Stable operation of induction-type bearingless motors under loadedconditions

A control method, which realizes stable operation in induction-type bearingless motors under loaded conditions, is proposed. Without a proposed control method, both the amplitude and the direction of radial force vary in accordance with loads. These variations result in interferences of radial force commands in two perpendicular axes. The interferences become a serious problem as a load increases. A shaft finally touches down without proper compensation. It is found that a decoupling of the radial force commands can be realized with compensation of both amplitude and phase angle of four-pole motor current. A decoupling controller is proposed with the principle of a field-oriented controller of conventional induction motors. The results of experiments show the validity of compensations. It is shown that induction-type bearingless motors can be operated stably under loaded conditions, as well as transient conditions     

Java heap protection for debugging native methods

Java virtual machine (JVM) crashes are often due to an invalid memory reference to the JVM heap. Before the bug that caused the invalid reference can be fixed, its location must be identified. It can be in either the JVM implementation or the native library written in C invoked from Java applications. To help system engineers identify the location, we implemented a feature using page protection that prevents threads executing native methods from referring to the JVM heap. This feature protects the JVM heap during native method execution; if the heap is referred to invalidly, it interrupts the execution by generating a page-fault exception. It then reports the location where the exception was generated. The runtime overhead for using this feature depends on the frequency of native method calls because the protection is switched on each time a native method is called. We evaluated the runtime overhead by running the SPECjvm98, SPECjbb2000, VolanoMark, and JFCMark benchmark suites on a PC with two Intel Xeon^® 1.6 GHz processors. The performance loss was less than 2% for the benchmark items that do not call native methods so frequently (10⁴ times per second) and 5%–20% for the benchmark items that do (10⁴–10⁵ times per second). The worst performance loss was 54%, which was recorded for a benchmark item that calls native methods 2.0×10⁶ times per second.     

DomainProfiler: toward accurate and early discovery of domain names abused in future

Domain names are at the base of today’s cyber-attacks. Attackers abuse the domain name system (DNS) to mystify their attack ecosystems; they systematically generate a huge volume of distinct domain names to make it infeasible for blacklisting approaches to keep up with newly generated malicious domain names. To solve this problem, we propose DomainProfiler for discovering malicious domain names that are likely to be abused in future. The key idea with our system is to exploit temporal variation patterns (TVPs) of domain names. The TVPs of domain names include information about how and when a domain name has been listed in legitimate/popular and/or malicious domain name lists. On the basis of this idea, our system actively collects historical DNS logs, analyzes their TVPs, and predicts whether a given domain name will be used for malicious purposes. Our evaluation revealed that DomainProfiler can predict malicious domain names 220 days beforehand with a true positive rate of 0.985. Moreover, we verified the effectiveness of our system in terms of the benefits from our TVPs and defense against cyber-attacks.     

Markov-type velocity field for efficiently animating water stream

In computer graphics, one of the most challenging tasks is continuously varying phenomena such as waving, swaying, and flowing motions. In this paper, we present a novel hybrid model (physical-stochastic) to create an endless animation in which offline simulation is used to produce an infinitely varying real-time animated result. In this particular case, a water stream model is proposed. Most fully 3D physically based simulation methods for depicting fluid flows are very time and memory consuming. Thus, these methods are still reserved for offline simulations and small-domain real-time simulations, especially in the case of fluid flows with irregularly repeating patterns. The proposed model is based on the tracer particle technique, uses a non-static velocity field, and consists of two main phases. In the first phase, we construct the stochastic velocity field by using the physically based method. The second phase is the main part, in which we create real-time endless animation. Here, we introduce a new type of velocity field which we refer to as a Markov-type velocity field (MTVF). MTVF allows us to animate a water stream endlessly in real-time by avoiding the time-consuming process of solving the corresponding equations for every simulation step.     

Graft polymerization of 1,1-dihydroperfluorobutyl acrylate onto cellulose fabric

1,1-Dihydroperfluorobutyl acrylate was prepared and grafted onto cellulose fabric by means of γ-ray preirradiation to produce oil and water repellency. A 2% grafting was sufficient to impart oil and water repellency to cellulose fabric. In the case of 19.1% or 29.4% grafted fabrics, the grafting decreased rapidly with laundering, and the laundered fabric which possessed no oil and water repellency still had about 10% grafting of the perfluoro compound. Thus, the apparent graft polymerization takes place both on the surface and in the inner structure of the cellulose fiber; the polymer grafted onto the surface imparts oil and water repellency and is easily removed by laundering, whereas the polymer grafted onto the inner structure is hardly eliminated but does not impart oil and water repellency to the fabric.     

Parameter estimation for a solids mixing|segregation model for gas fluidised beds

Gibilaro and Rowe in their model of a gas fluidised bed containing two particulate species introduced four arbitrary parameters linked to the known mechanisms giving rise to particle movement in such beds. In this work, these parameters are made deterministic by linking them to the physics of the bubbling bed and the particle properties. Segregation patterns are predicted from these parameters iteratively by allowing for the accompany variation with bed height of minimum fluidisation velocity and local bed behaviour. Good agreement between predicted and experimental patterns is found. So too is the dependence of mixing index on gas velocity. Agreement breaks down for high jetsam concentration systems. The model is not directly applicable when a defluidised region forms at the bottom of the bed.     

Development of Reaction-Bonded Electroconductive Silicon Nitride-Titanium Nitride and Resistive Silicon Nitride-Aluminum Oxide Composites

Reaction-bonded Si₃N₄· TiN and Si₃N₄· Al₂O₃ composites were successfully fabricated by heating mixed powder compacts of Si and TiN or Si and Al₂O₃ in a nitrogen atmosphere. The former showed electrical conductivity, owing to the presence of TiN. An electrical resistivity of 2.6 × 10⁻⁵Ω· m was obtained for the Si₃N₄· TiN composite with 70 vol% TiN. The composite with 20 vol% TiN showed an electrical resistivity of 0.22 Ω· m and a bending strength of 460 MPa. On the other hand, the Si₃N₄· Al₂O₃ composite had insulating properties. The use of an appropriate amount of resin binder resulted in a higher green density and, consequently, a higher bending strength. Moreover, electroconductive Si₃N₄· TiN/resistive Si₃N₄· Al₂O₃ complex ceramics could be fabricated by heating green compacts composed of two different portions, one composed of mixed powders of Si and TiN and the other of Si and Al₂O₃. Attainment of such complex ceramics was attributed to the small dimensional change at the nitriding stage, under 0.3% and the similarity of the thermal expansion coefficients of the two composites.     

Synthesis of carbon nanotubes on carbon fibers by means of two-step thermochemical vapor deposition

The present study aimed at development of a method for synthesizing multi-walled carbon nanotubes (CNTs) on carbon paper substrates (CP) at densities as high as those so far reported for CNTs formed on quartz substrates. Applying conditions optimized for CNTs synthesis on quartz substrates, in which CP was heated at 1073 K, being placed parallel to the flow of m-xylene/ferrocene vapor, resulted in formation of extremely few deposits on CP. Forced vapor flow through the CP greatly improved the frequency and homogeneity of deposition of the Fe-bearing nanoparticles, but these became encapsulated by carbon and deactivated. The addition of H₂S to the vapor further enhanced nanoparticle deposition. Moreover, it enabled the subsequent formation of CNTs at densities as high as 2-6 × 10⁹ cm⁻². In order to realize such high population densities, it was found essential to perform CVD in a two-stage sequence commencing with nanoparticles deposition at 1073 K followed by the formation and growth of CNTs at 1273 K, with the H₂S concentration in the vapor phase optimized throughout within a range of 0.014-0.034 vol%.