Advances in data-driven models for fatigue life prediction of metallic materials

Validation of a high-cycle fatigue model via calculation/test comparisons at structural scale: Application to copper alloy sand-cast ship propellers

A modified damaged stress model for fatigue life prediction based on load interaction

Bayesian uncertainty quantification and propagation for validation of a microstructure sensitive model for prediction of fatigue crack initiation

Aiming to improve the predictive ability of Walker model for fatigue life prediction and taking the turbine disc alloy GH4133 as the application example, this paper investigates a new approach for probabilistic fatigue life prediction when considering parameter uncertainty inherent in the life prediction model. Firstly, experimental data are used to update the model parameters using Bayes’ theorem, so as to obtain the posterior probability distribution functions of two parameters of the Walker model, as well to achieve the probabilistic life prediction model for turbine disc. During the updating process, Markov Chain Monte Carlo (MCMC) technique is used to generate samples of the given distribution and estimating the parameters distinctly. After that, the turbine disc life is predicted using the probabilistic Walker model based on Monte Carlo simulation technique. The experimental results indicate that: (1) after using the small sample test data obtained from turbine disc, parameter uncertainty of the Walker model can be quantified and the corresponding probabilistic model for fatigue life prediction can be established using Bayes’ theorem; (2) there exists obvious dispersion of life data for turbine disc when predicting fatigue life in practical engineering application.

In order to obtain comprehensive assessment of the factors influencing fatigue life and to further improve the accuracy of fatigue life prediction of welded joints, soft computing methods, including entropy-based neighborhood rough set reduction algorithm, the particle swarm optimization (PSO) algorithm and support vector regression machine (SVRM) are combined to construct a fatigue life prediction model of titanium alloy welded joints. By using an entropy-based neighborhood rough set reduction algorithm, the influencing factors of the fatigue life of titanium alloy welded joints such as joint type, plate thickness, etc. are analyzed and the reduction results are obtained. Fatigue characteristic domains are proposed and determined subsequently according to the reduction results. The PSO-SVRM model for fatigue life prediction of titanium alloy welded joints is established in the suggested fatigue characteristic domains. Experimental results show that by taking into account the impact of joint type, the PSO-SVRM model could better predict the fatigue life of titanium alloy welded joints. The PSO-SVRM model indicates the relationship between fatigue life and fatigue life influencing factors in multidimensional space compared with the conventional least-square S-N curve fitting method, it could predict the fatigue life of the titanium alloy welded joints more accurately thus helps to the reliability design of the structure.

Textile composite are extensively used as structural materials for automotive, aerospace, energy, transportation and construction applications. During their service life these structures are subjected to different types of static and cyclic loading. For structural health monitoring of these structures, it is important to know the fatigue life and damage occurred at any stage of the life of the structure. Fatigue life is generally estimated using suitable life prediction model, while fatigue damage can be predicted by monitoring measurable damage parameters such as stiffness and strength. Two mathematical models namely fatigue life prediction model and stiffness degradation model are proposed for plain weave glass/epoxy composite subjected to flexural fatigue loading. Three different functions namely linear, exponential and sigmoid are evaluated to represent S-N diagram for plain weave glass/epoxy composite. Using predicted fatigue life along with initial modulus as inputs, the stiffness degradation model can predict residual stiffness at any stage of the fatigue loading life cycle. Logarithmic function used to represent stiffness degradation in the model is derived by inverting Boltzmann sigmoid function. The results of both, fatigue life model and stiffness degradation model were found to be in good agreement with those of the experimental results.

