PSI - Issue 26

V.N. Kytopoulos et al. / Procedia Structural Integrity 26 (2020) 113–119 V.N. Kytopoulos / Structural Integrity Procedia 00 (2020) 000 – 000

115

Fig. 1. Block diagram of the experimental MBE – setup

Table 1: The chemical composition (%) of the low – carbon steel used for the construction of the specimens

C

Mn

Si

Ni

Cr

Mo

S

P

N

0.05

0.40

0.015

0.017

0.01

0.001

0.014

0.01

0.003

All corrosion experiments were performed at room temperature (25 o C). The exposure time in the SSF-apparatus were 200, 400, 600, 800, 1000 hours. The final corrosion product formed on the specimen surface was in a greatest part, as obtained by X-ray diffraction analysis, a ferric oxyhydroxide (FeO(OH)) component. The produced corrosion layer was removed by dry air blast and soft natural bristle brush revealing underneath a black and strong adhering layer in form of magnetite (Fe 3 O 4 ) product. Afterwards, the magnetite layer was removed by appropriate chemical etching procedure. Otherwise, this layer obscures the investigating hydrogen effect. It is pointed out that during corrosion a parallel pro duction of magnetite and atomic hydrogen takes place by the following electrochemical reactions (Moeller et al. 2006): Fe = Fe +2 + 2e - , Fe +2 + 2FeO(OH) = Fe 3 O 4 + 2H + , H + + e - = H It is known that for non-destructive investigations of ferromagnetic materials using the Barkhausen ME-response among several measurement parameters the common count rate and corresponding root-mean-square voltage, V rms , signal are mostly used (Stefanita et al. 2000; Sulliran et al. 2004; Blaow et al. 2007). It is noted that the mean square is closely related to the energy rate of the micromagnetic events. Nevertheless, some apparent ‘problems’ may arise by using these parameters separately. This is because the V rms as a single parameter reflects the qualitative behavior of micromagnetic activity whereas count rate the quantitative one. To overcome, at least in part, this apparent ‘inherent weak ness’ we propose to couple both behaviors by using an operational variable parameter J=V rms /N, where the energy rate, V rms is the measured root mean-square voltage and N the corresponding detected number of micromagnetic events or counts (pulses). This means that this parameter has dimension of energy rate per micromagnetic event and as such reflects a kind of specific energy indicative of the strength of an elementary jumping or displacement step of domain wall. Thus, the above parameter can be seen as a measure of the specific ME-response. It has been observed that an increase (decrease) in the number of pining sites or obstacles (N) would lead to an effective domain wall multiplication (reduction), associated with a decrease (increase) in the effective wall velocity and thus in the measured signal energy (Sulliran et al. 2004). From these factors one can finally deduce that an increase (decrease) , in the number, N, of pining sites or pinning density, would result in a decrease (increase), in the above ratio of J- parameter. However, as shown elsewhere (Sulliran et al. 2004), this would happen only in an applied stress-free or 3. Results and Discussion