A parametric probabilistic approach to quantify uncertainties in a non-linear cumulative fatigue damage model considering limited data
| Autor: | Dias, João Paulo, Ekwaro-Osire, Stephen, Cunha Jr, Americo, Alemayehu, Fisseha, Dabetwar, Shweta, Nispel, Abraham |
|---|---|
| Přispěvatelé: | Texas Tech University [Lubbock] (TTU), Universidade do Estado do Rio de Janeiro [Rio de Janeiro] (UERJ), West Texas A&M University, Texas Tech University, Universidade do Estado do Rio de Janeiro, CUNHA JR, Americo |
| Jazyk: | angličtina |
| Rok vydání: | 2018 |
| Předmět: |
[ SPI.MECA.GEME ] Engineering Sciences [physics]/Mechanics [physics.med-ph]/Mechanical engineering [physics.class-ph]
[MATH.MATH-PR] Mathematics [math]/Probability [math.PR] [SPI.MECA.SOLID]Engineering Sciences [physics]/Mechanics [physics.med-ph]/Solid mechanics [physics.class-ph] [SPI.MECA.MEMA] Engineering Sciences [physics]/Mechanics [physics.med-ph]/Mechanics of materials [physics.class-ph] [SPI.MECA.GEME]Engineering Sciences [physics]/Mechanics [physics.med-ph]/Mechanical engineering [physics.class-ph] [MATH.MATH-PR]Mathematics [math]/Probability [math.PR] [MATH.MATH-ST]Mathematics [math]/Statistics [math.ST] [ SPI.MECA.MEMA ] Engineering Sciences [physics]/Mechanics [physics.med-ph]/Mechanics of materials [physics.class-ph] [SPI.MECA.MEMA]Engineering Sciences [physics]/Mechanics [physics.med-ph]/Mechanics of materials [physics.class-ph] [ SPI.MECA.SOLID ] Engineering Sciences [physics]/Mechanics [physics.med-ph]/Mechanics of the solides [physics.class-ph] [SPI.MECA.SOLID] Engineering Sciences [physics]/Mechanics [physics.med-ph]/Solid mechanics [physics.class-ph] [ MATH.MATH-ST ] Mathematics [math]/Statistics [math.ST] [SPI.MECA.GEME] Engineering Sciences [physics]/Mechanics [physics.med-ph]/Mechanical engineering [physics.class-ph] [MATH.MATH-ST] Mathematics [math]/Statistics [math.ST] [ MATH.MATH-PR ] Mathematics [math]/Probability [math.PR] |
| Zdroj: | Twelfth International Conference on Fatigue Damage of Structural Materials (ICFDSM 2018) Twelfth International Conference on Fatigue Damage of Structural Materials (ICFDSM 2018), Sep 2018, Hyannis, United States Twelfth International Conference on Fatigue Damage of Structural Materials (ICFDSM 2018), Sep 2018, Hyannis, United States. 2018 |
| Popis: | International audience; The linear damage rule (LDR) have been widely used in engineering applications involving cumulative fatigue damage problems, however, this rule present limitations regarding the prediction of sequential variable loading effect. In order to overcome these deficiencies, non-linear damage models, such as the double linear damage rule (DLDR), has been proposed which intend to keep the simplicity of the LDR approach. Moreover, given the acknowledged stochastic nature of the cumulative fatigue damage, the few attempts to model probabilistically fatigue damage considering non-linear approaches mostly rely on the collection of a large amount of experimental data, which allow them to model uncertainties using certain probability distribution families, such as Weibull and log-normal. However, determination of large experimental datasets is time-consuming and expensive, and non-linear probabilistic fatigue approaches should also consider statistical methods to model uncertainties when limited amount of data is available. This work proposes a parametric probabilistic approach to quantify the uncertainties of material and model parameters in a non-linear cumulative fatigue damage model considering that limited data for the random variables is available. Firstly, a probabilistic DLDR model for two-loading level sequence is developed in which the equations for the calculation of the knee-point for the high-low and low-high loading sequences are derived as functions of the distributions of the constant amplitude fatigue lives and the model parameters. Second, considering that only the limit values, mean and variance are known, the probability distributions of the constant amplitude fatigue lives are determined using the maximum entropy principle, and their uncertainties propagated through the model parameters using a Monte Carlo sampling method to determine the distribution of the knee-points of the DLDR model. The proposed approach is validated using existing fatigue life data for two-loading level experiments available in the literature. Furthermore, the results are compared with the predictions obtained with an existing probabilistic DLDR methodology which considers Weibull distribution to model the fatigue random variables. The main contribution of this work relies on the introduction of uncertainty quantification techniques on the probabilistic modeling of cumulative fatigue damage considering the limited availability of data for the fatigue random variables. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|