Narrow Results

The emergence of Grid infrastructures like EGEE has enabled the deployment of large-scale computational experiments that address challenging scientific problems in various fields. However, to realize their full potential, Grid infrastructures need to achieve a higher degree of dependability, i.e., they need to improve the ratio of Grid-job requests that complete successfully in the presence of Grid-component failures. To achieve this, however, we need to determine, analyze and classify the causes of job failures on Grids. In this paper we study the reasons behind Grid job failures in the context of EGEE, the largest Grid infrastructure currently in operation. We present points of failure in a Grid that affect the execution of jobs, and describe error types and contributing factors. We discuss various information sources that provide users and administrators with indications about failures, and assess their usefulness based on error information accuracy and completeness. We describe two real-life case studies, describing failures that occurred on a production site of EGEE and the troubleshooting process for each case. Finally, we propose the architecture for a system that could provide failure management support to administrators and end-users of large-scale Grid infrastructures like EGEE.

The use of BGA (Ball Grid Array) interconnects utilizing the lead-free solder joint has grown rapidly because of its small volume and diversity of application. Thus, it requires the continuous quantification and refinement of lead-free solder joint reliability. The lead-free solder creep and cyclically applied mechanical loads cause metal fatigue on the lead-free solder joint which inevitably leads to an electrical discontinuity. In the field application, BGA solder joints experience mechanical loads during temperature changes caused by power up/down events as the result of the CTE (Coefficient of Thermal Expansion) mismatch between the substrate and the Si die. In this paper, extremely small resistance changes at joint area corresponding to through-cracks generated by thermal fatigue were measured. In this way, the failure was defined in terms of anomalous changes in electrical resistance of the joint. Furthermore the reliability of BGA solder joints in thermal cycling is evaluated by using the modified coffin-Manson criterion which may define and distinguish failure. Any change in circuit resistance according to the accumulated damage induced by the thermal cycling in the joint was recorded and evaluated in order to quantitate reliability of solder joint.

Contrary to what often seems to be accepted, 100% Tit-for-Tat is not an ESS, even in the three-pure-strategy game Tit-for-Tat-Always Cooperate-Always Defect, for which 100% Always Defect finally remains the only attractor for the composition of “asexual” populations. The present paper firstly investigates the dynamics of the Tit-for-Tat-Always Cooperate-Always Defect game for asexual populations, with the more realistic assumption that Tit-for-Tat sometimes fails to apply its algorithm. Surprisingly, this perturbation of the original game leads to different attractor patterns, in relation to the numerical values of the expected number of meetings and of Tit-for-Tat’s failure rate. These patterns generally include one punctual attractor for which most dyads at least partially cooperate. Nevertheless, the attractor alternative to 100% Always Defect in each pattern may become unstable when certain mixed strategies (cooperating and defecting at random) are introduced into the game, and the whole flow of strategy frequencies may then converge towards 100% Always Defect. Beyond the reiterated Prisoner’s Dilemma, the present analysis illustrates how self-organizing processes at infra-populational levels, like that occurring in dyads including at least one fallible Tit-for-Tat, can influence the evolution of populations.

The purpose of this research is to understand: (1) the main themes that appear to contribute to entrepreneurial success, (2) the various combinations of antecedents that can lead to entrepreneurial success, and; (3) the role that travel plays in entrepreneurial success. We first use a qualitative methodology to assess the themes that emerge in our conversations with 14 highly-successful Canadian entrepreneurs. The main categories that emerged from our interviews that contribute to entrepreneurial success involve: learning, travel, adversity quotient, and mentorship. From these results, we conduct a qualitative comparative analysis (QCA) and find that the input variables that were most important to entrepreneurial success were: learning, experiencing failure, learning from mentors, and adversity quotient. The contributions to knowledge of this research are twofold. First, we show that travel is an important construct to entrepreneurial success, which is significant as travel has largely been omitted from the entrepreneurship literature. Second, we show that entrepreneurial success is dependent on a complex combination of variables of varying levels of importance.

