Results for “FAILURES” — Page 8

NASA NTRSDate not supplied

More Data Needed for Failure Rate Estimation, Validation, and Uncertainty Reduction

Current Environmental Control and Life Support System (ECLSS) development and test activities are not generating data fast enough to provide statistically-supportable precise Orbital Replacement Unit (ORU) failure rate estimates for future missions. Accurate and precise failure rate estimates are critical for missions beyond Low Earth Orbit (LEO) because current risk mitigation approaches – namely regular resupply and rapid abort capabilities – will not be available. Safe operations will depend on mission planners’ ability to accurately forecast spares demand and efficiently provide the necessary resources. However, even after more than a decade of operations on board the International Space Station (ISS), a significant amount of uncertainty remains in failure rate estimates. Uncertain or inaccurate failure rates result in increased risk and spares mass for future missions. A Bayesian failure rate estimation approach, such as the one currently implemented by the ISS Program, can help reduce uncertainty by incorporating engineering judgement into failure rate estimates. However, experience on the ISS and with other complex systems shows that these prior failure rate estimates are often inaccurate. In addition, prior failure rate estimates are typically point values; some level of uncertainty must be added to convert these into probability distributions for Bayesian updating, and there are several potential methods for doing so. Due to the low rate of data collection, these subjective (and often inaccurate) prior estimates currently have a strong influence on the end result. This paper examines the challenges associated with failure rate estimation, validation, and uncertainty reduction in the context of ECLSS development for beyond-LEO missions. A variety of techniques for generating and updating Bayesian priors are discussed and evaluated using both real-world and simulated data. Potential solutions for improving failure rate estimation, including testing additional units, are analyzed and discussed, and a set of recommendations are provided for next-generation system development activities.

Supportability↗

NASA NTRSDate not supplied

Discovery and Analysis of Rare High-Impact Failure Modes using Adversarial RL-Informed Sampling

Adaptive learning agents have tremendous potential to handle critical tasks currently performed by humans. Unfortunately, due to their complexity, it can be difficult to verify that these learning agents do not have critical failure modes. Standard verification and validation methods often do not apply directly to learning agents and Monte Carlo methods have difficulty covering even a small fraction of the state space, especially in multiagent systems or over long time horizons. To overcome this difficulty, we demonstrate an adaptive stress-testing method based on reinforcement learning of correlations that raise the probability of failure. This approach has three key properties: (1) it is able to find rare failure modes with far greater sample efficiency than Monte Carlo methods, (2) it can estimate the true probability of a failure mode despite the inherent bias in the learning method, and (3) it is capable of learning and resampling compact representations of multimodal failure spaces. These properties are important in practice as we need to find disparate failure modes while accounting for their actual relevance. This is a significant advantage over traditional adaptive stress testing methods that give abstract likelihoods of particular failure instances, but cannot estimate the probability of a broader failure mode. We test our algorithm on a simple problem from the aviation domain where an autonomous aircraft lands in gusty wind conditions. The results suggest that we can find failure modes with far fewer samples than the Monte Carlo approach and simultaneously estimate the probability of failure.

reinforcement learning↗

DOE OSTI2021

Assessment of Cladding-Cladding Delamination Failures in U-Mo Fuel Plates

Monolithic fuel is a candidate fuel form for the conversion of high-performance research reactors. This plate-type fuel consists of a high-density, U-Mo alloy-based fuel in a monolithic form that is sandwiched between zirconium diffusion barriers and encapsulated in an aluminum cladding. To qualify this fuel system, the program must demonstrate that the fuel meets the safety standards, has a predictable behavior, and performs well in a reactor. Although many plates have shown satisfactory thermo-mechanical performance in reactor, a small number of experimental plates exhibited failures, rendering performance of these plates unacceptable. Post-irradiation examinations of failed plates have revealed that various failures modes are present, including pillowing, blistering, buckling, warping, delamination, fracture, rupture, oxide spallation, etc. Previous studies have implied that such various operational and loading conditions could generate a complex stress-strain state, facilitating distinct failure modes that could impact the structural integrity. To understand such various failure modes, a comprehensive set of studies focusing on failure modes is being performed. As a part of these studies, this work studies the cladding-cladding separation failures that were observed in some of previous irradiation tests. To understand the root cause of cladding separation failures, selected plates from previous irradiation experiments were simulated under comparable operational conditions. Stress-strain state of the plates at the interface were examined. The results were comparatively evaluated to understand possible effects of stress state of the bond-line on delamination failures. This paper presents these findings, discusses such failure modes, and examines the influence of various factors on previously observed cladding delamination failures.

