The effect of the stress ratio on fatigue crack growth in a 3% NaCl solution

Corrosion fatigue crack growth tests have been carried out at various stress ratios for a low alloy steel SNCM 2 and type 304 stainless steel.

Measurements of the effective stress intensity factor range ratio U were performed to explain the effect of stress ratio R.

The corrosive environment decreased da/dN at R = 0.1, 0.4 and little affected da/dN at R = 0.9 for SNCM 2 and increased da/dN at all R ratios for SUS 304.

It was confirmed that there exists a threshold stress intensity factor ΔK_thCF in 3% NaCl solution for both materials tested.

The corrosive environment decreased ΔK_thCF for all conditions tested except at R = 0.1 and 0.4 for SNCM 2, where ΔK_thCF-values were nearly equal to ΔK_th-values in air. ΔK_thCF/ΔK_th was 0.6 at R = 0.9 for SNCM 2 and 0.8, 0.5 and 0.7 at R = 0.1, 0.7 and 0.9 for SUS 304, respectively.

It was shown that the complicated effect of stress ratios on crack growth for SNCM 2 can be explained using effective stress intensity factor ΔK_eff.     

Threshold range and opening stress intensity factor in fatigue

The fatigue threshold, ΔK_th, is strongly influenced by the stress-ratio, ie by the loading conditions. Results for a Ti6A14V alloy show that a ΔK exists for non-propagating fatigue cracks which is independent of loading conditions. This ΔK is called the fatigue tolerance range and is denoted by ΔK^K. The fatigue tolerance range corresponds to that part of the ΔK_th during which the fatigue crack is open. Arguments that the fatigue tolerance range has to be explicitly incorporated in equations predicting fatigue crack growth rates are presented.     

Relation between fatigue threshold and fatigue limit for surface-rolled specimens

After cylinder notch fatigue specimens of 40 CrNiMo steel were rolled, their fatigue limit increased by 41%. The rolled specimens did not fracture, even though they had been loaded for 10⁷ cycles under fatigue limit stress, but a non-propagating fatigue crack was generated. Thus the value of the fatigue limit depends on the fatigue threshold value ΔK_th of the metal of the rolled layer. Plastic deformation increased ΔK_th in these experiments. It can be inferred that ΔK_th of the rolled layer increases from the occurrence of plastic deformation and microvoids on the layer. Calculation of the effect of residual stress in the crack wake on the stress intensity factor ΔK indicates that residual compression stress decreases ΔK by 21.5 MPa √M. It was calculated that rolling induced both the length of the non-propagating crack and the increase of fatigue limit. The calculated values are in accord with experiment. Analysis and calculations indicate that the non-propagating crack is generated on the rolled layer. Thus the fatigue limit is improved because rolling produces residual compression stress in the layer (which decreases the stress intensity factor), and increases ΔK_th of the layer.     

The effect of load ratio on the threshold stresses for fatigue crack growth in medium carbon steels

A study of the effect of load ratio, R, on the low crack growth rates and the threshold stress conditions exhibited by five medium carbon steels has been conducted. It was found that decreased values of R retarded crack growth to an increasing degree as a defined threshold for crack growth was approached, such that this threshold showed a marked dependence on K max rather than ΔK. Comparisons of the five materials showed that the threshold increased as yield strength increased for a given R, but this effect could be normalised in terms of a constant value of the maximum plastic zone size at the crack tip.     

Fatigue threshold and crack growth in an electron beam weld made of steel and bronze

In this research topic some experimental tests with single-edge notched beams were performed to determine the threshold load value for fatigue crack growth and to characterize fatigue crack behaviour of an electron beam weld made of steel and bronze. Subsequently, numerical analyses were done to estimate the threshold value ΔK_th and to simulate the fatigue crack growth. The calculated crack path was compared to those determined experimentally. The objective was to find out the necessary fracture properties for an analysis of an electron beam welded worm wheel and to asses the capability of usual fracture analysis software to simulate fatigue crack growth in welds.     

Influence of R on effective stress range ratio and crack growth

The influence of stress ratio R and stress intensity range ΔK on crack closure and fatigue crack growth were studied. Crack closure and crack growth experiments were performed on 6063-T6 Al alloy. Crack closure stresses were measured using a surface-measurement technique with a COD gauge. The gauge was placed at different locations behind the crack tip, and it was found that the location of the gauge does not influence the closure load. The closure load was however found to be a function of R and ΔK. Fatigue crack growth rate is found to depend upon R, U and ΔK. A model for both U and da/dN has been developed.     