This paper presents facts and figures, directly or indirectly, related to medical equipment reliability and reviews various important aspects, directly or indirectly, concerned with medical equipment reliability including classifications of medical devices/equipment, human error in medical equipment, useful guidelines for reliability and other professionals to improve medical equipment reliability, and medical equipment maintenance. A number of methods considered useful for performing medical equipment reliability analysis are also presented. Useful sources and organizations for obtaining medical equipment failure-related data are listed.

We evaluate the impact of strategic orientation on the failure probability of financial institutions. Using the US credit union industry as the empirical setting, we find that credit unions which exhibit preferential treatment to borrowers are more likely to fail, whereas those who set operational strategies towards balancing the benefits between savers and borrowers experience a lower failure probability. The impacts appear to be more pronounced in small credit unions and in credit unions which have a lower operating experience. We also find that borrower-oriented credit unions generate lower interest margins while neutral behavior credit unions generate higher margins.

Software-Defined Networking disassociates the control plane from data plane. The problem of deciding upon the number and locations of controllers and assigning switches to them has attracted the attention of researchers. Foreseeing the possibility of failure of a controller, a backup controller has to be maintained for each switch so that the switches assigned to the failed controller can immediately be connected to their backup controllers. Hence, the switches cannot experience disconnections in case of failure of their controller. In this paper, two mathematical models are proposed. The first model focuses on minimizing the average of latencies from all switches to their backup controllers while considering the failure of the controllers. The second model aims at minimizing both the average and worst-case of latencies from all switches to the corresponding backup controllers. Both of our models are evaluated on three networks and are compared (in terms of two metrics, viz., average and worst-case latencies) with an existing model that focuses on minimizing only worst-case latency. The first model gives better average latency compared to the reference model. The second model also gives better average latency and almost equal worst-case latency compared to the reference model.

This paper presents a simplified approximate analysis of the overall collapse of the towers of World Trade Center in New York on September 11, 2001. The analysis shows that if prolonged heating caused the majority of columns of a single floor to lose their load carrying capacity, the whole tower was doomed. Despite optimistic simplifying assumptions, the structural resistance is found to be an order of magnitude less than necessary for survival.

Collapse of transmission towers due to downbursts is often initiated by local failure of key structural members, while the local failure of key structural members is related to local material and geometrical nonlinearities. This paper presents a multi-scale finite element (FE) model for the failure analysis of transmission towers under downburst-induced wind loading. The potential local failure areas of the tower are modeled by shell or solid elements, and the remaining parts by beam elements. In this way, the failure of the tower can be accurately simulated on the one hand and the computational effort can be reduced on the other hand. This paper first introduces how to determine the downburst-induced wind loading on transmission towers. Both the conventional beam and multi-scale FE models of the transmission tower are then developed and used in the failure analysis. A comparison of the failure results obtained by the two FE models show that the multi-scale FE model can effectively simulate the stress concentration of angle members around the bolt connections and the cross-section plastic collapse of key structural members, leading to a different failure pattern for the tower from the conventional FE method. It is suggested that the multi-scale FE model should be used for better accuracy in the failure analysis of transmission towers under downburst loading.

We present an extended radial point interpolation method (XRPIM) for modeling cracks and material interfaces in two-dimensional elasto-static problems. Therefore, partition of unity enrichment is incorporated into RPIM. We employ both step enrichment and crack tip enrichment for cracks. The studies are restricted to stationary cracks though the method can be extended easily to moving boundaries. We compare the results to the extended finite element method to show the superiority of our method. We show for two selected problems that the error is of magnitudes lower compared to XFEM simulations.

We present a Smoothed Finite Element Methods (SFEM) for thermo-mechanical impact problems. The smoothing is applied to the strains and the standard finite element approach is used for the temperature field. The SFEM allows for highly accurate results and large deformations. No isoparametric mapping is needed; the shape functions are computed in the physical domain. Moreover, no derivatives of the shape functions must be computed. We implemented a visco-plastic constitutive model and validate the method by comparing numerical results to experimental data.