42 ENGINEERING↗

DOE OSTI2025

Correlation Between Weather Alerts and Grid Component Failures for Grid Alert

Weather events cause most grid failures. Often, we even get notifications on our phones to take cover or be prepared for an imminent event. If electric grid utilities had a similar warning that also included probable scenarios and the equipment involved, they could prepare and minimize the effects. Recent research at Idaho National Laboratory into electric grid risk analysis methods resulted in a tool that allows for the development of the most likely scenarios given failure probabilities of grid components. INL has a project with the U.S. Department of Energy’s Cybersecurity, Energy Security, and Emergency Response (CESER) program to develop a Grid Alert application that receives messages from the existing emergency alert system, filters and determines components possibly affected by the emergency event, calculates probable scenarios uses MASTERRI and then notifies the utility if there is significant risk. Historical failure data of elements that comprise the U.S. electric grid have been compiled by utilities and organizations such as the international regulatory body North American Electric Reliability Corporation (NERC). Nominal failure rates are obtained from this data. To make this tool possible, estimated failure rates are needed for different component types given the alert type, severity, and location. Historic weather-related grid element failures are correlated with historic weather events from Integrated Public Alert & Warning System (IPAWS). These correlated events and failures are used along with Bayesian updates from the historical norms to provide a modified failure rate for grid elements in the alert areas and calculate probable scenarios. This discusses the Grid Alert project plan but focuses on the data gathered and process used in determining failure rates for possible grid failure scenarios.

24 - POWER TRANSMISSION AND DISTRIBUTION↗

DOE OSTI2024

Prognostic model and failure mechanisms of steam generators in Sodium-Cooled fast reactors

This paper presents a prognostic model for sodium-cooled fast reactor (SFR) steam generators (SGs). Here, the purpose of the model is to estimate the remaining useful life of SFR SGs and thus to support the decision-making of autonomous control. SFR SGs are of great interest for plant integrity due to their harsh operation environment. They operate at higher temperatures and higher coolant-to-steam pressure differences than those of current light water reactors (LWR). The severity of the SFR SG failure consequences, which include water-sodium contact, is another reason. Understanding its failure mechanisms is important for the development of its prognostic model. Based on our literature review and physics-based analysis, we concluded that creep would be a dominant degradation mode of SFR SGs due to SFR’s high-temperature and high-pressure environment. Thus, creep is the focus of the prognostic model. Various other failure modes were also investigated in this study. The mechanical fatigue due to flow-induced vibration is usually observed in early developmental phases and would not be an important issue during normal operation. The thermal fatigue due to thermal stripping is occasionally observed in other components in SFRs but does not affect SG integrity. Pure water stress corrosive crack, fretting, and so on are commonly observed in LWR SGs but are significantly less important in SFR SGs because of the high temperatures, high pressure differences, and chemical properties of liquid sodium. Based on these investigations, a prognostic model focusing on creep failures was developed. It estimates the failure probability profile by sampling the Larson-Miller parameter (LMP) and temperature and associated uncertainties through Monte Carlo methods. Two case studies were presented. The first one demonstrated the model’s capability to calculate the failure probability for a new specimen within a given time. If the tube is working under temperatures of 500C ± 3 and pressures of 15.2 MPa, which leads to LMP of 20,100 ± 50, the failure probabilities within 50, 70, and 90 years are approximately 0.1 %, 3.2 %, and 17.5 %. The second demonstrated how adjusting the workload can help to protect the integrity of the component. For an old tube reaching 99 % of its lifespan, continuing to run at temperatures of 500C ± 3 and LMP of 20,100 ± 50 leads to a failure probability of about 29.0 % within a year. If the temperature is reduced by 5C, the failure probability can be reduced to 2.5 %, and, if the pressure is also reduced such that the LMP is increased by 100, the failure probability can further be reduced to 0.12 %.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