A microcomputer control system for the measurement of fatigue crack growth data

A microcomputer-based system for the measurement of fatigue crack growth da/dn versus cyclic stress intensity factor ΔK data using compact-tension test specimens is described. The procedure has been developed to allow automatic measurement of crack growth rate under any specified combination and sequence of load conditions, i.e. ΔK and R (stress ratio) and includes the capability of establishing the threshold cyclic stress intensity factor ΔK₀. Crack extension measurement is effected from the elastic compliance evaluated from the AC component of the load and displacement signals to an accuracy of -3 μm every 1000 load cycles. Results from a typical low-alloy-steel rotor forging are presented to illustrate the use of the system.     

Understanding fatigue crack propagation in AISI 316 (N) weld using Elber’s crack closure concept: Experimental results from GCMOD and acoustic emission techniques

An experimental study of fatigue crack propagation and crack closure behaviour, in compact tension specimens of AISI 316 (N) weld has been conducted. The crack closure load was determined from the changes in the slope of the load–displacement curves using global crack mouth opening displacement (GCMOD) type gauge. The results were compared with those measured by acoustic emission technique which showed good agreement with each other. The experimental data bear clear evidence of fatigue crack closure. The crack opening force was found to increase moderately with crack length and increasing R-ratio, under a constant P_max of 5 kN. Above a critical R-ratio of 0.45 (approximately), the crack closure load is smaller than the minimum applied load. A good correlation was obtained for ΔK_eff/ΔK = 0.6684 – 2.4135R + 7.0077R² in the range 0 R  0.5. The magnitude of crack closure is used to interpret observed crack growth behaviour at different R-ratios.     

Effects of corrosive environments on near-threshold fatigue crack growth behavior of T1-6A1-4V

The near-threshold fatigue crack growth behavior of Ti-6A1-4V alloy has been investigated in low O₂ steam (< 1 ppm), high O₂ steam (40ppm), and boiling water with various concentrations of Nad and/or Na₂₂SO₄. At load ratio (R) of 0.5, high O₂ steam increased the crack propagation rates in the threshold region, relative to low O₂ steam. However, at R = 0.8, the near-threshold crack growth rates in low and high O₂ steam were comparable. Values of threshold stress intensity range, ΔK_th, slightly increased with an increase in the concentration of NaCl in the solution. Varying solution pH from 5.0 to 10.0 in a 0.1 g NaCl plus 0.1 g Na₂SO₄ per 100ml H₂O solution had no effect on the rates of near-threshold crack propagation. Increasing the hydrazine level from 30 to 10⁷ ppb in the same salt solution also did not change the resistance to crack growth. Comparing the present results with the previous data on 403 stainless steel, the near-threshold crack propagation rate performance in Ti-6Al-4V alloy is superior to that in 403 steel in both the steam and salt solution environments.     

Fatigue crack initiation and growth under mixed mode loading in aluminum alloys 2017-T3 and 7075-T6

Fatigue crack initiation and growth characteristics under mixed mode loading have been investigated on aluminum alloys 2017-T3 and 7075-T6, using a newly developed apparatus for mixed mode loading tests. In 2017-T3, the fatigue crack initiation and growth characteristics from a precrack under mixed mode loading are divided into three regions—shear mode growth, tensile mode growth and no growth—on the ΔK_I-ΔK_II plane. The shear mode growth is observed in the region expressed approximately by ΔK_II > 3MPa√m and ΔK_II/ΔK_I > 1.6. In 7075-T6, the condition of shear mode crack initiation is expressed by ΔK_II > 8 MPa√m and ΔK_II/ΔK_I > 1.6, and continuous crack growth in shear mode is observed only in the case of ΔK_I/ΔK_II, 0. The threshold condition of fatigue crack growth in tensile mode is described by the maximum tensile stress criterion, which is given by Δσ_θmax √2πr 1.6MPa√m, in both aluminum alloys. The direction of shear mode crack growth approaches the plane in which K_I decreases and K_II increases towards the maximum with crack growth. da/dN-ΔK_II relations of the curved cracks growing in shear mode under mixed mode loading agree well with the da/dN-ΔK_II relation of a straight crack under pure mode II loading.     

