Effetto della diffusione dell'idrogeno sui fenomeni di Environmental Assisted Cracking di acciai per pipeline in condizioni di protezione catodica

Pipeline steels can show Environmental Induced Cracking phenomena under slow straining with Hydrogen Embrittlement mechanism under cathodic protection [1]. Hydrogen evolution can take place due to cathodic protection normally used in order to protect the pipeline against general corrosion. The steel is polarized at cathodic potentials in the range -0.8 to -1.1 V vs SCE, but very negative values could be reached in overprotected areas close to the impressed current anodes. The hydrogen ions reduction reaction takes place on the metal surface, creating adsorbed hydrogen. The adsorbed hydrogen diffuses into the metal owing its solubility in the metal lattice. In this paper hydrogen permeation tests were carried out according to the electrochemical methods proposed by Devanathan-Stachurski [8] on different pipeline steels (Tab. 1 and Fig. from 1 to 4). The hydrogen diffusivity depends on microstructure of the steel (Tab. 2). Also the tensile properties depend from the microstructure of the steel, so it is possible to observe a correlation between the hydrogen diffusion coefficient and the yield strength of the different steels (Fig. 5). Because hydrogen mainly diffuses through pre-existent paths as ferritic grain boundary [13], the diffusivity increases for the control rolled steel with respect to hot rolled steel owing to their fine ferritic grain. In the tempered martensite steel, the effect of very fine structures is contrasted by very fine and dispersed precipitates that can act as traps subtracting hydrogen to diffusion. Owing to yield strength depends on ferritic grain dimension and formation of hardening precipitates, a correlation between yield strength and diffusivity can be expected. However, figure 5 also shows important differences in behavior between steel with similar microstructures. The hot rolled steels by recent production have much lower diffusivities than the X60 steel produced in '60s, that has rather high content of carbon and sulfur and coarse microstructure with very pronounced pearlite band. In order to explain such a difference, less efficient hydrogen trapping could be supposed. The hydrogen diffuses in the metal lattice and concentrates in plastic deformation areas like the crack tip. The EAC phenomena can take place after the hydrogen concentration reaches a critical value. The hydrogen diffusion coefficients were correlated to the EAC parameters obtained in the SSR tests and in the corrosion fatigue tests. In the SSR tests the environmental effects are related to the decreasing of the reduction of area (Fig. 6-7 and Tab. 3). The ratio between the reduction of area in environment and in air, called normalised reduction of area, is a quantitative index of hydrogen embrittlement: the HE phenomena are more pronounced as this values is lower than one. The fatigue curves crack growth in air are in agreement with the Paris' law, (da/dN)F = CxΔK". (Tab. 4). In synthetic seawater under cathodic protection, the crack growth rate becomes much higher than the rate measured in air, at intermediate values of stress intensity factor range and low loading frequencies (Fig. 8). The crack growth morphology changes owing to the environmental effects as shown in Fig. 9. The effect is more pronounced as the cathodic polarization increases from -900 to -1050 mV vs SCE. At high stress intensity factor range, the curves in air and in synthetic seawater approach together. Mechanical factors prevail under crack instability conditions, as stress intensity factor approaches material fracture toughness. The crack growth rate becomes higher than the correspondent value in air above a threshold (ΔKeth) that can be related with a critical value of the maximum stress intensity factor. According to the Model of Wei and Landes [10], such a value can be assumed as an estimation of critical stress intensity factor for SCC (KSCC) under cyclic loading (7) KSSC = ΔKeth / (1-R) Table 5 describes the estimated values of ΔKeth and KISCC. The steels show Kssc of about 30-40 MPa.m1/2, slightly decreasing as the cathodic polarization increases from -900 mV to -1050 mV vs SCE. According to the model proposed by Wei and Landes, the crack growth rate can be considered the sum of a pure fatigue contribution (da/dN)F evaluated by means of Paris' law, and a corrosion contribution (da/dN)c. Thus the contribution of corrosion can be evaluated by relation (9). (9) (da/dN)c = (da/dN)cF -(da/dN)F Fig. 10 shows the corrosion contribution to fatigue crack growth rate calculated by the experimental fatigue and corrosion fatigue curves at -1050 mV vs SCE, as a function of ΔK. As ΔK increases above 20 MPa.m1/2, the contribution due to corrosion approaches a constant value illustrated in the Tab. 6. It is well known that crack growth rate due to SCC is related to stress intensity factor (K). Crack growth occurs at almost a constant rate (daldt)scc above K SCC up to stress intensity factor approaches material fracture toughness. During fatigue cycle, SCC phenomena can only take place during the period of load increasing because during the decreasing period the crack tip is under compression. Thus, in order to estimate (da/dt)SCC by the experimental corrosion-fatigue curves the following relation can be adopted, valid for K always above KSCC during the loading cycle (that is ΔK>ΔK*), by considering the frequency (f): (10) (da/dt)scc = (da/dN)c·2·f At -900 mV vs SCE potential, the differences between the steels are comparable to the experimental scattering. At -1050 mV vs SCE, the steels with microstructures of banded ferrite with pearlite or bainite or martensite inside the bands presented increasing of the cracks rate with the mechanical properties. A similar behaviour, but less pronounced differences, was also observed for the martensitic steels. Fig. 11 and 22 relate the hydrogen diffusion coefficient with the results of SSR test and the crack growth rates for stress corrosion cracking estimated by corrosion fatigue tests. As hydrogen diffusivity increases, the effect of hydrogen embrittlement in SSR tests and crack propagation becomes more evident. Doubling in crack growth rates can be noted as the diffusivity increases from 2.7.10-7 Cm2/s to 5.10-7 Cm2/s at the potential of-1050 mV vs SCE. It usually assumed that Hydrogen embrittlement occurs through mechanisms involving hydrogen transport to the plastic strain zone at the crack tip and a critical concentration for failure initiation. An increase of hydrogen transportation to the plastic zone of a propagating crack is expected to arise crack growth rates, because the critical content of hydrogen can be reached in a short time. However, a complex situation has to be hypothesized as far as the relation between material susceptibility and hydrogen diffusivity. Literature data [14,15] confirm an increase of 4 times in crack growth rates of C-Mn steels during corrosion-fatigue tests in seawater under cathodic protection, at 20 MPa.m1/2 ΔK in the range from 0°C a 25°C. This is in accordance with the increasing effect of temperature on diffusion coefficient [16]. On the other hand, it is also well known that with an increase in temperature from room temperature the hydrogen deleterious effect minimizes, in despite of a considerable increase of diffusivity. Actually, hydrogen transport also depends on trapping into the metal and surface concentration of hydrogen adsorbed on steel. The surface concentration decreases with temperature because recombination reaction to produce hydrogen molecule is favored. Decreasing of probability of being trapped by reversible and irreversible traps, and increasing of hydrogen solubility in the lattice also occur. Under the environmental conditions adopted for testing, the surface concentration of adsorbed hydrogen mainly depends on cathodic polarization, in absence of any hydrogen recombination poison like sulfide. Thus, in order to interpret the experimental data a constant surface concentration could be roughly assumed. The data show an effect of microstructure that cannot only be related to diffusivity. Tempered martensite steels show low susceptibility to environmental effects during both SSR test and corrosion-fatigue test, owning to their fine microstructure with very fine and dispersed precipitate. These fine microstructures are able to distribute hydrogen over spread irreversible traps and reduce probability of reaching local critic hydrogen concentration for cracking initiation. Finally, the diffusivity measured in this research has been obtained on steel without any deformation during test. Hydrogen diffusion is dramatically influenced by the presence of deformation as found in previous research [17]. Thus, the correlation between steel susceptibility and hydrogen diffusivity shown in Fig. 11 and 12 could be related, at least partially, to the influence of microstructure both on diffusivity and critical condition for the initiation of environmental assisted cracking.