Modeling high-performance computing application-failure and hardware-failure data
Abstract Not Provided
SEARCH · Search NASA
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Abstract Not Provided
NASA has been engaged with additively manufactured (AM) process and component development since the 2000’s. AM offers various technical advantages, such as enhanced hardware design complexity, part consolidation, and processing of novel alloys in addition to programmatic advantages for reduction in processing time and cost. The focus of much of the AM development at NASA has been to mature the various processes, characterize material properties, develop standards, produce demonstrator parts, and integrate AM hardware in liquid rocket engines. These aspects have been demonstrated through process and design iterations using a methodical characterization, test-fail-fix cycles, as well as application and dissemination of lessons learned. In addition to these fundamental demonstrations of the AM process and hardware development, alloys that provide performance advantages in the high temperature and high-pressure environments have been matured for use in rocket engines. These environments are challenging for any alloy and any design, and the AM process is required to fully meet the intended design requirements. The importance of proper AM process was made evident in the failure of a Laser Powder Bed Fusion (L-PBF) copper-alloy combustion chamber during a hot-fire test due to a degraded material quality resulted from an AM process issue. The hot-fire test aimed to demonstrate high duty cycle under a risk-tolerant development project, where consequences of component failure would be minimal. However, the unintentional component failure emphasized the necessity of robust material characterization and rigorous process control procedures for the safe use of AM components in critical applications. In part, such concerns motivate the AM certification approach that NASA has recently adopted in NASA-STD-6030 “Additive Manufacturing Requirements for Spaceflight Systems”. This presentation provides an overview of the previously mentioned failure, a discussion on the evaluation of the failed chamber and supplemental chambers produced at the same time, a representative material samples that included intentional build witness lines, and a summary of the key results and recommendations from the evaluations. NASA continues to approach AM processes and designs with a level of risk and acceptance of failures that is appropriate for the project objectives, with the overall goal of safe implementation of AM technology and transferring AM technology into commercial space applications. The objective of this presentation is to provide awareness to the community working critical and non-critical AM components and the lessons learned on proper implementation of AM.
The collection and analysis of data concerning intermittent dailures in digital devices was performed using data from a computer design for shipboard usage. The failure data consisted of actual field failures classified by failure mechanisms and their likelihood of having been intermittent, potentially intermittent, or hard. Each class was studies with respect to computer operation in the ranges of 0 to 2,000 hours, 0 to 5, hours, and 0 to 10,000 hours. The study was done at the computer level as well as the microcircuit level. Results indicate that as age increases, the quasi-intermittent failure rate increases and the mean time to failure descreases.
UNKNOWN
Explore the source record for details and available documents.
In the probabilistic safety assessments (PSAs) used in the nuclear industry, human failure events (HFEs) are determined as a subset of hardware failures, namely those hardware failures that could be triggered by human action or inaction. This approach is top-down, starting with hardware faults and deducing human contributions to those faults. Elsewhere, more traditionally human factors driven approaches would tend to look at opportunities for human errors first in a task analysis and then identify which of those errors is risk significant. The intersection of top-down and bottom-up approaches to defining HFEs has not been carefully studied. Ideally, both approaches should arrive at the same set of HFEs. This question remains central as human reliability analysis (HRA) methods are generalized to new domains like oil and gas. The HFEs used in nuclear PSAs tend to be top-down— defined as a subset of the PSA—whereas the HFEs used in petroleum quantitative risk assessments (QRAs) are more likely to be bottom-up—derived from a task analysis conducted by human factors experts. The marriage of these approaches is necessary in order to ensure that HRA methods developed for top-down HFEs are also sufficient for bottom-up applications.
