PSI - Issue 10

D. Triantis / Procedia Structural Integrity 10 (2018) 11–17 D. Triantis / Structural Integrity Procedia 00 (2018) 000 – 000

14

4

log ( N

) log ( N

)    

  

 

(1)

I



1

2

b

(

)     

1

2

where σ is the standard deviation, μ is the mean value of the amplitude distribution, α 1 is the coefficient related to the smaller amplitude, α 2 is the coefficient related to the fracture level. The values of α 1 , α 2 vary between 0.5-5.0. Con sidering that changing the values of α 1 and α 2 did not significantly affect the I b -value, their value was here assumed constant and equal to 1 (α 1 =α 2 =1) (Rao and Lakschmi (2005)). In order to demonstrate the variability of the I b -value during the experiment, the total number of AE hits were divided into groups of sequential hits. Each group included n=100 hits from which the first 99 belonged to the previous group and the last one to the next. The I b -values that were calculated using equation (1), were attributed to a time value that corresponds to the mean time of the moments of occurrence of the sequential hits of each group. In Fig.1 the I b -value time history is depicted for all stages. During stage A the I b -values increase until the moment t=70 s, where the corresponding axial stress ( σ ) is not higher than 50% of the failure stress ( σ f ). Then, and until the time instant t=100 s ( σ≈0.70σ f ), the I b -values show a trend to remain constant at values higher than 2.5. This kind of I b -values variations during the first stages of loading may be attributed to AE generated due to the closure and friction phenomena of pre-existing microcracks in the material. At the following time instants (t>100 s) while the elastic de formation stage is practically exceeded, a gradual decrease of the I b -values starts, maintaining high values exceeding 2.0. This behaviour is continued until the time instant of t=130 s ( σ≈0.90σ f ). In this time period, the AE are mainly related to the generation of a number of new microcracks that are spread in the whole bulk of the tested specimen. While approaching the end of stage A, and as the stress reaches the value of 60.5 MPa ( σ≈0. 96 σ f ), a sudden decrease of the I b -values, is observed (see Fig.3) and a minimum value (Ib≈1.0) is recorded. This means that during this time period, a transitional state from formation, to growth stage of newly formed cracks takes place, at the end of which a preliminary crack-coalescence stage is possible.

150

2.5

2.0

100

hits per sec Ib -value

1.5

50 hits per sec

Ib - value

1.0

0

0.5

285

290

295

300

305

310

315

time (s)

Fig. 3. The acoustic activity (hits per second) and the I b -values during the last seconds before failure. .

During stage B, the I b -values continuously rise and after 35 s they are stabilized -with some fluctuations- between 3.0 and 3.3, given that the occurrence rate of AE decreases and the ΑΕ hits are characterized by lower amplitudes. A similar pattern is also observed in stages C and D, except that in stage D, for t>290 s, instead of a further increase of I b -value, a gradual reduction appears. During this final stage, the AE activity described by the number of hits per second, starts becoming more intense (see Fig.3). Two to three seconds before fracture of the specimen, and while the rate of the AE hits remains constant, the I b – values show a rapid decrease from the value of 1.5 down to values lower than 1 ( ≈ 0.85) at the instant of fracture. This behaviour clearly shows that abrupt events take place. These events indicate the existence of dynamic and unstable cracks that lead to the macroscopic failure.

3.2. The variability of the average frequency and RA value

Considering that the Acoustic Emission technique, can be used for the characterization of the typical crack modes, the Recommendation of RILEM TC 212-ACD was followed, that has been satisfactorily applied in concrete (Ohtsu