DOE OSTI2016

SU-F-T-247: Collision Risks in a Modern Radiation Oncology Department: An Efficient Approach to Failure Modes and Effects Analysis

Purpose: Collisions between treatment equipment and patients are potentially catastrophic. Modern technology now commonly involves automated remote motion during imaging and treatment, yet a systematic assessment to identify and mitigate collision risks has yet to be performed. Failure modes and effects analysis (FMEA) is a method of risk assessment that has been increasingly used in healthcare, yet can be resource intensive. This work presents an efficient approach to FMEA to identify collision risks and implement practical interventions within a modern radiation therapy department. Methods: Potential collisions (e.g. failure modes) were assessed for all treatment and simulation rooms by teams consisting of physicists, therapists, and radiation oncologists. Failure modes were grouped into classes according to similar characteristics. A single group meeting was held to identify implementable interventions for the highest priority classes of failure modes. Results: A total of 60 unique failure modes were identified by 6 different teams of physicists, therapists, and radiation oncologists. Failure modes were grouped into four main classes: specific patient setups, automated equipment motion, manual equipment motion, and actions in QA or service mode. Two of these classes, unusual patient setups and automated machine motion, were identified as being high priority in terms severity of consequence and addressability by interventions. The two highest risk classes consisted of 33 failure modes (55% of the total). In a single one hour group meeting, 6 interventions were identified. Those interventions addressed 100% of the high risk classes of failure modes (55% of all failure modes identified). Conclusion: A class-based approach to FMEA was developed to efficiently identify collision risks and implement interventions in a modern radiation oncology department. Failure modes and interventions will be listed, and a comparison of this approach against traditional FMEA methods will be presented.

60 APPLIED LIFE SCIENCES↗

DOE OSTI2022

Experiments and modelling of adhesive failure initiation of an epoxy underfill: a study of electronic packaging survivability

Epoxy underfills can be implemented in electronic packaging to enhance solder joint reliability of surface mounted components. However, it is important for an engineer to have a failure criterion that can be used for failure predictions and redesign of electronic assemblies. For this study, data from epoxy bond failure in mock electronic part assemblies were correlated to finite element analyses to predict adhesive failure initiation. Experiments were performed to determine failure loads for various loading locations and nonlinear viscoelastic analyses were performed for the same loading locations to determine a maximum principal strain failure parameter. Predictions showed that a maximum principal strain failure parameter defined from one test could be used as an indicator of adhesive failure of an epoxy bond undergoing other modes of loading. Failure initiation predictions matched experimental data using a maximum principal strain failure parameter for an epoxy bond undergoing mixed modes of loading for both unfilled and alumina oxide filled 828DEA epoxy. Such experimental setup is deemed appropriate for future epoxy testing.

42 ENGINEERING↗

NASA NTRS1980

Sensor failure and multivariable control for airbreathing propulsion systems

A new sensor/actuator failure analysis technique for turbofan jet engines was developed. Three phases of failure analysis, namely detection, isolation, and accommodation are considered. Failure detection and isolation techniques are developed by utilizing the concept of Generalized Likelihood Ratio (GLR) tests. These techniques are applicable to both time varying and time invariant systems. Three GLR detectors are developed for: (1) hard-over sensor failure; (2) hard-over actuator failure; and (3) brief disturbances in the actuators. The probability distribution of the GLR detectors and the detectability of sensor/actuator failures are established. Failure type is determined by the maximum of the GLR detectors. Failure accommodation is accomplished by extending the Multivariable Nyquest Array (MNA) control design techniques to nonsquare system designs. The performance and effectiveness of the failure analysis technique are studied by applying the technique to a turbofan jet engine, namely the Quiet Clean Short Haul Experimental Engine (QCSEE). Single and multiple sensor/actuator failures in the QCSEE are simulated and analyzed and the effects of model degradation are studied.

