幂函数退化曲线快速测试方法

针对工程中常见的“先快后慢”的退化曲线,提出一种幂函数退化曲线快速测试方法,能够通过快速退化阶段试验数据或加上少量缓慢退化阶段试验数据对整条退化曲线进行拟合和外推。大量工程实例计算和Monte Carlo模拟结果表明:即使只有快速退化阶段试验数据,该方法得到的整条曲线也能很好地逼近退化曲线真值或偏于保守,工程上安全可用。分别对三参数幂函数退化曲线(橡胶老化曲线、S -N曲线、ε -N曲线、LED光强衰减曲线等)和四参数幂函数退化曲线(应力松弛曲线、蠕变持久曲线等)快速测试方法进行了详细讨论,给出相应的中值退化曲线和高置信度、高可靠度退化曲线。从所给两个实例对比可以看到,所提方法可以比传统方法节省2/3左右的试验时间和成本。 

Rapid determination method for degradation curve subject to power function

Many degradation curves illustrate a two-stage property where the deterioration is fast in the first stage and then becomes slow in the second stage. For this sort of degradation curves subject to power function, a rapid determination method was proposed, which can fit and extrapolate the complete degradation curve with the test data from fast degradation stage, or adding a small amount of test data from slow degradation stage. Abundant analyses of real applications and Monte Carlo simulations showed that, under this circumstance the obtained curve can also approach the true value, or be more conservative, which is safe and useable in engineering. Rapid determination methods for three-parameter power function degradation curve (rubber aging curve, S -N curve, ε -N curve, light intensity degrading curve of LED, etc) and four-parameter power function degradation curve (stress relaxation curve, creep rupture curve, and so forth) were discussed, corresponding median degradation curve and confidence limit curve with high confidence level and high reliability were further derived. According to the comparative analyses of two real applications, it can be concluded that the proposed method can save about two-third of test time and cost.