Fatigue crack growth characteristics of rotor steels

Room temperature fatigue crack growth rate data were generated for Ni-Mo-V (ASTM A469, Cl-4), Cr-Mo-V (ASTM A470, Cl-8) and Ni-Cr-Mo-V (ASTM A471, Cl-4 and a 156,000 psi yield strength grade) rotor forging steels. Testing was conducted with WOL type compact toughness specimens and the results presented in terms of fracture mechanics parameters. Data show that the Ni-Cr-Mo-V steels exhibit slower fatigue crack growth rates at a given stress intensity range (ΔK) than do the Ni-Mo-V steels. In addition, the Cr-Mo-V steel was found to exhibit slower growth rates than the other alloys at ΔK levels below 40 ksi √in but somewhat foster rates at ΔK levels in excess of 45 ksi √in. The fatigue crack growth rate properties of the alloys studied conform to the generalized fracture mechanics crack growth rate law where da/dN = C₀ΔK^R. It was noted that the fatigue crack growth rate parameters n and C₀ tend to decrease and increase, respectively, with increasing material toughness, K_ic.     

Automatic near-threshold fatigue crack growth rate measurements at liquid helium temperature

The development of a fully automated test apparatus for near-threshold fatigue crack growth rate measurements in a liquid helium environment is described, and some initial results for AISI 300 series stainless steels are presented. The experimental apparatus consists of a servohydraulic test machine and a cryostat, complete with a minicomputer, a programmable arbituary waveform generator, a programmable digital oscilloscope and a fully automatic liquid helium refill system. The technique uses 6.4 mm thick compact specimens subjected to systematically decreasing loads, with 24 h operation at 40 Hz, the crack growth being continuously monitored by specimen compliance measurements. The results presented in this study include da/dN vs ΔK curves and threshold fatigue stress intensity factors, ΔK_th, at 4 K for AISI 304L, 304LN and 316 stainless steels. The near-threshold fatigue behaviours of these materials are similar, and the fatigue crack growth rate trends at intermediate ΔK levels nearly agree with published results.     

A study of the effects of mechanical and environmental variables on fatigue crack closure

Using the potential drop technique, fatigue crack closure has been monitored in pin loaded SEN specimens of -titanium, a titanium alloy and EN24 steel. The specimens were tested in tension-tension under conditions closely approximating to plane strain, and closure was only detected in vacua of better than 133 mN m⁻² (10⁻³ torr). No significant closure was detected in air. The extent of the crack area closed at minimum load varied with air pressure, applied stress, R ratio (R = L_min/L_max), crack length, material, and loading mode. Additional experiments made with a dip gauge showed that the COD/applied load response of the crack was non-linear in vacuum above minimum load in the fatigue cycle confirming that crack closure was occcurring. It is shown that for a given material, loading mode and air pressure, the effect of loading and crack length variables on crack area closed at minimum load can be characterised in terms of the parameter (K²_min-ΔK², this being proportional to the calculated COD at minimum load. The extent of closure in vacuum is influenced extensively by this parameter.     

Crack growth in ultrafine-grained AA6063 produced by equal-channel angular pressing

Crack growth behaviour of ultrafine-grained AA6063, processed by equal-channel angular pressing (ECAP) via route E at room temperature, was evaluated with special emphasis on the effect of grain size distribution and work hardening. A bimodal, two times ECAPed condition and a monomodal ultrafine-grained condition after eight ECAP passes are compared with the coarse grained peak aged material. Depending on their microstructure, the ECAPed materials show significantly lower fatigue threshold values (ΔK _th) and higher crack growth rates (da/dN) than their coarse grained counterparts. Micrographs of the crack propagation surfaces reveal the reduced grain size as major key to increased crack growth rates of the ECAPed material, as it influences roughness-induced crack closure and crack deflections. Furthermore, the effects of other features, such as ductility, work hardening capability and grain boundary characteristics, are discussed.     

Effect of load ratio on the deformation behaviour of fatigue crack tips

To investigate the change in the manner of plastic deformation at a fatigue crack tip under different loading conditions, the fresh surface area which is produced in one fatigue cycle at the crack tip was measured by an electrochemical method under various load ratios. The fresh surface area produced in each fatigue cycle is in proportion to the crack propagation rate in the Paris regime. This indicates that the crack tip geometry at K_max maintains a similar profile through the Paris regime. Furthermore, the fresh surface area is independent of load ratio at a given da/dN in the Paris regime. This suggests that the shape of the fatigue crack tip at K_max should be independent of load ratio at a given da/dN, and that the shape at K_min is also independent of load ratio, whereas the size of the deformed region at the crack tip is in proportion to da/dN.     