This paper presents a multi-scale model that can predict the ballistic impact behavior of multi-layer plain-woven fabrics using the finite element method (FEM). Multi-layer fabrics of 30.5 × 30.5 cm, woven by high performance yarns Kevlar^® 29 3000 denier, are impacted by a 0.3 fragment simulating projectile (FSP). Using a multi-scale approach, behavior of multi-layer fabrics subjected to different impact velocities is numerically analyzed. Ballistic limit of the fabric can also be predicted. The multi-scale model shows an effective gain of computation time in comparison with current mesoscopic ones. Computational results show a good agreement with experimental data.

We present impact simulations with the Smoothed Finite Element Method (SFEM). Therefore, we develop the SFEM in the context of explicit dynamic applications based on diagonalized mass matrix. Since SFEM is not based on the isoparametric concept and is based on line integration rather than domain integration, it is very promising for events involving large deformations and severe element distortion as they occur in high dynamic events such as impacts. For some benchmark problems, we show that SFEM is superior to standard FEM for impact events. To our best knowledge, this is the first time SFEM is applied in the context of impact analysis based on explicit time integration.

Failure among SMEs has been attributed to size, age, location and being part of particular industries and not part of a network. These determinants, if used as surrogates for resources, suggest firms that lack such resources are more susceptible to failure. Using the resource-based view of the firm, this paper aims to do a post-mortem examination on the attributes of failed SMEs by analyzing data of more than 13,000 failed firms in the UK between 1999 and 2009. The findings reveal that failure is not merely a function of lack of resources. Although these findings shed new light on attributes of failed SMEs, there are still some limitations to the work that must be taken into consideration when applying the results.

This paper, using a flexible and more robust analytical tool, neural networks models, analyzes the most important factors in determining business failure among small, high-technology firms. Using the resource-based view and organizational ecology as the theoretical lenses through which to view the problem, the paper modeled a number of variables that stood as proxies for resources. The results suggest profit, in the form of retained earnings, is the most significant factor that determines failure among these firms. Other factors of importance are governance structure, location and firm size. Importantly, the issue of governance structure as a resource has received very little attention in existing works on business failure, and as such, this paper makes a contribution to the literature by adding governance as an important resource.

New entrepreneurs often face failures that can erode confidence and self-efficacy, thereby obstructing progress. This study considers the effects of failure on entrepreneurial self-efficacy and proposes a model based on entrepreneurial learning of how passion and resilience may mitigate these effects. Using data from 423 entrepreneurs (both successful and unsuccessful) in North America, it tests a model via structural equation modeling, in which entrepreneurial self-efficacy is directly affected by failure, and indirectly affected by passion and resilience. The results indicate the negative direct effects of failure on entrepreneurial self-efficacy may be offset by strongly positive effects of entrepreneurial passion and by resilience. This appears to be the first empirical study to test directly the moderating effects of entrepreneurial passion and resilience on the relationship between failure and entrepreneurial self-efficacy. In the presence of sufficient passion and resilience, failure may be viewed as a positive influence on self-efficacy. The results suggest that entrepreneurial failure may act as a precursor to entrepreneurial passion. They also suggest that the practical, negative effects of setbacks can be mitigated, or even reversed, by focusing on developing entrepreneurial passion and resilience in new entrepreneurs.

While innovation is an attractive path, it is also a rocky path made up of numerous challenges, even failures. This study provides new knowledge for understanding innovation failure. It seeks answers to the question: What are the perceived factors of innovation failure in SMEs? Every individual who has experienced an innovation failure has a story to tell. Therefore, the research question of this study is answered based on these stories. The main data are collected through narratives produced by individuals who have been involved in the development of completely failed innovation initiatives. In addition, four expert interviews are conducted. The results demonstrate that the most common factor for innovation failure is the occurrence of several incidents during the innovation process that slowly contribute to complete failure. In addition, the results reveal three SME-typical narratives of failed innovations as narrators the Passionate Innovator, the Solo Innovator, and the Developer Innovator.

