Studio della NN-SCC di acciai per tubazioni interrate tramite prove di flessione in tre punti

Near Neutral Stress Corrosion Cracking (NN-SCC) is a particular form of stress corrosion cracking that takes place on coated buried pipelines It occurs under particular circumstances, in presence of slow plastic straining, in areas of disbondment in which the coating remains intact but a crevice is formed between it and metallic surface Disbonded coating prevents the CP current from reaching deep inside the crevice and allows the accumulation on steel of a dilute bicarbonate solution with CO(2) content from ground moisture, having pH 6-7 Thus, critical conditions are achieved [1]. NN-SCC phenomena are usually reproduced in laboratory experimental study by SSR tensile tests on cylindrical specimens. Cracking is found after the onset of necking, at high plastic strain, i.e. under conditions that are not representative of real loading. Another limitation is related with the corroding surface, as NN-SCC only initiates after several years of service, from localized attacks. In order to reproduce the nucleation sites that promote NN-SCC on real structures an electrochemical technique for pre-corrosion of has been Improved [2]. This work deals with the initiation and growth of crack for NN-SCC. Initiation conditions were investigated by means of 3 point slow bending tests on full thickness specimens and, for comparison purpose, slow strain rate tests on cylindrical specimens the NN-SCC propagation was investigated through low frequency corrosion-fatigue tests on single notch three point bend beam specimens. Tests were carried out on a API 5L X65 controlled rolled ferritic-pearlitic steel for pipelines (Table I) in NS4 solution (0 483 g/L. NaHCO(3), 0 122 g/L KCl, 0 18 g/L. CaCl(2) and 0 1 g/L MgSO(4)). This solution was proposed by Parkins [4] and represents the average composition of the water found on Canadian pipelines, where NN-SCC phenomena were observed NS4 solution show 8 3 pH that reduces to 7 1 saturation with CO(2)/N(2) gas mixture at 0.05 atm CO(2) partial pressure The electrochemical pre-corrosion procedure involves 10 cycles of cyclic voltammetry from -1.8 to +1.8 V vs SCE (scan rate 20 mV/s) in NS4 solution modified by increasing the CO(2) and bicarbonate content (1 atm and 12.4 g/L, respectively) to obtain neutral pH. Near Neutral Stress Corrosion Cracking (NN-SCC) is a particular form of stress corrosion cracking that takes place on coated buried pipelines It occurs under particular circumstances, in presence of slow plastic straining, in areas of disbondment in which the coating remains intact but a crevice is formed between it and metallic surface Disbonded coating prevents the CP current from reaching deep inside the crevice and allows the accumulation on steel of a dilute bicarbonate solution with CO(2) content from ground moisture, having pH 6-7 Thus, critical conditions are achieved [1]. NN-SCC phenomena are usually reproduced in laboratory experimental study by SSR tensile tests on cylindrical specimens. Cracking is found after the onset of necking, at high plastic strain, i.e. under conditions that are not representative of real loading. Another limitation is related with the corroding surface, as NN-SCC only initiates after several years of service, from localized attacks. In order to reproduce the nucleation sites that promote NN-SCC on real structures an electrochemical technique for pre-corrosion of has been Improved [2]. This work deals with the initiation and growth of crack for NN-SCC. Initiation conditions were investigated by means of 3 point slow bending tests on full thickness specimens and, for comparison purpose, slow strain rate tests on cylindrical specimens the NN-SCC propagation was investigated through low frequency corrosion-fatigue tests on single notch three point bend beam specimens. Tests were carried out on a API 5L X65 controlled rolled ferritic-pearlitic steel for pipelines (Table I) in NS4 solution (0 483 g/L. NaHCO(3), 0 122 g/L KCl, 0 18 g/L. CaCl(2) and 0 1 g/L MgSO(4)). This solution was proposed by Parkins [4] and represents the average composition of the water found on Canadian pipelines, where NN-SCC phenomena were observed NS4 solution show 8 3 pH that reduces to 7 1 saturation with CO(2)/N(2) gas mixture at 0.05 atm CO(2) partial pressure The electrochemical pre-corrosion procedure involves 10 cycles of cyclic voltammetry from -1.8 to +1.8 V vs SCE (scan rate 20 mV/s) in NS4 solution modified by increasing the CO(2) and bicarbonate content (1 atm and 12.4 g/L, respectively) to obtain neutral pH. Both solutions were degassed with nitrogen for at least 24 hours, then were saturated with CO(2)/N(2) gas mixture for at least 2 hours, until a stable desired pH value was achieved Slow strain rate tests were carried out, according to ISO 7539-7, using cylindrical specimens with 3 mm diameter gauge " length, at 10(-6) S(-1) constant strain rate Loading curves were drawn. After the tests, reduction of area was measured on specimens and fracture surfaces were examined Three-point slow bending tests (SB) were performed with the device shown in Figure 1, under displacement control on full thickness specimens, obtained in transversal direction. The strain field during the bending test has been studied by computer simulation with the software DEFORM-2D (R), developed for the 2D finite element modeling of manufacturing processes. Plane strain was considered [5] An elastoplastic behavior of the material has been considered, defined by tensile tests Details of the numerical simulations are reported elsewhere [5,6] Through computer simulations it was possible to obtain total strain and strain rate of the Most strained fibers during bending A displacement rate of 0.02 mm/min corresponding to a strain rate about 10(-6) s(-1) for the most strained fibers, was adopted Besides monotonic bending tests, Ripple Loading (RL) test were performed They were carried out similarly to SB tests, but bending was stopped after plastic strains was reached at 105% of yield point in the loading curve (maximum load; after the in linear elastic strain Afterwards, loading cycles with 0.9 minimum to maximum load ratio (R=P(mm)/P(max)) were applied at frequency of 10(-2) Hz, for about 14 days After testing, the specimens were observed for detecting nucleation of stress corrosion cracks Fatigue and corrosion-fatigue