Durability assessment is essential for assuring the reliability and safety of automotive structures, which subjects to complex random variable amplitude loadings. In this article, the application of different damage accumulation rules were discussed through the example of torsion beam rear axle from a sport utility vehicle, which is an assembly of welded thin sheet metals. Fatigue lives were calculated with these rules together with the structural stress at failure locations and the stress–life curve determined through the constant amplitude test of specially designed standard specimens stress distributions in the vicinity of failure locations. By comparing the calculated lives with test results, it is found that all the calculations with an stress–life curve for the probability of survival of Ps = 50% are conservative compared to the experimental mean value even the 95% confidence interval, and that with linear damage accumulation hypothesis is of better estimate, followed by the damage curve approach, while the double-linear rule will result in the most conservative estimation. Moreover, the modified Miner Rule, which extends the stress–life curve with a flatter slope k′ = 2 k − 1 below the fatigue limit, is recommended for structural fatigue design for its appropriate estimations corresponding to a low failure rate with high confidence level.
Lightweight design requires an accurate life prediction for structures and components under service loading histories. However, predicted life with the existing methods seems too conservative in some cases, leading to a heavy structure. Because these methods are established on the basis that load cycles would only cause fatigue damage, ignore the strengthening effect of loads. Based on Palmgren-Miner Rule (PMR), this paper introduces a new method for fatigue life prediction under service loadings by taking into account the strengthening effect of loads below the fatigue limit. In this method, the service loadings are classified into three categories: damaging load, strengthening load and none-effect load, and the process for fatigue life prediction is divided into two stages: stage I and stage II, according to the best strengthening number of cycles. During stage I, fatigue damage is calculated considering both the strengthening and damaging effect of load cycles. While during stage II, only the damaging effect is considered. To validate this method, fatigue lives of automobile half shaft and torsion beam rear axle are calculated based on the new method and traditional methods, such as PMR and Modified Miner Rule (MMR), and fatigue tests of the two components are conducted under service loading histories. The tests results show that the percentage errors of the predicted life with the new method to mean life of tests for the two components are −3.78% and −1.76% separately, much lesser than that with PMR and MMR. By considering the strengthening effect of loads below the fatigue limit, the new method can significantly improve the accuracy for fatigue life prediction. Thus lightweight design can be fully realized in the design stage.
Motor rotor magnetic bridges operate under multiple physical field loads, such as electromagnetic force, temperature, and centrifugal force. These loads can cause fatigue and aging failure of the bridges, especially when the rotor is operating continuously at high speeds and high temperatures. Therefore, the failure analysis and accelerated test cycle development of magnetic bridges is a major aspect of their reliability evaluation. This paper studies rotor multi-physics load transfer characteristics and establishes a rotor magnetic bridge failure physical model. A simulation analysis is conducted from the electromagnetic field, thermal field, structural field, and thermomechanical coupling field to determine the risk point load responses and failure-dominant loads. In addition, the accuracy of the simulation model is verified by actual bench tests. Considering the influence on the rotor bridge’s life under the coupling of multiple failure modes, the fatigue failure model under alternating loads and the fatigue aging coupling failure model are established, respectively. Through a damage analysis, the whole life cycle damage targets for both failure modes are determined, and the test condition levels are screened based on the load frequency distribution and damage distribution. The multi-objective optimization method is used to calculate the number of test cycles and finally develop accelerated test cycle conditions that can reproduce multiple failure modes. This research can provide support for rotor bridge reliability design and verification, as well as product quality development.