Soft Failure Modes Research Articles

Detection and Identification of System Level Soft Failure Induced by Radio Frequency Interference in Small UAV System

A system level methodology is presented for soft failure detection and analysis in small unmanned aerial vehicle (UAV) system, induced by radio frequency interference (RFI). This method includes a radio frequency test system based on the equivalent dynamic test method, and soft failure detection and identification method based on ground control station observation records and system log analysis. By designing signal forwarding system for global positioning system, the equivalent dynamic flight state of the UAV is realized. Through the application of standard radiated susceptibility test methods, the effects of RFI with multiple interference levels covering the frequency range from 2 MHz to 3 GHz on small UAV systems have been investigated. RFI is coupled into the UAV system through the antenna coupling and cable coupling. Soft failure modes are classified and failure analysis is mainly carried out through system log analysis, combined with full-wave simulation analysis, the root cause of the soft failure is determined. The understanding of the soft failure mode and susceptibility is beneficial to the design of UAV system, improving the robustness of software and hardware.

Robust Through-Fin Avalanche in Vertical GaN Fin-JFET With Soft Failure Mode

We study the inherent ruggedness of the avalanche through the fin channel, a new avalanche mode in a vertical GaN power Fin-JFET, through single-pulse and repetitive avalanche circuit tests. By turning on the gate during avalanche, the major avalanche current path migrates from the p-GaN gate to the n-GaN fin channel. The single-pulse critical avalanche energy density ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${E}_{\text {AVA}}$ </tex-math></inline-formula> ) was measured to be 10 J/cm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> , which is the highest reported in GaN transistors. The Fin-JFET withstood over 3700 repetitive avalanche pulses at 70% of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${E}_{\text {AVA}}$ </tex-math></inline-formula> . It exhibits a failure-to-open-circuit signature in single and repetitive avalanche. This soft failure mode allows the device to retain its full breakdown voltage, which is highly desirable for system robustness. By contrast, the through-gate avalanche in Fin-JFETs and the reported avalanche in Si and SiC transistors all show a destructive, failure-to-short-circuit signature. These results show the viability of soft avalanche failure in power devices, provide key robustness references for GaN devices, and suggest the fundamental superiority of moving the avalanche path away from the major blocking junction.

Statistical analysis and optimal inspection planning for lifetime delayed gamma degradation processes

AbstractDegradation measurements of highly reliable products can be used to extrapolate the failure times, and subsequently, predict the remaining useful time‐to‐failure distribution. Different degradation failure mechanisms should be treated specifically to achieve an accurate estimation and prediction. In this study, a typical sequential hard and soft failure mode is investigated while applying the lifetime delayed degradation processes (LDDP) modeling framework. This study focuses on the gamma process, which is combined with location‐scale‐family lifetime distributions and a general reliability inference approach that is applied based on the likelihood. Furthermore, the optimal inspection planning issue with the corresponding process is formulated. The determination of Fisher information (D‐optimality) is used to evaluate the efficiency of the different inspection plans to obtain more information during the degradation process. This study consisted of comprehensive simulations with a more accurate mean time‐to‐failure (MTTF) estimation for the reliability analysis and the inspection starting time that is optimized for planning. A practical crack inspection dataset is also used in this investigation. Herein, the optimal time was simplified while considering the first inspection time and the equivalent inspection interval. It has been proved from the simulation that the inspection number affects the peak value of the Fisher information. This study helps to reduce the experiment sample size and improve the estimation of reliability effectively in engineering.

Truncated Sequential Sampling Plans of Electrical Products Having Two Failure Modes

It is of great significance to design an economic sampling plan for the electrical products in mega engineering projects. The existing traditional plan normally considers single failure mode of the item being inspected. However, the major products in mega engineering projects usually have different failure modes. To overcome the drawbacks of traditional plans, a truncated sequential sampling plan of electrical products having two failure modes is proposed in this paper. The newly proposed sequential sampling plan takes two different failure modes of an electrical product into consideration, consisting of the sequential sampling scheme for soft-failure mode and the sequential sampling scheme for catastrophic-failure mode. The concept of cut-off line is introduced, limiting the sample size within a certain range. At last, a numerical example is provided. This newly proposed sequential sampling plan provides a reference for the design of economic acceptance sampling plans for expensive electrical products and the products with multi-failure modes.

Reliability modeling and evaluation of lifetime delayed degradation process with nondestructive testing

The sequential hard and soft failure mode, a typical failure phenomenon, involves degradation that starts after an initiation period. According to this situation, a random initiation effect is introduced to the normal degradation stage. Intuitively, a lifetime delayed degradation process (LDDP) provides a general framework for this typical complex failure mode. In the present study, general reliability inference approaches involving the joint likelihood function are developed for the LDDP with repeated measurements, according to the expectation maximization (EM) and stochastic EM algorithms, along with numerical simulations and practical application based on real data. Additionally, statistical inferences were obtained using a bootstrap method on the basis of parameter estimations. The proposed method was compared with the traditional two-step strategy under different sample sizes and inspection frequencies and exhibited enhanced performance.

Selective maintenance modeling and analysis of a complex system with dependent failure modes

In this paper, we propose a statistical and optimization framework for selective maintenance of a complex system with components that have multiple failure modes. Specifically, we study the case where each component has multiple instantaneous hard failure modes and a single soft failure mode. An imperfect maintenance action reduces the effective age and degradation level of a component while increasing the mean degradation parameter and the failure rates of hard failures. A numerical example illustrates the application of the proposed framework and the effectiveness of the Genetic Algorithm and Differential Evolution heuristics in solving such maintenance optimization problems.

Identification of Soft Failure Mechanisms Triggered by ESD Stress on a Powered USB 3.0 Interface

The objective of this work is to identify electrostatic discharge (ESD) related soft failure mechanisms early in the product life cycle. We compare different methods of injecting ESD stress into USB 3.0 interfaces which are regularly exposed to ESD stress during end-user operation. A directional injection method is applied which is compatible to high-speed line operation and capable to provide quantitative information about failure levels. Soft failure modes are investigated depending on the operational state of the system under test. Five typical categories of soft failure modes of a USB 3.0 interface could be identified. A method to use the characterization data gained from system verification boards for final system design is presented.

Soft failure mode and effects analysis using the OWG operator and hesitant fuzzy linguistic term sets

Soft failure mode and effects analysis using the OWG operator and hesitant fuzzy linguistic term sets