tests were carried out on Single-Edged-Notched-Three Points Bending (SENB3) specimens at 0 6 R, with sinusoidal cyclic loading. Delta K decreasing tests and constant P test were carried out. The frequency was 18 Hi. in air and 0.2 Hz in NN-SCC solution at free corrosion potential In order to evaluate the contribution of NN-SCC to corrosion fatigue crack growth rate, the Wet and Landes superposition model was considered [3]. It assumes corrosion-fatigue crack growth rate (da/dN)(CF) is sum of the fatigue contribution (da/dN)(F), obtained by Paris law, and stress corrosion contribution (da/dN)(C) Thus, the corrosion contribution can be evaluated by Eq.1 The results show loading curves of SSR tests in NN-SCC solution very similar to those obtained in air (Figure 2) but slightly lower reduction of area was observed due to small brittle crack propagations in necking region (Figure 3). In all cases, high reduction of area was approached. The SB test allows studying the material behavior at high uniform plastic deformations. Figure 4 shows the SB curves in air and in NN-SCC solution There was no change in the curves due crack initiation and propagation However, SEM analysis showed several small cracks on pre-corroded specimens whereas the specimens that have not undergone any pre-corrosion show only rare points of initiation (Figure 5) All cracks were perpendicularly to tensile loading, and occurred on external fibers, from localized attacks, showing a morphology similar to that observed in field, after very long time of exposure Thus, the data confirm the ability of the technique of pre-corrosion to promote suitable surfaces for laboratory studies on NN-SCC Owing to the cracking start from shallow pits, with very small depth, a mayor effect of local variations of solution composition connected with occluded cell can be assumed Inside pit, hydrogen ion concentration significantly inreases by the hydrolysis of Fe(2+) ions Consequently, the formation of atomic hydrogen significantly increases too and NN-SCC may occur. On the contrary, the amount of hydrogen is too small for embrittlement at free corrosion potential, outside occluded cell. The hydrogen embrittlement mechanism in NN-SCC was individuated by Parkins, and confirmed by other works [7,8] However, the mechanism requires continuous plastic deformations In SSR tests and SB tests, plastic deformation exceeding 20% were reached, but cracking were observed at much lower values in RL tests Although no significant changes in the loading curves were found, SEM examinations of the pre-corroded specimens after RL tests showed initiation of a single small crack, within a localized attack (Fig 6) It has been established that crack initiation is related with slow and continuous plastic straining, never being observed under static stress [9,10]. Thus, the cyclical variation of the load during the RL tests represents a promoting factor of Occurrence of the phenomenon In the pipeline, pressure variations may promote both initiation and propagation of micro-cracks, producing local stress concentration and plastic straining at crack tip The significantly lower number of crack embryos on RL specimens with respect to SSR and SB specimens seems to indicate that nucleation of NN-SCC cracks is related to plastic strain, going up with it [11] Figure 7 shows the results of corrosion fatigue crack growth rate tests in NN-SCC solutions compared to fatigue behavior in air. These tests were adopted in order to evaluate crack growth assisted by environment under continuous straining at the crack up The propagation curves show typical behavior of stress corrosion fatigue Crack growth rate increases with respect to fatigue in air at stress intensity factor variation (Delta K) between 1617 MPa root m and 24 MPa root m, as maximum stress intensity factor exceeds 40 MPa root m For K above 25 MPa root m curves obtained in air and in NN-SCC solution (at low frequency) approach together, because mechanical effect becomes prevalent on corrosion contribution The effect is evident by analyzing the data according to the Wet and Landes model (Figure 8): the contribution of corrosion to CF crack growth rate increases from Delta K threshold, reaching a plateau typical for stress corrosion phenomena. At high Delta K values, the increase of in contribution hides the corrosion contribution NN-SCC crack growth rate can be evaluated by Eq 2 by assuming that crack propagation for corrosion can only occur during increasing phase of loading cycle, with tensile stress at crack tip Plateau crack growth rate for stress corrosion cracking can be estimated at 3.2-10(5) mm/s Significant crack propagation occurs due to synergic action of corrosion and fatigue, differently from RL. tests It should be emphasized that the ratio between maximum load and in load for CF. tests is lower than RI, tests Moreover, stress concentration at crack tip is much higher with respect to the pit bottom. Finally, in the CF tests very high plastic straining occurs at crack tip at every cycles On the base of these results it can be point out that crack initiation seems to be mainly affected by the presence of localized attacks and the amount of plastic deformation achieved during continuous straining, while load variation amplitude and plastic strain rate are more important for the propagation from crack embryos.

Scopo del lavoro è stato la valutazione delle condizioni di innesco della NN-SCC di un acciaio per tubazioni interrate API 5L X65, adottando provini di flessione in tre punti prelevati in pieno spessore in direzione trasversale rispetto alla tubazione. Sono state effettuate prove di flessione lenta, monotone o monotone fino al raggiungimento di condizioni di deformazione plastica, con successiva variazione ciclica del carico a bassa frequenza. Allo scopo di riprodurre una superficie corrosa con morfologia simile a quella osservata in campo è stata adottata una particolare procedura di precorrosione elettrochimica, sviluppata in precedenti lavori. I risultati ottenuti hanno permesso di osservare l’innesco di numerose microcricche sui provini con superficie precorrosa, mentre tali microcricche sono state osservate sui provini con superficie tal quale in quantità significativamente inferiore. La morfologia delle microcricche innescate dagli attacchi localizzati è risultata molto simile ai casi di NN-SCC osservati in esercizio.