In an attempt to increase resource efficiency and reduce carbon emissions, the development of lightweight designs in structural applications is essential. In addition, the lightweight structures often follow complex topologically optimized designs which are more suitable for the application of additive, in contrast to conventional manufacturing techniques. Within the additive manufacturing (AM) process, constituents may be combined to design and produce durable and lightweight materials with predefined mechanical, electrical, and thermal properties, while also accounting for their sustainability and recyclability. Regretfully, due to the lack of research in material behavior, the AM technology implementation in engineering applications is still limited in comparison to traditional manufacturing methods. While the potential of additively manufactured continuous fiber composites has already been recognized in the scientific community, constitutive modeling and damage resistance are seldom reported. Since fiber-reinforced composite structures are rarely designed as unidirectional (UD), this study is focused on numerical analysis of failure for multi-directionally reinforced composite laminates loaded in a uniaxial direction. Specimens are modeled and evaluated using a progressive damage model, proposing guidelines for safer design and application.

Structural hierarchies are universal design paradigms of biological materials, e.g., several materials in nature used for carrying mechanical load or impact protection such as bone, nacre, dentin show structural design at multiple length scales from the nanoscale to the macroscale. Another example is the case of diatoms, microscopic mineralized algae with intricately patterned silica-based exoskeletons, with substructure from the nanometer to micrometer length scale. Previous studies on silica nano-honeycomb structures inspired from these diatom substructures at the nanoscale have shown a great improvement in plasticity, ductility and toughness through these designs over macroscopic silica, though along with a substantial reduction in stiffness. Here, we extend the study of these structural designs to the micron length scale by introducing additional hierarchy levels to implement a multilevel composite design. To facilitate our computational experiments we first develop a mesoscale particle-spring model description of the mechanics of bulk silica/nano-honeycomb silica composites. Our mesoscale description is directly derived from constitutive material behavior found through atomistic simulations at the nanoscale with the first principles-based ReaxFF force field, but is capable of describing deformation and failure of silica materials at tens of micrometer length scales. We create several models of randomly-dispersed fiber-composite materials with a small volume fraction of the nano-honeycomb phase, and analyze the fracture mechanics using J-integral and R-curve studies. Our simulations show a dominance of quasi-brittle fracture behavior in all cases considered. For particular materials with a small volume fraction of the nano-honeycomb phase dispersed as fibers within a bulk silica matrix, we find a large improvement (≈4.4 times) in toughness over bulk silica, while retaining the high stiffness (to 70%) of the material. The increase in toughness is observed to arise primarily from crack path deflection and crack bridging by the nano-honeycomb fibers. The first structural hierarchy at the nanometer scale (nano-honeycomb silica) provides large improvements in ductility and toughness at the cost of a large reduction in stiffness. The second structural hierarchy at the micron length scale (bulk silica/nano-honeycomb composite) recovers the stiffness of bulk silica while substantially improving its toughness. The results reported here provide direct evidence that structural hierarchies present a powerful design paradigm to obtain heightened levels of stiffness and toughness from multiscale engineering a single brittle — and by itself a functionally inferior material — without the need to introduce organic (e.g., protein) phases. Our model sets the stage for the direct simulation of multiple hierarchical levels to describe deformation and failure of complex biological composites.

The quasi-static and dynamic compression responses and failure of fiber-reinforced syntactic foams were investigated. The role of fiber volume fraction on the compression response of syntactic foams was examined in terms of mechanical behavior and energy absorption under both quasi-static and dynamic conditions. Results showed that the mechanical behavior and energy absorption of the reinforced specimens increased with increasing fiber volume fraction. The syntactic foams exhibited distinct strain rate sensitivity; and their yield strength and elastic modulus increased by 41.1% and 85.1%, respectively, as strain rate increased from 0.0011 s${}^{\mathrm{\text{-1}}}$ to 1070 s${}^{\mathrm{\text{-1}}}$. The deformation and failure processes of the syntactic foams were also examined, and the underlying mechanisms were discussed.