Probabilistic methods developed at Garrett Auxiliary Power Division of Allied-Signal Aerospace Company under the “Life Prediction Methodology for Ceramic Components of Advanced Heat Engines” program sponsored by The Department of Energy/Oak Ridge National Laboratory (DOE/ORNL) under contract No. 86X-SC674C (WBS Element 3.2.2.3) are presented. Statistical methods have been developed to estimate Weibull strength parameters and component reliability with confidence limits for structural ceramics. Estimates can be made using pooled strength data from specimens of multiple sizes and loading conditions, from multiple test temperatures, and from material with multiple strength distributions. Bootstrap and likelihood ratio techniques are used to calculate confidence intervals on parameters and reliability estimates from these complex pooled data sets. A large database was generated on one ceramic (NT154 silicon nitride) to verify the methods. These statistical methods guide the development of standards for more accurate parameter estimation, to define component reliability requirements with confidence limits, and to plan specimen tests for more efficient estimation of Weibull parameters and component reliability.
The capability to perform accurate fast-fracture strength predictions for ceramic components under complex stress states must be available in order to transition the use of advanced, high-strength ceramic materials from the laboratory to the high-strength/high-temperature applications they are intended for. Multiaxial strength prediction theories have provided the prediction capabilities, but only limited testing of these theories under complex states of stress and stress gradient conditions has been performed previously. Presented here are comprehensive test results and strength predictions for ceramic components subjected to complex states of stress and stress gradient conditions. The results show excellent agreement of the predictions from the multiaxial theories with test results for volumetrically distributed flaws. An important finding of this work is the problem that arises in performing component surface strength predictions from database-type specimens. Database-type specimens and component surface properties are not necessary correlated, and in many cases it may be completely inaccurate to use database-type specimen surface properties for component surface strength predictions.
Statistical methods for design of ceramic components for fast fracture and time-dependent failure modes have been advanced to significantly enhance component reliability and reduce baseline data generation costs. The statistical methods are incorporated in AlliedSignal Engines CERAMIC and ERICA computer codes. Use of the codes facilitates generation of material strength and fatigue parameters by pooling data from specimens that are of varying sizes, stressed by dissimilar loading conditions, or tested at a variety of temperatures, and by including specimens that have been proof tested. The codes include approaches to calculate confidence bounds for Weibull and fatigue parameters of censored data as well as for predicted reliability of ceramic components. Rather than presenting highly detailed, previously published theoretical derivations (Cuccio, et al., 1994, Schenk, et al., 1998, Schenk, et al., 1999), this paper presents a summary of general methodology and a synopsis of code features and capabilities. The work was sponsored by the Department of Energy/Oak Ridge National Laboratory (DoE/ORNL) under Contract DE-AC05-84OR21400.
Statistical methods for the design of ceramic components for time-dependent failure modes have been developed which can significantly enhance component reliability, reduce baseline data generation costs, and lead to more accurate estimates of slow crack growth (SCG) parameters. These methods are incorporated into the AlliedSignal Engines CERAMIC and ERICA computer codes. Use of the codes facilitates generation of material strength parameters and SCG parameters simultaneously, by pooling fast fracture data from specimens that are of different sizes, or stressed by different loading conditions, with data derived from static fatigue experiments. The codes also include approaches to calculation of confidence bounds for the Weibull and SCG parameters of censored data and for the predicted reliability of ceramic components. This paper presents a summary of this new fatigue data analysis technique and an example demonstrating the capabilities of the codes with respect to time-dependent failure modes. This work was sponsored by the U.S. Department of Energy Oak Ridge National Laboratory (DoE/ORNL) under Contract No. DE-AC05-84OR21400.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.