Abstract:

This is a special issue focusing on performance analysis of space vehicles. Methodologies, techniques, decision approaches, applications, and viewpoints are presented. The papers cover a diverse assortment of topics. Even though many are focused on performance assessments of space vehicles, the subjects covered are relevant to the field of performability engineering in general. I trust the readers will find the papers interesting and informative. A synopsis of each paper covering its unique and important contribution is outlined as follows:

Two Case Studies Illustrating the Management of Risks for the International Space Station

By: K. Carter-Journet, J. Calhoun, M. Raftery, and M. Lutomski, U.S.A.

Two case studies are described that illustrate how Probabilistic Risk Assessment (PRA) is practically used in risk-informed decision making for the International Space Station (ISS). The first case describes how risks from the impacts from micrometeoroid and orbital debris (MMOD) are controlled. The second study describes assessments of potentially increasing risks from commercial vehicles that will visit the ISS. These studies are some of the best examples of the actual implementation of risk-informed decision making.

Estimating the Risk of a New Launch Vehicle Using Historical Design Element Data

By: R. Cross, U.S.A.,

This paper describes an approach that has been developed by a working group consisting of NASA, the Federal Aviation Administration (FAA), and the Air Force. Historical data of new launch vehicles are analyzed to estimate the failure probabilities for given types of systems and functions that can then be appropriately synthesized to estimate the failure probability of a new launch vehicle. This approach offers an important alternative to traditional component-based models that generally underestimate the failure probability of a new launch vehicle.

Orbital Warehouse Design for an Extra-terrestrial Supply-Chain Distribution Model

By: N. Chari, U. Venkatadri, and C. Diallo, Canada

This is an interesting paper in the methodology used and the results that are obtained. A supply network is defined that involves the terrestrial launch facility, the ISS, a group of hotels, a warehouse, and a lunar base. Assuming given altitudes for the ISS and lunar base, optimal altitudes for the hotels and warehouse are determined by minimizing an expression for the total energy required in orbital transfers. Using these given altitudes, fuel and supply requirements are then determined. This approach has the potential for greater extensions and applications.

On the Estimation of Space Launch Vehicle Reliability

By: S. Guarro, U.S.A.

This paper reviews approaches used to estimate the reliability of a new space launch vehicle. These approaches are classified as two types-direct estimation approaches using past history and indirect bottom-up approaches using component level models and data. The features of the different approaches are then described along with the contributions they include as well as exclude. The significant value of the paper is the practical insights and conclusions that are derived based on past experience and past space launch vehicle history.

Shuttle Risk Progression-Focus on Historical Risk Increases

By: T. Hamlin, U.S.A.

The Space Shuttle PRA has been extensively used and updated to follow the upgrades made to the Space Shuttle. This paper retraces the evolution of the estimated Probability for the Loss of Crew and Vehicle (PLOCV). The focus is on the increased estimates of PLOCV that are obtained by backward retracing of the upgrades that were made up through the first flight. The risk progression evolution that is thereby obtained is a stepwise progression. The risk progression evolution can serve as a useful reference in estimating the risk evolution of a new launch vehicle.

Fiber Breakage Model for COPV Reliability Estimation

By: R. P. Heydorn and P. L. N. Murthy, U.S.A.

Because of their strength and light weight, composite overwrapped pressure vessels (COPVs) are employed in new systems including a number that are used on the ISS. An issue with COPVs is the potential occurrence of catastrophic stress rupture. This paper develops a formula for the reliability of a COPV that models the basic mechanistic process involved in the failure occurrence. Comparisons are made with the purely empirical Weibull model that is often used. The work provides a potential, significant improvement in COPV reliability estimation.

Use of a PRA in Supporting the Design of a GOES Weather Satellite and Ground System

By: P. Kalia, R. Pair, J. Uhlenbrock, V. Quaney, and Y. Shi, U.S.A.

This paper describes the employment of a PRA throughout development of design of the Geostationary Operational Environmental Satellite (GOES). The paper is interesting in describing the interactions between the PRA analysts and the designers and how specific PRA results were used to improve the design. The paper also describes the different considerations involved in modeling the ground support system for the satellite. The paper is a documented case of how a PRA provides cost-effective benefits in terms of the improvement of reliability during the design.