Behbehani, K.↗

NASA NTRS1984

The evaluation of failure detection and isolation algorithms for restructurable control

Three failure detection and identification techniques were compared to determine their usefulness in detecting and isolating failures in an aircraft flight control system; excluding sensor and flight control computer failures. The algorithms considered were the detection filter, the Generalized Likelihood Ratio test and the Orthogonal Series Generalized Likelihood Ratio test. A modification to the basic detection filter is also considered which uses secondary filtering of the residuals to produce unidirectional failure signals. The algorithms were evaluated by testing their ability to detect and isolate control surface failures in a nonlinear simulation of a C-130 aircraft. It was found that failures of some aircraft controls are difficult to distinguish because they have a similar effect on the dynamics of the vehicle. Quantitative measures for evaluating the distinguishability of failures are considered. A system monitoring strategy for implementing the failure detection and identification techniques was considered. This strategy identified the mix of direct measurement of failures versus the computation of failure necessary for implementation of the technology in an aircraft system.

Motyka, P.↗

NASA NTRS1987

Characteristics of Space Shuttle Main Engine failures

During development and operation of the Space Shuttle Main Engine (SSME), 27 ground test failures of sufficient severity to be termed 'major incident' have occurred. Resourecs including NASA Failure Investigation Board reports, contractor failure reports, originally recorded data, along with engineering notes, data bases, and presentations connected with the failures were available for compilation into the engine failure review presented in this paper. Most SSME failures were a result of design deficiencies stemming from inadequate definition of dynamic loads. High cycle fatigue was the most frequent mechanism leading to failure. Eighteen of the 27 failures occurred during constant power level operation. Formal board reports were not available for all failures. Therefore, the failure history presented in this paper is not complete or of uniform quality.

Cikanek, Harry A., III↗

NASA NTRS1992

Postbuckling failure of composite plates with holes

This paper summarizes a study focused on understanding the failure mechanisms present in a plate with a centrally located circular hole loaded inplane into the postbuckling range of deflections. The study was numerical and experimental in nature and had as a goal the a priori prediction of failure using existing failure data from several sources, together with a stress analysis. The maximum stress failure criterion was used to predict failure and both intra- and interlaminar stresses were considered. Four laminates were considered. The phenomenon of modal interaction, or the jumping from one deformed configuration to another, is discussed. With this jumping, the plate configuration may change from one half-wave in the loading direction to two half-waves in the loading direction. The study concludes that the failure in (+/-45/0/90)2s and (+/-45/2(2))2s laminates is due to fiber compression failure and is predictable. The failure load is, for the most part, independent of response configuration. The (+/-45/0/(6))s laminate fails due to interlaminar shear along the simple support, while the response and failure of the (+/-45)4s laminate is governed to a large degree by material nonlinearities and progressive failure.

Lee, H. H.↗

NASA NTRS1992

An improved approach for flight readiness certification: Probabilistic models for flaw propagation and turbine blade failure. Volume 1: Methodology and applications

An improved methodology for quantitatively evaluating failure risk of spaceflight systems to assess flight readiness and identify risk control measures is presented. This methodology, called Probabilistic Failure Assessment (PFA), combines operating experience from tests and flights with analytical modeling of failure phenomena to estimate failure risk. The PFA methodology is of particular value when information on which to base an assessment of failure risk, including test experience and knowledge of parameters used in analytical modeling, is expensive or difficult to acquire. The PFA methodology is a prescribed statistical structure in which analytical models that characterize failure phenomena are used conjointly with uncertainties about analysis parameters and/or modeling accuracy to estimate failure probability distributions for specific failure modes. These distributions can then be modified, by means of statistical procedures of the PFA methodology, to reflect any test or flight experience. State-of-the-art analytical models currently employed for designs failure prediction, or performance analysis are used in this methodology. The rationale for the statistical approach taken in the PFA methodology is discussed, the PFA methodology is described, and examples of its application to structural failure modes are presented. The engineering models and computer software used in fatigue crack growth and fatigue crack initiation applications are thoroughly documented.

