PSI - Issue 13

5

Ya. Khaburskyi et al. / Procedia Structural Integrity 13 (2018) 1651–1656 Author name / Structural Integrity Procedia 00 (2018) 000–000

1655

2

3

F

4

5

1



Fig. 4. Diagrams F –  for an evaluation of fatigue crack closure effect.

Table 1. Results of the fatigue crack growth tests.

Marks on fig. 4

Environment

Δ K , MPa m 1/2

op , MPa m

Δ K

eff , MPa m

da / dN, m/cycle

1/2

1/2

K

U

1 2 3 4 5

Air

15.5

2.2 3.4

0.86 0.81 0.43 0.25 0.11

13.3 13.6

1.8 10 -8

17 17 17 17

0 0 0

Special technological environment (aqueous solution)

10.2 12.7 15.1

6.8 4.3 1.9

0 To confirm definitely the chemical interaction of the proposed substance with crack surfaces metal, similar experiments were carried out on plexiglass, chemically inert material. Cyclic loading was interrupted at crack growth rate da / dN ~ 10 -7 m/cycle and the diagram F –  was registered, which did not show crack closure effect. Then the aqueous technological environment in the concentration of 20 g/l was supplied into the crack and the test was continued. However an effect on crack growth rate and its closure was not revealed. It means that the technological environment does not affect fatigue crack growth on chemically inert material under the same conditions of cyclic loading as for the steel. One can assume a certain role of water as a dissolvent for the proposed chemical substance in the processes which lead to crack growth retardation and arrest. Therefore similar experiments were carried out on the steel 20 using the substance solution in ethyl alcohol in concentration of 20 g/l. The experimental results are presented in Fig. 2 b . Total crack arrest was observed as in the case when testing in the aqueous solution under the same concentration. It indicates an interaction of crack surfaces metal with the proposed chemical substance, regardless of the solvent, either water or alcohol. To reveal the Δ K range where the crack arrest takes place, the specimen after total crack growth stop was periodically load up (increase Δ K ) and hold a certain number of cycles N , to be sure in possible crack growth start. Such start was observed at Δ K ~ 40 MPa m 1/2 , but the positive effect of treatment was removed at Δ K ~ 50 MPa m 1/2 only. Evidently, this SIF is close to the cyclic fracture toughness K fc . Thus, at a certain concentration of the substance in water or alcohol it is possible to attain crack arrest in the entire actual range of SIF. The peculiarities of the proposed method usage for corrosion fatigue crack growth retardation. At a low fatigue crack rate ( da / dN ~ 10 -8 m/cycle) crack closure effect is insignificant, as depicted in Fig. 5 a . A creation of artificial crack closure implies a decrease of effective range Δ K eff due to an increase of minimal effective SIF K min eff , and average effective SIF K av eff increases respectively (Fig. 5 b ). The proposed method can be so effective to straiten Δ K eff that virtually approximate K mid eff to maximum SIF K max (fig. 5 b ). This actually means almost static loading in the crack tip during cyclic loading, namely, with the nominal stress ratio R = 0. Then if material is sensitive to stress corrosion cracking or hydrogen embrittlement and K max is above thresholds of corrosion K scc or hydrogen K HE cracking, this method could provoke corrosion and/or hydrogen induced cracking. It should be noted that low strength steels with high plasticity are usually not susceptible to stress corrosion cracking or hydrogen embrittlement. However their long-term service, especially under action of hydrogenating environments, can lead to sharp decrease of a material resistance to brittle fracture, as shown by Nykyforchyn et al.