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.