Neural network (NN) models have made a significant impact on fatigue-related engineering communities and are expected to increase rapidly soon due to the recent advancements in machine learning and artificial intelligence. A comprehensive review of fatigue modeling methods using NNs is lacking and will help to recognize past achievements and suggest future research directions. Thus, this paper presents a survey of 251 publications between 1990 and July 2021. The NN modeling in fatigue is classified into five applications: fatigue life prediction, fatigue crack, fatigue damage diagnosis, fatigue strength, and fatigue load. A wide range of NN architectures are employed in the literature and are summarized in this review. An overview of important considerations and current limitations for the application of NNs in fatigue is provided. Statistical analysis for the past and the current trend is provided with representative examples. Existing gaps and future research directions are also presented based on the reviewed articles.

In this study, a combined low and high cycle fatigue (CCF) life prediction model, which considers the crack closure effect (CCE) of micro-defects, is proposed based on the continuous damage mechanics. The CCF life prediction model is decomposed into three sub-models: the low cycle fatigue (LCF), high cycle fatigue (HCF) under the maximum stress of LCF (HCFLM), and their coupled damage models. The CCE is considered by taking one CCE parameter into the HCFLM sub-model. The experimental CCF data of K403 full-scale turbine blades under different vibration stresses is used to verify the accuracy of the proposed model to compare with other life prediction models. The prediction life from the proposed model falls within the 2 times of scatter band compared with the experimental results. Further, there are the different damage evolution forms at different vibration stresses. When the vibration stress is below 64.48MPa, the CCF damage mainly is caused by the LCF damage. However, while the vibration stress is higher than 64.48MPa, the HCFLM damage plays a major role in the CCF damage accumulation, and it is predicted that the CCF damage of the first stage serration on the K403 turbine blades is mainly from LCF.

Even though friction stir welding (FSW) has been shown to produce high performing butt-joints, stress concentration at the weld edges in overlap FSW significantly reduces the performance of these joints. By combining FSW and adhesive bonding into a friction stir (FS) weld-bonding, joint mechanical performance is greatly improved. Quasi-static and fatigue strength of the proposed FS weld-bonding joints was assessed and benchmarked against overlap FSW and adhesive bonding. The characterization of the structural adhesive is also presented, including differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA), as well as mechanical characterization with curing temperature. A small process parameter study was made to select proper FSW parameters for AA6082-T6 overlap FSW and FS weld-bonded joints. The adhesive degradation temperature (357ºC) was found to be higher than reported temperatures in the adhesive during welding of FS weld-bonding joints. Higher curing temperatures were found to lead to increased strength while decreasing ductility of the adhesive. The addition of adhesive bonding to the overlap FSW to produce FS weld-bonding resulted in a significant increase in quasi-static and fatigue strength, achieving 79.9% of the fatigue strength of adhesive bonded joints at 106 cycles, while FSW had 41.6%.

This paper presents a study on the effect of microstructure on the fatigue crack growth rate (FCGR) in advanced normalised-rolled (NR) and thermomechanical control process (TMCP) S355 steels in the Paris Region of the da/dN vs. ΔK log-log plot. The environments of study were air and seawater (SW), under constant amplitude sinewave fatigue loading. Discussions were based mainly on the comparison between the crack path in the TMCP and NR steels. Fundamentally, three phenomena: crack-tip diversion, crack-front bifurcation and metal crumb formation were observed to influence the rate of fatigue crack growth (FCG). The prevalence of these phenomena appears to be a function of the nature of the material microstructure, environment and crack-tip loading conditions. The three factors appear to retard the crack growth by reducing or re-distributing the effective driving force at the main active crack tip. A crack path containing extensively the three phenomena was observed to offer strong resistance to FCG. Increase in the FCGR was observed with decrease in the crack-tip diversion angle, branched-crack length and metal crumbs formed. In SW, the degree of the electrochemical dissolution of the microplastic zone (or crack-tip blunting) appears to be an additional factor influencing crack growth in ferrite-pearlite (α-P) steel. This study, generally tends to present microstructural features that strongly influenced FCGR in α-P steels in the Paris Region both in air and SW. This work is very important in the design of fatigue resistant steel.

This article describes how a step-stress accelerated life test (SSALT) can be designed for testing the fatigue life and reliability of structural components with a single failure mode. With simple numerical simulations of the crack’s propagation in the notched area of the structural part for different loading levels, the slope of the S-N curve for a structural component is initially estimated. Then, a very few fatigue-life experiments are carried out in the high-cycle domain to determine the intercept of the structure’s S-N curve. By considering the scatter from the material’s P-S-N curve, different SSALT designs for the structural component can be composed and checked for their expected acceleration factor. The procedure is experimentally validated for the case of a notched specimen and two different SSALT designs. From the results it can be concluded that the predicted durations of the SSALT experiments correlate well with the real experiments.

The main aim of the current study is to evaluate the compressive quasi-static and fatigue properties of titanium alloy (Ti6Al4V) cellular materials, with different topologies, manufactured via Laser Powder Bed Fusion (LPBF) process. The topologies herein considered are lattice based regular and irregular configurations of cubic, star and cross shaped unit cell along with trabecular based topology. The results have indicated that the effective stiffness of all configurations are in the range of 0.3 – 20 GPa, which is desirable for implant applications. The morphological irregularities in the structures induce bending dominated behavior affecting more the topologies with vertical struts. The S – N curves normalized with respect to the yield stress indicate that the behavior of star regular structures is between purely stretching dominated cubic and purely bending dominated cross based structures. Trabecular structures have shown desirable quasi-static and fatigue properties despite the random distribution of struts.

Unloading compliance (UC) method and normalization method (NM) are two of the most commonly used methods for determining the fracture toughness of materials. However, considerable differences often exist in the fracture toughness determined by these two methods, which solicits a new method to determine the fracture toughness accurately. In this paper, the compliance of crack length differences as measured by the crack length difference ratio Si is discovered, analysed and verified by experiments. Based on this compliance, a new accurate method, known as AJR, is developed and verified by test results. Factors that exhibit the advantages of the developed new AJR method are also investigated. It is found that the J-R curves determined by the new AJR method are more accurate than those determined by UC and NM. The new AJR method should be the first choice for steels with a small strain hardening ratio and low effective yield strength, and thicker CT specimens with shallower initial crack length. This is because the disagreement between UC and NM is unacceptably large. The developed new AJR method and the results presented in this paper can assist engineers and researchers to determine J-R curves and fracture toughness of steels more accurately and can contribute to the body of knowledge of fracture mechanics.

The fatigue behaviour of notched and unnotched specimens produced by additively manufactured Inconel 718 are analysed in the as-built and heat-treated conditions. The surfaces display high roughness and defects acting as fatigue initiation sites. In the as-built condition, fine sub-grains were found, while in the heat-treated state, the sub-grains were removed and the dislocation density recovered. SN-curves are predicted based on tensile properties, hardness and defects obtained by fractography, using the √area-method.

The use of passive infrared thermography comprises great opportunities to improve understanding the fatigue damage process of crack-patched structures. Quasi-static and cyclic coupon tests are performed using metallic specimens with single-sided bonded patches and monitored with passive infrared thermography. Different test setups help to differentiate between metallic crack growth and adhesive damage on thermal images. Results show that metallic crack growth can be monitored from the patched side, also in combination with local delamination at the patch/metal interface. Thus, it is possible to analyse the overall degradation progress of the crack patched component under loading conditions and thereby to identify the driving damage mechanism of the particular repair configuration. Being able to understand the overall damage behaviour of crack patched components is essential to improve the ability of predicting its long-term behaviour.

The operational conditions in next generation power plants and other high temperature applications such as turbine blades in jet engines demand the component to perform under extreme conditions where metallic materials show time dependent deformation under cyclic loading conditions. Under creep-fatigue loading condition, the crack tip is exposed to both time dependent and independent plastic deformation. Conventional crack characterizing parameters such as (C t ) , avg has shown good correlation with dominant damage based crack velocity, (d a/d t ) avg . However, the true definition or prediction of crack driving forces under such scenario are vague due to limited theoretical validity of the conventional crack tip characterizing parameters, such as J or C t. In this work, the concept of configurational forces are applied for the first time to understand the creep-fatigue crack growth behaviour. The crack growth is simulated using node-release technique and the configurational forces are calculated using post processing the finite element results for calculation of dJ/dt.

Based on the physical phenomenon that the fatigue cracks initiate along specific slip plane, a slip plane damage based low cycle fatigue (LCF) lifetime model for the nickel- based single crystal superalloy is established. The predicted results indicate that the lifetime model can reflect the orientation effect. In addition, in order to characterize the dwell time dependence of the LCF lifetime, creep damage and compression-creep damage are introduced to the lifetime model. Finally, the lifetime predictions under LCF loading with tensile dwell time, compressive dwell time and LCF with tensile-compressive dwell time are conducted by employing the lifetime mode, respectively. The predicted lifetimes show a good agreement with the experimental data, which verifies the accuracy of the developed lifetime model in this paper.

Superaustenitic steel Sanicro 25 has been subjected to in-phase and out-of-phase thermomechanical fatigue cycles in the temperature range from 250°C to 700°C. Both constant strain rate cycling and cycling with 10 minutes dwell at peak temperature were applied. The effect of the dwells on the cyclic response, internal structure and damage mechanism was studied. Cyclic hardening/softening curves, cyclic stress-strain curves and fatigue life curves were evaluated. The transmission electron microscopy was used to find modifications of the internal structure and precipitation of the nanoparticles. 10 min dwell at maximum temperature modified substantially the dislocation arrangement. Various nanoparticles representing the obstacles for dislocation motion were analysed and identified by energy dispersive X-ray spectroscopy in scanning transmission electron microscope. The damage mechanism operating under specific loading conditions was investigated on the surface as well as in the interior of the cycled specimens. Scanning electron microscopy combined with focused ion beam and electron backscatter diffraction was adopted to reveal the respective mechanisms responsible for crack nucleation and propagation. Effect of dwells on fatigue behaviour, modification of internal structure and damage mechanisms are analysed and discussed.