Modelling and measurement of crack closure and crack growth following overloads and underloads

Ignoring crack growth retardation following overloads can result in overly conservative life predictions in structures subjected to variable amplitude fatigue loading. Crack closure is believed to contribute to the crack growth retardation, although the specific closure mechanism is debatable. The delay period and corresponding crack growth rate transients following overload and overload/underload cycles were systematically measured as a function of load ratio (R) and overload magnitude. These responses are correlated in terms of the local “driving force” for crack growth, i.e. the effective stress intensity factor range (ΔK_eff). Experimental results are compared with the predictions of a Dugdale-type crack closure model and improvements in the model are suggested.     

The growth of fatigue cracks through particulate sic reinforced aluminum alloys

Crack growth rates for large fatigue cracks in 12 variations of particulate silicon carbide reinforced aluminum alloy composites have been measured. Composites with seven different matrix alloys were tested, four of which were of precipitation hardening compositions, and those were tested in both as-extruded and peak aged conditions. Five of the materials were made by casting, ingot metallurgical methods and two of the alloys by mechanical alloying, powder metallurgical methods. For both manufacturing methods, primary fabrication was followed by hot extrusion. The fatigue crack growth curves exhibited an approximately linear, or Paris law, region, fitting the function da/dN = BΔK^s, and a threshold stress intensity factor, ΔK_th. As has been found for other materials, the coefficients B and s are correlated; for these composites In B= −16.4−2.1s. A correlation was also found between ΔK_th and s, and it was found possible to compute the magnitude of ΔK_th using a simple model for the threshold together with yield stress and SiC size and volume fraction. These results were explained using a relationship between ΔK_th and crack closure determined previously for unreinforced aluminum alloys. The path of fatigue crack growth is through the matrix for these composites, and SiC has the effect of altering the slip distance, therefore, the plasticity accompanying fatigue cracks. It was shown that all the crack growth rate curves were reduced to one equation having the form da/dN = B'ΔK_eff^s' where B' = 6.5 × 10^-9m/cy and s' = 1.7. A partly theoretical method for predicting fatigue crack growth rates for untested composites is given. Fatigue crack surface roughness was measured and found to be described by a fractal dimension, but no correlation could be obtained between surface roughness parameters and ΔK_th.     

Flexural fatigue characteristics of sandwich structures at different loading frequencies

The effects of frequency on the fatigue behavior of S2 glass fiber–vinylester reinforced sandwich composites with two different PVC cores have been investigated. Flexural fatigue tests were performed on sandwich beams with core densities of 130 and 260 kg/m³ at frequencies of 3 and 15 Hz, at a stress ratio, R=0.1 and at four different load levels viz. 90%, 85%, 80% and 75% of the ultimate load. S–N diagrams were generated and it was observed that the fatigue strength increased with core density, and the number of cycles to failure, N_f, increased with increase in frequency. In all cases failure was dominated by a primary shear crack in the core however, the crack path and crack propagation rates varied with frequency. The fatigue crack growth rate (FCGR) in the core of the H130 sandwich beams was subsequently investigated and the relationship between the crack growth rate, da/dN, and the cyclic stress intensity range, ΔK, was determined. It was found that crack growth rate decreased with increase in loading frequency.     

Endurance diagram

For the design of components that do not contain any defects and that are subjected to both alternating and mean stresses, a Soderberg diagram is normally used in a ‘safe life’ approach. However, structures may contain defects or cracks where a Soderberg diagram will not be sufficient. In this paper the development of an endurance diagram is presented for components that may contain crack-like defects left in the structure during the fabrication process. This diagram is based on the threshold stress intensity factor in fatigue crack growth, ΔK_th, and the critical stress intensity factor, K_c.     

Matrix fatigue cracking in fiber composites

A model is developed for fatigue growth of matrix cracks in metals reinforced with aligned continuous elastic fibers. The mechanics of elastic cracks bridged by frictionally constrained fibers is used to develop the model, which provides estimates of the tip value of the stress intensity factor amplitude, ΔK_TIP. It is found that when the applied load amplitude is held fixed during fatigue crack growth, ΔK_TIP, and thus the rate of growth approach an asymptotic value independent of crack length. The residual strength after fatigue crack growth is also discussed. In some cases, the residual strength is unaffected by prior fatigue growth. But, in another regime, the matrix crack length allows fibers to begin breaking before the matrix crack grows. The strength is then inversely proportional to the square root of fatigue crack length.