Printer Friendly

Acoustic emission behaviors of recovery for Mg alloy at room temperature.


Traditionally, the recovery of materials was investigated by the evolution of microstructure, mechanical properties, electronic or magnetic properties. In present research, the recovery behavior of AZ31B alloy was investigated by acoustic emission (AE) at room temperature. After deformation, the specimen was attached to an AE sensor with broad response frequency. Considerable AE signals were observed and numerous AE signals appeared at the initial stage of the recovery process. Theoretical analysis was conducted on the origin and mechanisms of AE. AE signals appear to be from the dislocation annihilations driven by the internal stress of materials after deformation. The results also show that the cumulative AE counts are in good agreement with a non-linear relationship derived from the traditional recovery kinetics.

Keywords: AE count, Elastic energy, Recovery, Mg alloy

1. Introduction

Acoustic emission (AE) is the radiation of stress waves by materials, in which local dynamic restructuring of the internal structure occurs with the release of elastic energy. During the second half of the 20th century, AE from materials was studied extensively, especially in the AE from deformation and fracture of materials. More recently, the application of AE was extended greatly in fields such as thermal cycling, in which dislocation motion and microstructure damage are generally recognized to produce significant AE, martensitic transformation where the AE signal was thought to accompany the rapid variation in the shape of deforming volume, solidification of liquid, weld controlling, and fatigue, etc. Nearly all of the mechanisms of AE can be explained by one of these two approaches: dislocation motion or elastic stress release by micro-fracture of materials. In general, the model of AE source by dislocation motion is adequate in explaining and predicting the AE behavior. Schaarwachter et al. [1] successfully applied dislocation model of AE in the deformation of copper and its dilute alloys in explaining the relation between the dislocation motion and the behaviors of AE and later this results were applied by Moorthy et al. in explaining the AE behaviors of solution-annealed AISI type-316 austenitic stainless steel [2] in detecting micro- and macro-yielding of the materials and predicting the AE response frequency by dislocation model. All of the above have shown that dislocation motion and AE are closely related with each other. So AE is generated from the motion of dislocations, or we can suppose that wherever there is dislocation motion, there will be AE from materials no matter how weak the signal strength is. It is this supposition that paves the way for the present research.

The term of recovery denotes any modification of properties during annealing before the appearance of new strain-free recrystallized grains. Investigations performed on monocrystals have proved that recovery takes places by means of annihilation of dislocations with opposite Burgers vectors, tightening of dislocations loops, ordering of dislocation configurations, as well as by means of formation of subboundaries and their motions [3]. Recovery in polycrystals, apart from intra-grain mechanism, can also take place by means of annihilation of dislocation, points defects on grain boundaries. From the above, we can conclude that recovery process is a source of dislocation motion and it is reasonable to expect that the recovery process should be accompanied with a release of elastic energy or acoustic emission. Traditionally, recovery behavior is characterized by the mechanical properties of materials such as strength, hardness, ductility or by the observation of microscopic mechanisms in dense dislocation tangles. The observation of microscopic mechanisms by electron microscopy is difficult for the large density of dislocations involved (between [10.sup.14] and 5x[10.sup.15]/[mm.sup.2]) and their heterogeneous distribution, which makes their energy density and their individual (cross-slip, climb, solute drag) or collective behavior difficult to handle either analytically using simple elementary dislocation models or by computer simulation. More recently, there is a growing interest in the characterization of microstructure by magnetic non-destructive techniques [4]. Structure sensitive magnetic macroscopic properties include coercive field, permeability, residual induction and power loss derived from a hysteresis loop. Recovery behavior of many materials can be evaluated properly by this method. However, like the method of mechanical properties, the detection of magnetic properties of materials is not so convenient for it has to be quenched each time after heat treatment in getting an instantaneous value of the property. That is to say, it is difficult to get continuous and synchronized behavior of recovery by this method. Furthermore, it is a time consuming and expensive process especially in some alloy containing noble metals because each datum shown in the recovery dynamic curve usually needs a specimen to be tested.

There has been much research about the recovery of Mg alloy [5, 6]. However, only few studies were specially conducted in the recovery of Mg alloy at room temperature. Koike et al. [7] reported that recovery can take place at room temperature by investigating the Kikuchi line and the deformation curve of the AZ31 Mg alloy. They analyzed the active role of dynamic recovery at room temperature to the deformation of Mg alloy. In fact, other researchers such as Miller in 1991 [8] and Conrad in 1957 [9] also proposed that the recovery of Mg alloy can take place at room temperature. The AE from phase transformation has been observed in many areas such as martensitic transformation, solidification of liquid, and so on. Many researches have shown that AE method is a powerful method in investigating the dynamitic behavior of phase transformation [10, 11]. However, until now, no research of AE was performed upon diffusion-controlled (thermally activated) structural transformations--recovery and recrystallization. So it is the aim of this research to get the continuous and synchronized AE signal for a better and more convenient material evaluation method in understanding the recovery behavior of AZ31B alloy at room temperature.