In the probabilistic risk assessments (PRAs) used in the nuclear industry, human failure events (HFEs) are determined as a subset of hardware failures, namely those hardware failures that could be triggered by human action or inaction. This approach is top-down, starting with hardware faults and deducing human contributions to those faults. Elsewhere, more traditionally human factors driven approaches would tend to look at opportunities for human errors first in a task analysis and then identify which of those errors is risk significant. The intersection of top-down and bottom-up approaches to defining HFEs has not been carefully studied. Ideally, both approaches should arrive at the same set of HFEs. This question is crucial, however, as human reliability analysis (HRA) methods are generalized to new domains like oil and gas. The HFEs used in nuclear PRAs tend to be top-down—defined as a subset of the PRA—whereas the HFEs used in petroleum quantitative risk assessments (QRAs) often tend to be bottom-up—derived from a task analysis conducted by human factors experts. The marriage of these approaches is necessary in order to ensure that HRA methods developed for top-down HFEs are also sufficient for bottom-up applications.
In the probabilistic safety assessments (PSAs) used in the nuclear industry, human failure events (HFEs) are determined as a subset of hardware failures, namely those hardware failures that could be triggered or exacerbated by human action or inaction. This approach is top-down, starting with hardware faults and deducing human contributions to those faults. Elsewhere, more traditionally human factors driven approaches would tend to look at opportunities for human errors first in a task analysis and then identify which of those errors is risk significant. The intersection of top-down and bottom-up approaches to defining HFEs has not been carefully studied. Ideally, both approaches should arrive at the same set of HFEs. This question remains central as human reliability analysis (HRA) methods are generalized to new domains like oil and gas. The HFEs used in nuclear PSAs tend to be top-down—defined as a subset of the PSA—whereas the HFEs used in petroleum quantitative risk assessments (QRAs) are more likely to be bottom-up—derived from a task analysis conducted by human factors experts. The marriage of these approaches is necessary in order to ensure that HRA methods developed for top-down HFEs are also sufficient for bottom-up applications. In this paper, I first review top-down and bottom-up approaches for defining HFEs and then present a seven-step guideline to ensure a task analysis completed as part of human error identification decomposes to a level suitable for use as HFEs. This guideline illustrates an effective way to bridge the bottom-up approach with top-down requirements.
A major requirement for an autonomous robot is the capability to diagnose faults during plan execution in an uncertain environment. Many diagnostic researches concentrate only on hardware failures within an autonomous robot. Taking a different approach, the implementation of a Telerobot Diagnostic System that addresses, in addition to the hardware failures, failures caused by unexpected event changes in the environment or failures due to plan errors, is described. One feature of the system is the utilization of task-plan knowledge and context information to deduce fault symptoms. This forward deduction provides valuable information on past activities and the current expectations of a robotic event, both of which can guide the plan-execution inference process. The inference process adopts a model-based technique to recreate the plan-execution process and to confirm fault-source hypotheses. This technique allows the system to diagnose multiple faults due to either unexpected plan failures or hardware errors. This research initiates a major effort to investigate relationships between hardware faults and plan errors, relationships which were not addressed in the past. The results of this research will provide a clear understanding of how to generate a better task planner for an autonomous robot and how to recover the robot from faults in a critical environment.
A major requirement for an autonomous robot is the capability to diagnose faults during plan execution in an uncertain environment. Many diagnostic researches concentrate only on hardware failures within an autonomous robot. Taking a different approach, the implementation of a Telerobot Diagnostic System that addresses, in addition to the hardware failures, failures caused by unexpected event changes in the environment or failures due to plan errors, is described. One feature of the system is the utilization of task-plan knowledge and context information to deduce fault symptoms. This forward deduction provides valuable information on past activities and the current expectations of a robotic event, both of which can guide the plan-execution inference process. The inference process adopts a model-based technique to recreate the plan-execution process and to confirm fault-source hypotheses. This technique allows the system to diagnose multiple faults due to either unexpected plan failures or hardware errors. This research initiates a major effort to investigate relationships between hardware faults and plan errors, relationships which were not addressed in the past. The results of this research will provide a clear understanding of how to generate a better task planner for an autonomous robot and how to recover the robot from faults in a critical environment.