Reliability Characteristics of a Satellite Communication System Including Earth Station and Terrestrial System

By: K. Nagiya and M. Ram, India

This is an interesting paper in that it shows how basic Markov models can be applied to a satellite system and associated support systems to determine expressions for the associated reliability, availability, mean time to failure, and expected profit. Even though the models are basic, what is useful is the set of insights gained and general behaviors obtained by varying the different parameters over the range of values of interest. Sensitivity measures are also defined which give the rate of change. The models also serve as a basis for extensions to more complex models.

Informative Bayesian Quantification of Design Reliability Based on Test Characteristics and Test Results

By: W. E. Vesely, U.S.A.,

Often the design reliability of a new system is statistically estimated using only the successes and failures observed in given tests. This paper shows how design reliability estimates can be significantly improved by incorporating the basic characteristics of the tests conducted, the faults detected, and the corrections made. The number of tests needed to achieve a given reliability criterion can thereby be significantly reduced. Reliability growth can also be more effectively monitored. The approach can have significant impacts as shown by the variety of applications.

Systematic Quantification of the Prior Risk Assurance of a New System Using Bayesian Evidence Analysis

By: W. E. Vesely, U.S.A.,

The total risk assurance of a new system is obtained from initial assessments plus additional tests conducted. The initial assessments provide the prior risk assurance which can be expressed as the confidence in an acceptable risk. This paper uses evidence analysis principles to systematically combine the diverse qualitative and quantitative results obtained from the initial assessments to obtain the measure of the overall risk assurance. Even though used in many other fields, the approach is generally not used for risk assurance. The benefits can be significant.

Modeling Rate of Occurrence of Failures with Log-Gaussian Process Models-A Case Study for Prognosis and Health Management of a Fleet of Vehicles

By: M. Wayne and M. Modarres, U.S.A.

Log-Gaussian Process Regression (GPR) handles general non-linear relationships by specifying a general form for the kernel function which defines the covariances of the dependent variables. The kernel function contains different possible behaviors with associated weights (hyperparameters) which are estimated from data. As the paper shows, GPR is a powerful technique for monitoring general trends in failure occurrences. It also has the important potential for monitoring fault occurrences as well as failure occurrences in reliability growth applications.

Bayes Linear Bayes Graphical Models in the Design of Optimal Test Strategies

By: K. J. Wilson., J. Quigley, T. Bedford, and L. Walls, U.K.

Bayesian networks (BN) are used in a wide variety of applications. The problem in many practical applications is the explosion of states and variables that result and that must be input. This paper presents an important linear Bayes approach for addressing this problem which uses optimal linear relationships among the states and variables. Conjugate distributions are also used for discrete state probabilities allowing efficient updating. The applications described for optimizing test strategies show the power of the approach in modeling multiple decision alternatives.

Finally, I wish to thank all the contributors for their papers. I was truly gratified by their excellent papers and timely responses.

William E. Vesely received his B.S. in Physics in 1964 from Case Institute of Technology and his M.S. and Ph.D. in Nuclear Engineering in 1966 and 1968, respectively, from the University of Illinois. Dr. Vesely has been in the risk assessment field for over 40 years. He was a principal author of the first major Probabilistic Risk Assessment (PRA) performed on nuclear plants, WASH-1400. He worked at the Nuclear Regulatory Commission as a risk specialist, and has been a PRA consultant for the Department of Defense, Department of Energy, various National Laboratories and various companies. Dr. Vesely has developed numerous approaches for risk and reliability evaluations, including techniques for data mining, pattern recognition, and risk trending.

Dr. Vesely has published over 100 papers and reports on PRA, statistical analysis, data analysis, and expert systems. He has been an adjunct professor for several universities.

Dr. Vesely has assurance responsibilities for risk assessments carried out by NASA. He also has responsibility for methods and tool developments for risk assessments and reliability assessments. He lectures at NASA's Probabilistic Risk Assessment (PRA) Courses and teaches NASA Fault Tree Courses.

He is the principal author of the Fault Tree Handbook with Aerospace Applications published by NASA. He served as a technical coordinator for Space Shuttle PRA.