2. Experimental Procedures

The materials system chosen was extruded bar of standard AZ31B Mg alloy from Osaka Fuji Magnesium Alloy Co.. Cylindrical samples with 15 mm in diameter and 15 mm in height were preliminarily subjected to annealing at 450[degrees]C for 1h. The samples were deformed to different strain levels at a deformation rate of 1.5 x [10.sup.-2] mm/min. As soon as deformation was finished, the specimen was attached to an AE sensor. AE sensor used in present study was low noise type (M304A, Fuji Ceramics, Japan). AE signals were measured by [mu]DISP (PAC, USA) with a threshold of 40 dB and high-pass filter of 50 kHz effectively reducing the noise from environment. The RMS voltage of AE signal was also recorded by the AE discriminator system with a gain of 60 dB as shown in Fig. 1.



3. Results

Figure 2 shows the RMS voltages of AE signals during the recovery stage of AZ31B alloy at strain of 9.3% along with environmental noise level. Both the AE signals and noise were continuous waves. The RMS value of recovery decreased to the noise level eventually, which indicates that this AE system is effective in recording the AE signal from recovery.


It is interesting to note that all of the AE waveforms are similar as shown in Fig. 3. They are of low level and some kind of continuous and the peak frequency is about 550 kHz. AE from recovery is similar to that from deformation process in Mg alloy [12]. With increasing recovery time, the amplitude or count of AE signal decreased gradually as shown in Fig. 4.

Figure 5 shows the cumulative AE event as a function of recovery time. It can be seen that most of the AE event occurred at the beginning of the recovery process where a steep slope in the initial section (<2000s) exists. Later, the cumulative AE event increased slowly and reached a horizontal stage during the final recovery stage (>4000s). It is found that no systematic relationship between the strain of AZ31B alloy and the overall AE events during the entire recovery process exists. Probably this is due to the difference in the time of specimen transfer from the compression jig to AE sensor since the initial AE event rate is so high that a little difference of time can resulted in a large discrepancy in the overall AE events. We will discuss it later in detail.



4. Discussion

4.1 Theoretical Analysis of AE from Recovery

Only limited research has focused on the recovery of materials at room temperature [7, 8], even though recovery at room temperature is critical in affecting the deformation properties of materials. Some research focused on the evolution of yield stress during recovery stage [13, 14]. Friedel model [15] predicts a logarithmic decrease of the yield stress and assumes that applied stress is equal and opposite to the internal stress of the dislocation structure. If one assumes further that the relaxation rate of this structure, and that of the internal stress, occurs by thermally activated mechanism according to

d[[sigma].sub.i]/dt = -K exp(-U([[sigma].sub.i])/kT) (1)

where K is a constant, and that the activation energy decreases with stress U = [U.sub.0] - v[[sigma].sub.i], v being the activation volume of the elementary recovery events, then a logarithmic time decay of the yield stress is

[sigma] = [[sigma].sub.i] = [[sigma].sub.t=0] - (kT/v)ln(1 + t/[t.sub.0]) (2)

and the plastic relaxation strain rate from recovery is

[[??].sub.p] = [rho][b.sup.2][v.sub.D]/M exp(-[U.sub.0]/kT sinh([[sigma].sub.i]v/kT) (3)

where [v.sub.D] is the Debye frequency, M the Taylor factor, and the U0 and v the parameters to be decided.

From above equations we can think the recovery process to be a deformation process, in which the force is applied opposite to the direction of deformation and decreases with increasing time, resulting into a plastic relaxation of materials. If we assume that the internal stress relaxation corresponds to dislocation annihilation and reorganization, and therefore to plastic relaxation strain, the AE activities from recovery process can also be thought as a result of dislocation motions and annihilations applied by the internal stress.

During the recovery stage, it is also assumed that the recovery is controlled mainly by the disappearance of the dislocations with opposite Burgers vectors (although many phenomena occurred in this process such as tightening of dislocations loops, ordering of dislocation configurations), and the disappearance of the dislocation will give rise to micro-slip in the slip plane and at the same time a certain strain energy will be released which results in AE signal (some of the strain energy will be released as a form of heat). During the AE generation process, it can be thought that the elastic wave was produced just as the origin of earthquake waves [16].