After the Fukushima Daiichi power plant disaster in 2011, the United States Congress ordered the Department of Energy (DOE) to investigate new fuel and cladding types that would be more tolerant in an accident similar to the accident at Fukushima. The resulting program was dubbed the Accident Tolerant Fuels (ATF) program. As a part of this program, insertion is planned for an experiment to go into an ATR loop, and physics analysis must be performed. Discussed are the test fission power, void worth, and hardware failure reactivity insertion. For the sensor qualification test, the test fission power was 0 kW, the void worth was 0.42$, and the hardware failure was 0.01$ and 0.04$. For the fueled experiment, the safety case fission power was 170 kW, the void worth was 0.34$, and the hardware failure was 0$ and 0.03$.
The strategy of using multiple versions of independently developed software as a means to tolerate residual software design faults is suggested by the success of hardware redundancy for tolerating hardware failures. Although, as generally accepted, the independence of hardware failures resulting from physical wearout can lead to substantial increases in reliability for redundant hardware structures, a similar conclusion is not immediate for software. The degree to which design faults are manifested as independent failures determines the effectiveness of redundancy as a method for improving software reliability. Interest in multi-version software centers on whether it provides an adequate measure of increased reliability to warrant its use in critical applications. The effectiveness of multi-version software is studied by comparing estimates of the failure probabilities of these systems with the failure probabilities of single versions. The estimates are obtained under a model of dependent failures and compared with estimates obtained when failures are assumed to be independent. The experimental results are based on twenty versions of an aerospace application developed and certified by sixty programmers from four universities. Descriptions of the application, development and certification processes, and operational evaluation are given together with an analysis of the twenty versions.
Modern gas turbine engines are typically operated in a lean-premixed configuration in order to meet emissions requirements. Lean operation makes gas turbines more susceptible to combustion instability, which can cause catastrophic hardware failure if left unchecked. The spatially uneven distribution of fuel in a gas turbine combustor, known as fuel staging, is an effective way of suppressing instabilities. This dissertation examines the effect of both steady-state and transient fuel-staging in a model gas turbine combustor. First, the steady-state effect of fuel staging on nominally unstable operating conditions is examined. At 200 degrees Celsius and a bulk flow velocity of 26 m/s, the combustor undergoes self-excited instabilities when the equivalence ratio is between φ = 0.70 and φ = 0.74 when all nozzles are fueled equally. These instabilities are suppressed when the center nozzle equivalence ratio is increased to φ Staging = 0.85 and the outer nozzle equivalence ratio maintained at φ = 0.70 or when the global equivalence ratio is maintained at φ = 0.70 by increasing the center nozzle equivalence ratio to φ Staging = 0.74 and decreasing the outer nozzle equivalence ratio to φ Staging = 0.69. The staged p' RMS amplitudes are found to decrease with increasing equivalence ratio and the damping rates found to increase with increasing equivalence ratio. Next, axisymmetric fuel staging transients are conducted valve opening/closing timescales of 1, 16, 58, 4,000, and 10,000 ms, and equivalence ratios of φ staging = 0.75, 0.80, 0.85. It should be noted that these valve opening/closing timescales, referred to as τ Transient , differ from the actual fuel convection timescale that are summarized in Table 5.2. It is found that the combustor pressure fluctuation at the end of the transient strongly depends on staging equivalence ratio but not on τ Transient . Increasing the staging equivalence ratio decreased the staged- stable combustor pressure amplitude, similarly to the steady-state results. The staged-stable damping rate trends are insensitive to to τ Transient and increased with increasing staging equivalence ratio. The timescale of the transition, τ, is found to be sensitive to both τ Transient and the staging equivalence ratio. For a Modern gas turbine engines are typically operated in a lean-premixed configuration in order to meet emissions requirements. Lean operation makes gas turbines more susceptible to combustion instability, which can cause catastrophic hardware failure if left unchecked. The spatially uneven distribution of fuel in a gas turbine combustor, known as fuel staging, is an effective way of suppressing instabilities. This dissertation examines the effect of both steady-state and transient fuel-staging in a model gas turbine combustor. First, the steady-state effect of fuel staging on nominally unstable operating conditions is examined. At 200 degrees Celsius and a bulk flow velocity of 26 m/s, the combustor undergoes self-excited instabilities when the equivalence ratio is between φ = 0.70 and φ = 0.74 when all nozzles are fueled equally. These instabilities are suppressed when the center nozzle equivalence ratio is increased to φ Staging = 0.85 and the outer nozzle equivalence ratio maintained at φ = 0.70 or when the global equivalence ratio is maintained at φ = 0.70 by increasing the center nozzle equivalence ratio to φ Staging = 0.74 and decreasing the outer nozzle equivalence ratio to φ Staging = 0.69. The staged p' RMS amplitudes are found to decrease with increasing equivalence ratio and the damping rates found to increase with increasing equivalence ratio. Next, axisymmetric fuel staging transients are conducted valve opening/closing timescales of 1, 16, 58, 4,000, and 10,000 ms, and equivalence ratios of φ staging = 0.75, 0.80, 0.85. It should be noted that these valve opening/closing timescales, referred to as τ Transient , differ from the actual fuel convection timescale that are summarized in Table 5.2. It is found that the combustor pressure fluctuation at the end of the transient strongly depends on staging equivalence ratio but not on τ Transient . Increasing the staging equivalence ratio decreased the staged-stable combustor pressure amplitude, similarly to the steady-state results. The staged-stable damping rate trends are insensitive to to τ Transient and increased with increasing staging equivalence ratio. The timescale of the transition, τ, is found to be sensitive to both τ Transient and the staging equivalence ratio. For a Modern gas turbine engines are typically operated in a lean-premixed configuration in order to meet emissions requirements. Lean operation makes gas turbines more susceptible to combustion instability, which can cause catastrophic hardware failure if left unchecked. The spatially uneven distribution of fuel in a gas turbine combustor, known as fuel staging, is an effective way of suppressing instabilities. This dissertation examines the effect of both steady-state and transient fuel-staging in a model gas turbine combustor. First, the steady-state effect of fuel staging on nominally unstable operating conditions is examined. At 200 degrees Celsius and a bulk flow velocity of 26 m/s, the combustor undergoes self-excited instabilities when the equivalence ratio is between φ = 0.70 and φ = 0.74 when all nozzles are fueled equally. These instabilities are suppressed when the center nozzle equivalence ratio is increased to φ Staging = 0.85 and the outer nozzle equivalence ratio maintained at φ = 0.70 or when the global equivalence ratio is maintained at φ = 0.70 by increasing the center nozzle equivalence ratio to φ Staging = 0.74 and decreasing the outer nozzle equivalence ratio to φ Staging = 0.69. The staged p' RMS amplitudes are found to decrease with increasing equivalence ratio and the damping rates found to increase with increasing equivalence ratio. Next, axisymmetric fuel staging transients are conducted valve opening/closing timescales of 1, 16, 58, 4,000, and 10,000 ms, and equivalence ratios of φ staging = 0.75, 0.80, 0.85. It should be noted that these valve opening/closing timescales, referred to as τ Transient , differ from the actual fuel convection timescale that are summarized in Table 5.2. It is found that the combustor pressure fluctuation at the end of the transient strongly depends on staging equivalence ratio but not on τ Transient . Increasing the staging equivalence ratio decreased the staged- stable combustor pressure amplitude, similarly to the steady-state results. The staged-stable damping rate trends are insensitive to to τT Transient and increased with increasing staging equivalence ratio. The timescale of the transition, τ, is found to be sensitive to both τ Transient and the staging equivalence ratio. For a given τ Transient , the decay timescale decreased with increasing staging equivalence ratio. For 1, 16, and 58 ms transients, τ does not depend on τ Transient , though it does strongly depend on direction, where instability onset cases have longer and more variable τ than instability decay cases. The same dependence of τ on direction is seen for the 4,000 and 10,000 ms transients. However, unlike the short duration transients, τ does depend on τ Transient as would be expected. Phase space reconstruction shows the instability is a harmonic limit cycle with a single frequency. The instantaneous phase images shows the center region of the combustor goes out-of-phase first followed by the rest of the combustor. Finally, asymmetric fuel staging tests are conducted. It is found that non-axisymmetric staging can be as effective as axisymmetric staging, although small hardware variations between nozzles can affect the the minimum staging equivalence ratio that consistently suppresses the instability. These changes in the bifurcation equivalence ratio can make staging configurations appear less effective; however, both axisymmetric and non-axisymmetric staging configurations exhibit similar staged-stable p' RMS , damping rates, and instability transition times for equivalence ratios greater than the bifurcation equivalence ratio. Instantaneous phase images show similar instantaneous phase relationship structures in the staged nozzle (center or outer), suggesting that the instability suppression mechanism is similar in each case. This result is surprising given that the time-averaged flame shapes are very different between axisymmetric and non-axisymmetric staging configurations. The similar effectiveness of axisymmetric and non-axisymmetric staging in this geometry is attributed to the fact that the instability mode is a nominally axisymmetric plane wave. It is hypothesized that non-axisymmetric staging configurations will likely be more effective than axisymmetric staging configurations when the instability mode shape is asymmetric (such as a circumferential wave).