Moore, N. R.↗

NASA NTRS1992

An improved approach for flight readiness certification: Probabilistic models for flaw propagation and turbine blade failure. Volume 2: Software documentation

An improved methodology for quantitatively evaluating failure risk of spaceflights systems to assess flight readiness and identify risk control measures is presented. This methodology, called Probabilistic Failure Assessment (PFA), combines operating experience from tests and flights with analytical modeling of failure phenomena to estimate failure risk. The PFA methodology is of particular value when information on which to base an assessment of failure risk, including test experience and knowledge of parameters used in analytical modeling, is expensive or difficult to acquire. The PFA methodology is a prescribed statistical structure in which analytical models that characterize failure phenomena are used conjointly with uncertainties about analysis parameters and/or modeling accuracy to estimate failure probability distributions for specific failure modes. These distributions can then be modified, by means of statistical procedures of the PFA methodology, to reflect any test or flight experience. State-of-the-art analytical models currently employed for design, failure prediction, or performance analysis are used in this methodology. The rationale for the statistical approach taken in the PFA methodology is discussed, the PFA methodology is described, and examples of its application to structural failure modes are presented. The engineering models and computer software used in fatigue crack growth and fatigue crack initiation applications are thoroughly documented.

Moore, N. R.↗

NASA NTRS2001

Evaluation of a Linear Cumulative Damage Failure Model for Epoxy Adhesive

Recently a significant amount of work has been conducted to provide more complex and accurate material models for use in the evaluation of adhesive bondlines. Some of this has been prompted by recent studies into the effects of residual stresses on the integrity of bondlines. Several techniques have been developed for the analysis of bondline residual stresses. Key to these analyses is the criterion that is used for predicting failure. Residual stress loading of an adhesive bondline can occur over the life of the component. For many bonded systems, this can be several years. It is impractical to directly characterize failure of adhesive bondlines under a constant load for several years. Therefore, alternative approaches for predictions of bondline failures are required. In the past, cumulative damage failure models have been developed. These models have ranged from very simple to very complex. This paper documents the generation and evaluation of some of the most simple linear damage accumulation tensile failure models for an epoxy adhesive. This paper shows how several variations on the failure model were generated and presents an evaluation of the accuracy of these failure models in predicting creep failure of the adhesive. The paper shows that a simple failure model can be generated from short-term failure data for accurate predictions of long-term adhesive performance.

Richardson, David E.↗

NASA NTRS2004

Designing to Mitigate Food Growing Failures in Space

Future space life support systems may use crop plants to grow most of the crew s food. A harvest failure can reduce the food available for future consumption. If the previously stored food is insufficient to reach the next harvest, the crew may go hungry. This paper considers how the overall food supply system should be modified to cope with food production failures. The food supply concept for a mission will use grown food, or stored food, cIr both. The optimum food supply mix depends on the costs and failure probabilities of stored and grown food. A simple food system model assumes that either we obtain the nominal harvest or a failure occurs and no food is harvested. Given the probability that any particular harvest fails, it is easy to compute the expected number of failures and the total food shortfall over a mission. If some food is grown and the probability of harvest failure is high, a non-redundant system has an unacceptable likelihood that the crew will have no food for a full harvest period. Food supply reliability must be increased either by supplying more food initially or by increasing food production capacity. We can obtain a very reliable food supply even when the harvest failure rate is high. If the cost of growing food is much less than the cost of providing stored food, it is better to provide redundant food growing capacity than to increase initial storage. A more realistic biomass production failure model allows the harvest amount or time to vary around the nominal values, using stochastic modeling with repeated Monte Carlo simulation, but such failures have minor impact compared to a complete harvest failure.