4.2 AE Behaviors and Recovery Kinetics

Theoretical considerations indicate that some properties of the materials are proportional to the square root of dislocation density during recovery process, e.g., the flow stress ([sigma] - [[sigma].sub.y]) [14]. Take [R.sub.y] as the fraction of recovery at a certain temperature, then

[R.sub.y] = ([sigma] - [[sigma].sub.y])/ ([sigma] - [[sigma].sub.0]) [infinity] k [square root of [rho]] (4)

where ([sigma] - [[sigma].sub.y]) is the decrease of flow stress in recovery, and ([sigma] - [[sigma].sub.0]) the overall decrease of flow stress after recovery. Then a general relation is satisfied to a specific property of materials at a fixed temperature [17] as follows:

[R.sub.y] = a + b ln (t) (5)

where a and b are constants for the specific property at each given temperature at a fixed pre-strain. Figure 6 shows the relative cumulative AE counts curve of both the linear and logarithmic relationship with time. The relative cumulative AE counts increase with recovery time. In this figure we find a similar trend of AE at different pre-strain levels indicating the same mechanism of AE during recovery process.

Assuming the AE event counts proportional to the number of annihilated dislocation, the cumulative AE event ([N.sub.i]) have the same relationship with dislocation density as event counts, and the relative cumulative event counts, [N.sub.i]/N has the following relation with dislocation:

[N.sub.i]/N [infinity] K([rho] - [[rho].sub.r]) (6)

where [rho] is the dislocation density just after the deformation process, [[rho].sub.r] is the dislocation density during recovery. From equations 4, 5 and 6, we can get

[N.sub.i]/N = A [[ln(t)].sup.2] + Bln(t)+C (7)

Here A, B, C are constants. The exact meaning of these constants needs further research. The non-linear fit results of relative cumulative AE counts are shown in Fig. 7 at strains of 9.3% and 5.3%, respectively. It can be seen that the prediction of theory is in good agreement with the experimental data.


5. Conclusions

1. AE signals were successfully detected in the recovery of AZ31B alloy at room temperature with a high pass filter of 50 kHz.

2. Cumulative AE events depend on the recovery time. Most of the AE events were recorded at the initial stage of recovery (<2000s), and AE event rate decreases gradually with the increase of recovery time.

3. The cumulative AE counts were assumed to be proportional to the dislocation density that decreases during recovery. A non-linear equation describes the relation between cumulative AE counts and the recovery time.



[1] W. Schaarwachter, H. Ebener, Acta metal. Mater. 38 (1990), 195.

[2] V. Moorthy, T. Jayakumar and Baldev Raj, International J. of Pressure Vessels and Piping, 64 (1995), 161.

[3] T. Hasegawa and U. F. Kocks, Acta metal. 27 (1979), 1705.

[4] A. Martinez-de-Guerenu, F. Arizti, et al., Acta Materialia, 52 (2004), 3657.

[5] C. J. Beevers, Acta Metal., 11 (1963), 1029.

[6] J. V. Sharp, A. Mitchell, Acta Metal., 13 (1963), 965.

[7] J. Koike, T. Kobayashi, Acta Materialia, 51 (2003), 2055.

[8] W.K. Miller. Metall Trans A, 22A (1991), 873.

[9] H. Conrad, W.D. Robertson, J. Metals, 1957, 503.

[10] F. J. Perez-Reche, E. Vives, J. Phys. IV France, 112 (2003), 597.

[11] J. Baram, J. Avissar, Scripta Metallurgica, 14 (1980), 1013.

[12] K. Ono, Acoustic Emission-Beyond the Millennium, 2000, 57.

[13] M. Verdier, Y. Brechet, Acta Mater. 47 (1999), 127.

[14] E. Nes, Acta Metall. Mater., 43 (1995), 2189.

[15] J. Friedel, Dislocation. Pergamon Press, Oxford, 1964.

[16] K. Aki, P. G. Richards, Quantitative Seismology, theory and methods, Volume 1, 1980.

[17] A. Martinez-de-Guerenu, F. Arizti, Acta Materialia, 52 (2004), 3667.

Y. P. LI and M. ENOKI

Department of Materials Engineering, The University of Tokyo, Hongo, Bunkyo, Tokyo, Japan.
COPYRIGHT 2005 Acoustic Emission Group
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2005 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Li, Y.P.; Enoki, M.
Publication:Journal of Acoustic Emission
Date:Jan 1, 2005
Previous Article:Plastic region bolt tightening controlled by acoustic emission monitoring.
Next Article:An acoustic emission test system for airline steel oxygen cylinders: system design and test program.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters