Mechanical Engineering
Permanent URI for this communityhttp://hdl.handle.net/1903/2263
Browse
2 results
Search Results
Item A RISK-INFORMED DECISION-MAKING METHODOLOGY TO IMPROVE LIQUID ROCKET ENGINE PROGRAM TRADEOFFS(2013) Strunz, Richard; Herrmann, Jeffrey W.; Reliability Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This work provides a risk-informed decision-making methodology to improve liquid rocket engine program tradeoffs with the conflicting areas of concern affordability, reliability, and initial operational capability (IOC) by taking into account psychological and economic theories in combination with reliability engineering. Technical program risks are associated with the number of predicted failures of the test-analyze-and-fix (TAAF) cycle that is based on the maturity of the engine components. Financial and schedule program risks are associated with the epistemic uncertainty of the models that determine the measures of effectiveness in the three areas of concern. The affordability and IOC models' inputs reflect non-technical and technical factors such as team experience, design scope, technology readiness level, and manufacturing readiness level. The reliability model introduces the Reliability- As-an-Independent-Variable (RAIV) strategy that aggregates fictitious or actual hotfire tests of testing profiles that differ from the actual mission profile to estimate the system reliability. The main RAIV strategy inputs are the physical or functional architecture of the system, the principal test plan strategy, a stated reliability-bycredibility requirement, and the failure mechanisms that define the reliable life of the system components. The results of the RAIV strategy, which are the number of hardware sets and number of hot-fire tests, are used as inputs to the affordability and the IOC models. Satisficing within each tradeoff is attained by maximizing the weighted sum of the normalized areas of concern subject to constraints that are based on the decision-maker's targets and uncertainty about the affordability, reliability, and IOC using genetic algorithms. In the planning stage of an engine program, the decision variables of the genetic algorithm correspond to fictitious hot-fire tests that include TAAF cycle failures. In the program execution stage, the RAIV strategy is used as reliability growth planning, tracking, and projection model. The main contributions of this work are the development of a comprehensible and consistent risk-informed tradeoff framework, the RAIV strategy that links affordability and reliability, a strategy to define an industry or government standard or guideline for liquid rocket engine hot-fire test plans, and an alternative to the U.S. Crow/AMSAA reliability growth model applying the RAIV strategy.Item A Predictive Model of Nuclear Power Plant Crew Decision-Making and Performance in a Dynamic Simulation Environment(2009) Coyne, Kevin; Mosleh, Ali; Reliability Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The safe operation of complex systems such as nuclear power plants requires close coordination between the human operators and plant systems. In order to maintain an adequate level of safety following an accident or other off-normal event, the operators often are called upon to perform complex tasks during dynamic situations with incomplete information. The safety of such complex systems can be greatly improved if the conditions that could lead operators to make poor decisions and commit erroneous actions during these situations can be predicted and mitigated. The primary goal of this research project was the development and validation of a cognitive model capable of simulating nuclear plant operator decision-making during accident conditions. Dynamic probabilistic risk assessment methods can improve the prediction of human error events by providing rich contextual information and an explicit consideration of feedback arising from man-machine interactions. The Accident Dynamics Simulator paired with the Information, Decision, and Action in a Crew context cognitive model (ADS-IDAC) shows promise for predicting situational contexts that might lead to human error events, particularly knowledge driven errors of commission. ADS-IDAC generates a discrete dynamic event tree (DDET) by applying simple branching rules that reflect variations in crew responses to plant events and system status changes. Branches can be generated to simulate slow or fast procedure execution speed, skipping of procedure steps, reliance on memorized information, activation of mental beliefs, variations in control inputs, and equipment failures. Complex operator mental models of plant behavior that guide crew actions can be represented within the ADS-IDAC mental belief framework and used to identify situational contexts that may lead to human error events. This research increased the capabilities of ADS-IDAC in several key areas. The ADS-IDAC computer code was improved to support additional branching events and provide a better representation of the IDAC cognitive model. An operator decision-making engine capable of responding to dynamic changes in situational context was implemented. The IDAC human performance model was fully integrated with a detailed nuclear plant model in order to realistically simulate plant accident scenarios. Finally, the improved ADS-IDAC model was calibrated, validated, and updated using actual nuclear plant crew performance data. This research led to the following general conclusions: (1) A relatively small number of branching rules are capable of efficiently capturing a wide spectrum of crew-to-crew variabilities. (2) Compared to traditional static risk assessment methods, ADS-IDAC can provide a more realistic and integrated assessment of human error events by directly determining the effect of operator behaviors on plant thermal hydraulic parameters. (3) The ADS-IDAC approach provides an efficient framework for capturing actual operator performance data such as timing of operator actions, mental models, and decision-making activities.