Defect sensitivity and high-cycle fatigue resistance of arc-welded 2219 aluminum alloy at 77 K

<scp>Fatigue &amp; Fracture of Engineering Materials &amp; Structures Virtual Special Issue</scp>: Data science and machine learning for fatigue and fracture assessment

This study investigated open-hole tension fatigue characteristic of carbon fiber-reinforced composite laminates subjected to constant amplitude cyclic loading. Two kinds of composite laminates, the CCF300/QY8911 and T300/QY8911, were used for fatigue testing with maximum loading of 70% notched tensile strength. Ultrasonic C-scan method was employed for fatigue damage evolution in composite laminates investigation, and specimens showed that the cracks initiated at the edge of the hole and free edge before final fracture. For most of existing fatigue models for determining material degradation and life of composite laminates have tended to be limited in application to specific materials, laminate stacking sequence and loading conditions; this paper proposed a probabilistic model for fatigue life prediction. This model utilized the equality relationship of failure probability between fatigue life and residual strength for fatigue life connection establishing under different loading stresses, by which the fatigue life or reliability life could be evaluated by the static tensile testing data that was one of the critical ultimate stress point on the S–N curve. In this paper, the tension-dominated fatigue lives of notched CCF300/QY8911 and T300/QY8911 composite laminates were evaluated by this model. To compare with the experimental data, the predicted results showed a good agreement.

A mechanistic model for fatigue life prediction of solder joints for electronic packages

Aiming to improve the predictive ability of Walker model for the life of turbine disc and taking an aircraft engine turbine disc made of GH4133 as the application example, this paper investigates the approach on probabilistic fatigue life prediction when considering parameters uncertainty inherent in the life prediction model, i.e. Walker model. Firstly, experimental data are used to update the model parameters with Bayes' theorem, so as to obtain the posterior probability distribution functions of two model parameters, as well to achieve the probabilistic model for life prediction of turbine disc. During the process of obtaining the posterior distribution, the Markov Chain Monte Carlo(MCMC) technique is employed for generating the samples of the given distribution and estimating the parameters distinctly; Secondly, the turbine disc life is predicted with the Walker probabilistic model by using MC (Monte Carlo) technique. The results show that: (1) under the condition of small scale data for turbine disc, parameters uncertainty of Walker model can be quantified and the corresponding probabilistic model for fatigue life prediction can be established by using Bayes' theorem; (2) There exists obvious dispersion of life data for turbine disc when predicting fatigue life in practical engineering application, which can be handled and calculated by the different survival rate of prediction life to meet the actual requirements.

Power-exponent function model for low-cycle fatigue life prediction and its applications – Part I: Models and validations

An intrinsic dissipation model for high-cycle fatigue life prediction

4 - Fatigue behavior of spot welded joints in steel sheets

Models commonly used in fatigue life prediction are based on cycles counted in different ways. The most used method is based on Rain Flow counting which takes care of the stress history in a very specific way. This method has three main drawbacks. It is an ad-hoc way to produce cycles from a continuously varying stress curve. It introduces a memory in the cycle counting in a very rigid way and the algorithm is quite complicated. A model based on level crossings is on the other hand easy to apply but the level crossing spectrum does not contain enough information about the stress history. Here a model is proposed where the damage accumulation depends on the actual level crossing and the stress history condensed in a state variable, as well. The proposed model has the following properties. Failure occurs when the total damage exceeds a given value. Every stress change causes a non-negative damage which depends only on the actual stress, its change and the stress state variable. In a specific application the state variable can be interpreted as the opening stress of a crack. The model is time invariant in the sense that the damage does not change if the time scale is transformed. Hence the life is determined by the sequence of maxima and minima of the stress. In general the dynamics of the state variable must be time invariant and stable in the sense that a stationary and ergodic random stress function shall generate a stationary and ergodic state variable. In this case it is possible to predict fatigue life in terms of a damage intensity, which is the expected damage per time unit. Transactions on Engineering Sciences vol 6, © 1994 WIT Press, www.witpress.com, ISSN 1743-3533 114 Localized Damage INTRODUCTION Fatigue life prediction under variable amplitude is usually based on the material properties, the stress history and some assumptions regarding the damaging process. The material properties are usually obtained from constant amplitude tests and the stress history is condensed in a suitable way by some counting technique. The damaging process is modelled by the Palmgren-Miner rule in connection with some additional assumptions. The commonly used model for fatigue life prediction at variable amplitude is the cycle counting approach. Here some cycle counting technique is defined and the stress history is condensed into a number of cycles at M levels. The total damage is then calculated according to the Palmgren-Miner rule,