Hardware failures are inevitable on large high performance computing systems. Faults or performance degradations in the high-speed network can reduce the entire system s performance. Since the introduction of the Gemini interconnect, Cray systems have become resilient to many networking faults that were fatal in their previous generation systems. These new network reliability and resiliency features have enabled higher uptimes on Cray systems by allowing them to continue running with reduced network performance. Oak Ridge National Laboratory has developed a set of user-level diagnostics that stresses the high-speed network and searches for components that are not performing as expected. Nearest-neighbor bandwidth tests check every network chip and network link in the system. Additionally, performance counters stored on the network ASIC s memory mapped registers (MMRs) are used to better understand the state of the network. Applications have also been characterized under various suboptimal network conditions to better understand what impact network problems have on user codes.
The USSR Cosmos Biosatellites are unmanned missions with durations of approximately 14 days. They are capable of carrying a wide variety of biological specimens such as cells, tissues, plants, and animals, including rodents and rhesus monkeys. The absence of a crew is an advantage with respect to the use of radioisotopes or other toxic materials and contaminants, but a disadvantage with respect to the performance of inflight procedures or repair of hardware failures. Thus, experiments hardware and procedures must be either completely automated or remotely controlled from the ground. A serious limiting factor for experiments is the amount of electrical powers available, so when possible experiments should be self-contained with their own batteries and data recording devices. Late loading is restricted to approximately 48 hours before launch and access time upon recovery is not precise since there is a ballistic reentry and the capsule must first be located and recovery vehicles dispatched to the site. Launches are quite reliable and there is a proven track record of nine previous Biosatellite flights. This paper will present data and experience from the seven previous Cosmos flights in which the US has participated as well as the key areas of consideration in planning a flight investigation aboard this Biosatellite platform.
The description and specifications for a digital avionics design and reliability analyzer are given. Its basic function is to provide for the simulation and emulation of the various fault-tolerant digital avionic computer designs that are developed. It has been established that hardware emulation at the gate-level will be utilized. The primary benefit of emulation to reliability analysis is the fact that it provides the capability to model a system at a very detailed level. Emulation allows the direct insertion of faults into the system, rather than waiting for actual hardware failures to occur. This allows for controlled and accelerated testing of system reaction to hardware failures. There is a trade study which leads to the decision to specify a two-machine system, including an emulation computer connected to a general-purpose computer. There is also an evaluation of potential computers to serve as the emulation computer.
Program detects hardware failure in servosystems by comparing actual servo valve position with predictions of software model. In addition, system will also pick up most computer input/output failures. Process presents faster and more reliable results than previous failure detection methods.
Mission critical systems typically employ multi-string redundancy to cope with possible hardware failure. Such systems are only as fault tolerant as there are many redundant strings. Once a particular critical component exhausts its redundant spares, the multi-string architecture cannot tolerate any further hardware failure. This paper aims at addressing such catastrophic faults through the use of 'Self-Reconfigurable Chips' as a last resort effort to 'repair' a faulty critical component.