the test temperature rising. This allows to propose the possibility of the elastic tensions initiation in the InP crystals as a result of the predeformation at 423K, that gives rise to the elastic waves emission. That is quite normal that such state of the mechanical stresses influences the acoustic emission, arising under indentor penetration. Additional deformation, introduced during indentor penetration, perhaps, assists the stress state change of the deformed regions of the InP crystals. Accordingly the AE signals number under indentor removal from the samples the other character of the temperature dependence revealed.
One more possibility of the AE method may be demonstrated by the corrosion monitoring of the metals and alloys. The dependencies between the applied load and the summary account of the acoustic emission signals are presented in Fig.6

Fig 6: The dependence between the applied load and the AE signals amount arising in the Fe alloys at microindentation. a - surface of the pure Fe alloy; b - surface of the Fe alloy affected by corrosion.


Figure 6 shows the difference in the values of the AE signals, generated in the pure Fe alloy (Fig.6 a) and in the Fe alloy undergone to corrosion (Fig.6 b). The amount of the AE signals is bigger for the samples with corrosion than for pure metal. To note that this effect takes place at both the indentor penetration (N₁) and its removal (N): N₁ and N of the pure alloy are less than those parameters for the corroded one.
The results obtained due to use of the apparatus schematic presented in Fig.1 illustrate the wide possibilities of the AE method for the various physical phenomena and regularities investigation.
At present our efforts are oriented toward the further progress of the acoustic equipment: to adapt a block of registration of the frequency spectra of AE signals, generated within the deformed region under external load action. This problem solving will promote to obtain added information about the physical processes, accompanying the material failure; besides, it made possible to estimate the plastic deformation contribution into the total deformation process, to define the material fatigue character, to recognize a pre-failure stage of materials in operation, etc.

References

1. Boyarskaya Yu. S., Averbuch I. I., Kats M. S., et al // Zavodskaya Laboratorya, 1974, v. 40, pp. 550-553.
2. Boyarskaya Yu. S., Kats M. S., and Purich E. I. //Procceding of the 1^st Conference "Acoustic Emission of Materials and Constructions", Rostov-na-Donu, 1984, pp. 46-47.
3. Kats M. S. //Zavodskaya Laboratorya, 1975, v. 41, pp. 113-115.
4. Yu. S. Boyarskaya, D. Z. Grabco, and M. S. Kats. Physics of the Microindentation Processes // Kishinev, "Shtiintsa", 1986, 294 p. (in Russian).
5. Ya. M. Olih. Acoustic Emission in Halcogenic Glass Gl_0.18As_0.28Sl_0.5 // FTT, 1998, 40, No.9, pp. 1623-1626.
6. C. S. Lee, V. M. Rhgim, D. Kwon, et al. Acoustic Emission Measurement of Fatigue Crack Closure // S'er. Met. et mater., 1995 32, No.5, pp. 701-706.
7. Yu. V. Milman, V. A. Goncharuk, and V. V. Danilov. Badania Wlaschiwosci Mechanicznych Ceramiki Metoda Emisji Akustycznej //Ceramics (Polish ceramic bulletin 7, Krakov), 1994, 45, pp. 197-211.
8. Yu. S. Boyarskaya, R. P. Zhitaru, M. S. Kats, et al. Study of Borum Layers and Ionic Crystals by the AE Method // Proceeding "Physics of the Composite Material Strength", LINF, Leningrad, 1979, pp. 228-234 (in Russian).
9. B. V. Tanikella , Scattergood O. Ronald. Acoustic Emission During Failure Process under Indentation // J. Amer. Ceram. Soc., 1995, 78, No.6, pp. 1698-1702.
10. Gul Kim Kwang and Sachse Wolfgang. Characteristics of Acoustic Emission Signals of Hertzion and Unloading Cracks in Glass // J. Appl. Phys., 1984, 55, No.8. pp. 2847-2856.
11. A. M. Orlov, A. A. Skvortsov, and V. A. Frolov. Acoustic Emission in the Dislocation Silicium by Current ant Thermal Affects // J. Pisma v ZhTF, 1999, 28, No.3, pp. 28-32, (in Russian).
12. Haruo Nadaj. Dislocation Velocities in Indium Phosphide // JJAP, 1981, 20, No.4, K793-K794.
13. D. Z. Grabco, V. S. Panfilov, and E. E. Maronchiuk. Micromechanichal Properties of the InP Single Crystals Doped by Iron // Proceeding "Indium Phosphide in the Semiconductor Electronics", "Stiinta", Kishinev, 1988, pp. 251-256.

© AIPnD , created by NDT.net

|Home|    |Top|