No abstract

No abstract

We propose the use of regression models as a tool to reduce time and cost associated with the development and selection of new metallic alloys. A multiple regression model is developed which can accurately predict tensile yield strength of high strength low alloy steel based on its chemical composition and processing parameters. Quantile regression is used to model the fracture toughness response as measured by Charpy V-Notch (CVN) values, which exhibits substantial variability and is therefore not usefully modelled via standard regression with its focus on the mean. Using Monte-Carlo simulation, we determine that the three CVN values corresponding to each steel specimen can be plausibly modelled as observations from the 20th, 50th and 80th percentiles of the CVN distribution. Separate quantile regression models fitted at each of these percentile levels prove sufficiently accurate for ranking steels and selecting the best combinations of composition and processing parameters.

We report on estimating probability of failure (PoF) of an FA/18 bulkhead based on empirically obtained and published in open literature fatigue-crack growth and equivalent pre-crack size (EPS) distribution data. We demonstrate that the tail of the EPS distribution has a significant effect on PoF. Considering each flight to be a successful (i.e., no failure) fatigue test, we use a Bayesian approach to obtain an updated (posterior) EPS distribution and provide more accurate estimates of PoF. We then consider monitoring fatigue cracks in high stress concentration areas using fatigue damage sensors and show that using such sensor data (notably "no crack found" response) leads to a significant reduction of uncertainty in estimating the PoF. We also show the effect of increasing sensor accuracy (i.e., reliable detection of smaller cracks) on PoF predictions and required sensor interrogation intervals. The reported approach allows us to perform tradeoff studies on sensor accuracy and interrogation frequency for maintaining required levels of PoF.

We describe the effect of the equivalent pre-crack size (EPS) distribution errors on probability of failure (PF) of an FA/18 bulkhead. We distinguish two types of errors: (1) measurement errors of the EPS distribution parameters and (2) errors that arise because PF is typically calculated based on an EPS distribution with an infinite tail, while in a flying aircraft the tail cannot be unbounded. Our results indicate that both errors have a pronounced effect on accuracy of predicted PF. We also show that because predictions of the evolving crack sizes are less sensitive to the EPS distribution errors, they could be used as a complementary measure for estimating accumulated aircraft fatigue damage.