Jones, Harry↗

NASA NTRS2008

Factors Influencing Progressive Failure Analysis Predictions for Laminated Composite Structure

Progressive failure material modeling methods used for structural analysis including failure initiation and material degradation are presented. Different failure initiation criteria and material degradation models are described that define progressive failure formulations. These progressive failure formulations are implemented in a user-defined material model for use with a nonlinear finite element analysis tool. The failure initiation criteria include the maximum stress criteria, maximum strain criteria, the Tsai-Wu failure polynomial, and the Hashin criteria. The material degradation model is based on the ply-discounting approach where the local material constitutive coefficients are degraded. Applications and extensions of the progressive failure analysis material model address two-dimensional plate and shell finite elements and three-dimensional solid finite elements. Implementation details are described in the present paper. Parametric studies for laminated composite structures are discussed to illustrate the features of the progressive failure modeling methods that have been implemented and to demonstrate their influence on progressive failure analysis predictions.

Knight, Norman F., Jr.↗

NASA NTRSDate not supplied

Discovery and Analysis of Rare High-Impact Failure Modes using Adversarial RL-Informed Sampling

Adaptive learning agents have tremendous potential to handle critical tasks currently performed by humans. Unfortunately, due to their complexity, it can be difficult to verify that these learning agents do not have critical failure modes. Standard verification and validation methods often do not apply directly to learning agents and Monte Carlo methods have difficulty covering even a small fraction of the state space, especially in multiagent systems or over long time horizons. To overcome this difficulty, we demonstrate an adaptive stress-testing method based on reinforcement learning of correlations that raise the probability of failure. This approach has three key properties: (1) it is able to find rare failure modes with far greater sample efficiency than Monte Carlo methods, (2) it can estimate the true probability of a failure mode despite the inherent bias in the learning method, and (3) it is capable of learning and resampling compact representations of multimodal failure spaces. These properties are important in practice as we need to find disparate failure modes while accounting for their actual relevance. This is a significant advantage over traditional adaptive stress testing methods that give abstract likelihoods of particular failure instances, but cannot estimate the probability of a broader failure mode. We test our algorithm on a simple problem from the aviation domain where an autonomous aircraft lands in gusty wind conditions. The results suggest that we can find failure modes with far fewer samples than the Monte Carlo approach and simultaneously estimate the probability of failure.

Validation↗

NASA NTRSDate not supplied

Discovery and Analysis of Rare High-Impact Failure Modes using Adversarial RL-Informed Sampling

Adaptive learning agents have tremendous potential to handle critical tasks currently performed by humans. Unfortunately, due to their complexity, it can be difficult to verify that these learning agents do not have critical failure modes. Standard verification and validation methods often do not apply directly to learning agents and Monte Carlo methods have difficulty covering even a small fraction of the state space, especially in multiagent systems or over long time horizons. To overcome this difficulty, we demonstrate an adaptive stress-testing method based on reinforcement learning of correlations that raise the probability of failure. This approach has three key properties: (1) it is able to find rare failure modes with far greater sample efficiency than Monte Carlo methods, (2) it can estimate the true probability of a failure mode despite the inherent bias in the learning method, and (3) it is capable of learning and resampling compact representations of multimodal failure spaces. These properties are important in practice as we need to find disparate failure modes while accounting for their actual relevance. This is a significant advantage over traditional adaptive stress testing methods that give abstract likelihoods of particular failure instances, but cannot estimate the probability of a broader failure mode. We test our algorithm on a simple problem from the aviation domain where an autonomous aircraft lands in gusty wind conditions. The results suggest that we can find failure modes with far fewer samples than the Monte Carlo approach and simultaneously estimate the probability of failure.

Validation↗