Advanced Materials Research Vols. 118-120

Abstract: Two-stage loading tests corresponding to high-low loading and low-high loading were performed to clarify the cumulative damage properties of two kinds of aluminum alloy welded joints such as the corner joint and butt joint. As a result, the logarithmic value of cumulative damage can better be characterized by the normal distribution in accordance with the Komogorov-Smirnov goodness of fit test. The proposed linear cumulative damage model can better reflect the scatter characteristic of test data, as well as the newly developed nonlinear cumulative damage model can better characterize the effects of load sequence and load level on fatigue damage value, which was verified with the obtained test data.

Abstract: Four kinds of aluminum alloy welded joints widely used in car-body for high-speed train were experimentally examined to clarify their probabilistic fatigue properties. Based on BS8118 and IIW standard criterions, the probabilistic fatigue limits for these four welded joints, corresponding to loading cycle of 2×106, are evaluated under two different loading modes. In terms of the extrapolation proposed by Eurocode 9 criterion, the corresponding probabilistic S-N curves in very high cycle regime are obtained, combined with grouping method and lifting and lowering method.

Abstract: The traditional loading-strength interference model is used to calculate the reliability of components and system when random loading act once. In fact, components always work under repeated random loading, so it is important to derive a reliability model considering the frequency of loading. The two-dimensional distribution of random loading is proposed in this paper. In engineering practice, only a few samples of time-loading process can be obtained because of all kinds of limitation, so the reliability model based on transverse distribution of random loading. In addition, when it is difficult to know the exact distribution of random loading and strength, a discrete reliablity model is developed through the method of universal generating function. Finally, the reliability of system is analyzed considering common cause failure.

Abstract: As everyone knows, reliability growth technology is an essential part in the mechanical reliability theory as well as an insurance of the products capability in usage. It exists throughout the entire lifespan of development, manufacturing and application. Concerning the reliability characters of mechanical products, that product life obeys Weibull distribution, which is mostly resulted from the test on the small sample, three parameters of life distribution are estimated by the grey estimation in this paper. Then according to the data acquired in the test, Duane growth model is surely developed to assess the situation of reliability growth. Furthermore, the following example ascertains that the developed model is in accordance with mechanical characters. From the result, Duane model is reasonable to evaluate the reliability growth level of mechanical products. It is obvious that the improved measure is effective to enhance the reliability and the value of MTBF can be calculated with the model.

Abstract: This paper explores the grey model based PSO (particle swarm optimization) algorithm for anti-cauterization reliability design of underground pipelines. First, depending on underground pipelines’ corrosion status, failure modes such as leakage and breakage are studied. Then, a grey GM(1,1) model based PSO algorithm is employed to the reliability design of the pipelines. One important advantage of the proposed algorithm is that only fewer data is used for reliability design. Finally, applications are used to illustrate the effectiveness and efficiency of the proposed approach.

Abstract: Kinematic accuracy is an important index to evaluate the quality of mechanisms. Kinematic reliability is a significant standard to examine dynamic kinematic accuracy of mechanisms. Having deduced inverse kinematics and calculated model for position errors of Delta mechanism, kinematic reliability of the mechanism is calculated considering the randomness of every error components. Analytic method sensitivity analysis is adopted to describe contribution of the randomness of each error component to kinematic reliability of the mechanism, and the calculating approach is exactly given that the sensitivity analysis which is for getting the error contributions to reliability of which the position errors during movement is in the range of allowable accuracy at any time. Kinematic reliability and its sensitivity analysis can search key factor influencing kinematic accuracy, which can provide the foundation for error compensate and accuracy assurance technology of the mechanism.

Abstract: Based on reliability physics theory, investigation on failure mechanism is a kernel job in reliability design and analysis of mechanical products. The basic conception and classification of failure mechanism are introduced. Researches on five kinds of fracture mechanisms (ductile fracture, cleavage fracture, quasi-cleavage fracture, fatigue fracture and intergranular fracture) are discussed in detail. At last, problems are summarized and suggestions for future investigations are also made.

Abstract: In order to promote the performance measurement technique of the diffusion vacuum pumps, one full-automatic test equipment was newly developed and patented. The control system is the core of the test equipment and its reliability analysis is necessary. The structure and the characteristics of the test equipment are introduced in paper. The fault tree of the test equipment is set up and illustrated in detail. Based on fault tree analysis, the reliability designs of the hardware and software of the test equipment are carried out.

Abstract: This paper proposes a new reliability modeling and analysis approach based on Stochastic Petri Nets by employing the logical relations in the RBD to cope with the inadequacy in various quantitative reliability analysis techniques, such as Reliability Block Diagram (RBD) and Markov analysis of control system in process industry. In this paper, the proposed new approach is elaborated and its feasibility and effectiveness is illustrated in a study case. As the results show, the proposed approach is demonstrated more straightforward and visual than Markov method. This research also bears significance in its application in the reliability analysis of general system in process industry.

Abstract: Given the existing difficulties in conventional reliability models and the limitations of the current SPN software tools in terms of modeling system reliability, a software tool for modeling system reliability based on SPN named RelSPN@zer is developed, describing both the general structure and the underlying numerical methods of the tool. RelSPN@zer provides a unified framework for the modeling and evaluation of SPN running under MATLAB environment and is especially tailed to the system reliability analysis. Many metrics of system reliability can be obtained both under stationary and transient state. An example is given to illustrate the use of this package.