Orthogonal experiment design together with the analysis of variance was used to examine the processing parameters (laser power, scan speed, layer thickness and hatch spacing) of selective laser melting (SLM) for superior properties of SLM parts, in which nine groups of specimens of Ti-6Al-4V were fabricated. The porosity for each group was measured and the results clarify that the influence sequence of individual parameter on the porosity is laser power > hatch spacing > layer thickness > scan speed. Ultrasonic fatigue tests (20 kHz) were conducted for the SLMed specimens in high-cycle fatigue (HCF) and very-high-cycle fatigue (VHCF) regimes. The S-N data show that the fatigue strength is greatly affected by the porosity: the group with the smallest porosity percentage having the highest fatigue strength in HCF and VHCF regimes. Moreover, the observations by scanning electron microscopy revealed that fatigue cracks initiate at lack-of-fusion defects in the cases of surface and internal crack initiation.

In the present investigation temperature dependence fatigue strength behaviour of Inconel 825 super alloys is investigated. Based on the experimental results different S-N models have been derived and suitable model for the prediction of fatigue strength have been proposed. An inverse power and exponential relation between fatigue strength and absolute temperature is demonstrated. The proposed models are used to predict the fatigue life using well known Palmgren-Miner rule. Based on high to low and low to high load steps test data sets under identical test conditions, Miner rule based statistical damage constant is stochastically modeled for fatigue life prediction at different level of probability and validated. The modeling process combines a probabilistic fatigue damage accumulation and a stress-life-temperature relation technique.

Gear contact fatigue is becoming a primary limitation for the growing demand of power density and service life in gear-driven equipment. The unchecked surface fatigue crack could further cause premature failure and put a serious risk to the safety and reliability of mechanical systems. In this work, an attempt is made to investigate the effects of rolling-sliding and mild wear on contact fatigue behavior. A comprehensive contact model is developed to capture the variation instantaneous pressure and stress field is calculated with the transient mixed EHL approach. Rolling-sliding contact is simulated with the time-varying roughness topography updated by Archard wear equation. The stress cycles are extracted and the relative contact fatigue life is obtained by using Zaretsky criterion. Results suggest that in rolling-sliding contact the contact fatigue life is obviously lower compared with pure rolling. The increases in the number and amplitude of stress micro-cycles is found to be the main contributors to the reduction of fatigue life. Mild wear tends to smooth the surface, subsequently mitigates the stress concentration and reduces stress cycles, then decrease the risk of surface contact fatigue.

To exploit the full material potential of short fibre reinforced PA6T/6I, specific component calculations including aniso- tropic material behaviour is necessary. For this, different failure criteria and fatigue models are used to describe the behaviour during a component service life. This paper deals with the determination and consideration of fibre orientations for failure criteria and fatigue calculations. Therefore, a novel method to determine fibre orientation (FO) distributions across injection moulded plates, is proposed. The developed method allows a forecast of FOs for different specimen extraction positions and angles on injection moulded plates by using only a few measured reference points. As a result, fatigue models can be calibrated with the strength values and the corresponding FO, calculated for fracture position. The performed tests show a non-negligible influence of failure positions, due to fibre orientation distributions along the specimens. So, the FO determination method delivers an improvement in strength values estimation.

A multilayer overlay coating system containing an intermediate intermetallic layer (designated 2IML) is an architecture expected to show good fatigue resistance. Experimental characterisation and modelling simulations were carried out to classify the different crack initiation mechanisms occurring during fatigue of this coating system and to reveal how changes in the layer architecture lead to fatigue improvement. Fatigue improvement is achieved by decreasing the IML-Top layer thickness due to the increased surface crack initiation resistance. However subsurface initiation mechanisms inhibit the improvement (dominated by surface initiation mechanism) achieved by locating the IML-Top layer closer to the top surface.

Staircase testing is a standard method for evaluating the fatigue strength of components. However, staircase testing assumes a normal distribution, while components can display bimodal behavior due to flaws in material, or issues during the manufacturing process. Three unique step loading data sets on different production crankshafts provide evidence that step loading reliably identifies material or manufacturing issues, that lower a component’s fatigue strength. Staircase testing has an 87% or greater chance of overestimating the component’s fatigue strength, which in turn overestimates the component’s expected reliability. For example, a component with a 99.9% reliability, based on staircase testing would only have a 74% reliability based on step loading. If a component contains an undetectable manufacturing defect, staircase testing has a 99% chance of overestimating the component’s fatigue strength. Step loading reliably improves the estimation of a component’s fatigue strength distribution while providing insights into a component’s defect tolerance.