Fatigue life prediction of aluminum using artificial neural network Academic Article in Scopus uri icon

abstract

  • © 2021, International Association of Engineers. All rights reserved.Lightweight materials are currently being more used in mechanical components due to their mechanical be-haviour and lighter weight. To prevent failures on the service life is necessary to predict its fatigue life. To perform the life evaluation, the durability assessment is used to establish the damage applied by fatigue loads and the number of load cycles or spectrum repetitions versus material fatigue properties using its S-N curves to calculate the accumulated damage. This information is evaluated using a fatigue damage hypothesis. An artificial neural network (ANN) is proposed to predict the fatigue life based on material ultimate tensile strength (UTS). This is the first part of research to develop aluminum alloy through ANN based on the expected fatigue strength. The evaluation performed with results in literature with different types of aluminum: 5056, 2198-T851, 2024-T3 and 7050-T7451 has been proved that ANN can predict the fatigue life-improving its accuracy over traditional and modified damage rules.

publication date

  • January 1, 2021