Dynamically identifying regenerative cycles in Simulation-based Optimization Algorithms for Markov chains

Simulation-based algorithms for maximizing the average reward of a parameterized Markov chain often rely upon the existence of a state which is recurrent for all choices of parameter values. The question of which recurrent state should serve to mark the end of a regenerative cycle is a very important practical consideration in applications. Even when all of the states of the process are recurrent, some states tend to be visited more often than others, and lengthy renewal cycles tend to result in high variance estimates of the gradient. To address this difficulty, we analyze a mechanism for adjusting this special state dynamically (i/sup */-adaptation) as applied to the "batch" simulation-based optimization algorithm of a previous paper. We show that the desirable convergence properties of the original "batch" algorithm are retained with i/sup */-adaptation, namely the almost sure convergence of the parameter vector to a critical point.     

Radiation consequences of seawater contamination during floating storage of reactor-compartment units from salvaged nuclear-powered submarines

The storage time before salvaging of decommissioned nuclear-powered submarines and the floating storage time of reactor units from salvaged submarines can total decades. During this time, radioactive contamination of seawater occurs as a result of the corrosion of the pressure hull in the region of the reactor system and rupturing of the pressure hull - as a result of corrosion of the components of the reactor system. A computational analysis shows that under standard storage conditions the specific activity of water is 10⁵–10⁶ times lower than the intervention level. In the case when the hull and the reactor system are destroyed even with remaining fuel, the specific activity of the sea water near the point of sinking also does not exceed the intervention level because of the slow emission of fission products and transuranium nuclides.Translated from Atomnaya Énergiya, Vol. 97, No. 3, pp. 210–218, September, 2004.     

Mechanisms and Special Features of the Thermodeformational Behavior of a Channel Pipe and Graphite Elements in RBMK Reactors During Heating Due to an Accident

Two- and three-parameter quadratic fitting curves, characterizing the basic thermal and deformational parameters of the state of the structural components of a RBMK technological channel heated as a result of an accident, are constructed on the basis of numerical simulation. The parameters characterizing the accidental heating are assumed to be the heat flux density on the inner surface of the channel pipe, the pressure in the channel, and the thermal resistance on contact surfaces. The output functions are the temperature of the channel pipe at the moment it sets on the graphite, the rate of heating, the average rate of radial expansion of the pipe before and after it sets on the graphite, and the temperature of the surface of the opening in the block at the moment the pipe sets on the graphite. The fitting relations constructed can be used to obtain real-time estimates of the basic thermal and deformational characteristics of the heating process due to the accident and to develop new integral criteria for the integrity of a damaged RBMK channel.     

A global output-feedback controller for stabilization and tracking of underactuated ODIN: A spherical underwater vehicle

This paper presents a method to design an output-feedback controller that simultaneously solves global asymptotic stabilization and tracking of an underactuated omni-directional intelligent navigator—a spherical underwater vehicle moving in a horizontal plane (i.e. at a constant depth). The vehicle does not have a sway actuator and has only position and orientation measurements available. The control development is based on Lyapunov's direct method, backstepping technique and use of interconnected structure of the vehicle dynamics. Numerical simulations demonstrate the results.     

Mg AZ80/SiC composite bars fabricated by infiltration of porous ceramic preforms

Semi-industrial samples of Mg AZ80/SiC/50p composites (machined bars with o.d.=80 mm and length=150 mm, having a maximum 2 vol pct of retained porosity) were successfully fabricated by pressureless, low-pressure (under overpressure of about 0.3 MPa) and moderate-pressure (under overpressure of about 0.8 MPa) infiltration. Pressureless infiltration was completed in 24 hours under the flowing atmosphere of nitrogen, whereas the moderate and low-pressure infiltrations were performed under a static atmosphere of pressurized nitrogen for 7 and 0.5 hours, respectively. The composite samples obtained by all applied infiltration paths posessed a homogeneous microstructure and superior mechanical properties as compared to the nonreinforced matrix. Due to the improved productivity of composite fabrication in comparison with the other two infiltration paths investigated in this work, the moderate-pressure infiltration process was recognized as the most competitive.     

Studies of the influence of alloying elements on the growth of ferrite from austenite under decarburization conditions: Fe-C-Nl alloys

We have evaluated controlled decarburization as a method for probing the effect of alloying elements on ferrite growth from austenite. The technique permits the exploration of longer-time ferrite layer growth; it minimizes the effects of interface structure on ferrite growth; and it permits the isolation of the effects of temperature and alloying element concentration on ferrite/austenite interface motion. The study of the decarburization of initially homogeneous Fe-C-Ni alloys was complemented by experiments using specimens with a controlled nickel concentration gradient. Although the decarburization method yields consistent results at longer times, it is found to be less appropriate for the study of initial ferrite growth. Nucleation in the gas/solid interface region, coupled with uncertainties about the precise time of decarburization, leads to large relative errors at the earliest times. For these reasons, the method is considered a valuable complement to studies based on precipitation boundary conditions. This article is based on a presentation given in the symposium “The Effects of Alloying Elements on the Gamma to Alpha Transformation in Steels,” October 6, 2002, at the TMS Fall Meeting in Columbus, Ohio, under the auspices of the McMaster Centre for Steel Research and the TMS-ASM Phase Transformations Committee.     

Cr-Mo solid solutions forced by high-energy ball milling

Mixtures of Cr and Mo elemental powders, with the nominal compositions Cr₂₅Mo₇₅, Cr₅₀Mo₅₀, and Cr₇₅Mo₂₅, are processed by high-energy ball milling at ambient temperature. Milling is observed to force the mixing of the immiscible bcc elements Cr and Mo into solid solutions. The lattice parameter of these solid solutions, measured by X-ray diffraction (XRD), displays the expected positive deviation from Vegard’s law. These deviations are compared to the ones predicted by Eshelby’s inclusion model for dilute alloys. The conventional Williamson-Hall approach is shown to fail to determine the grain size in as-milled samples, probably due to the high density of dislocations. Annealing at 700 °C for 10 hours under argon leads to a large reduction in structural defect density, without inducing any significant decomposition. The mixing measured in Cr-Mo is discussed in the broader context of the mechanical mixing forced by ball milling in moderately immiscible systems.     

The transient to steady-state transition during the spray-rolling process

From the geometrical standpoint, this article presents a qualitative theoretical analysis and prediction of the transient to steady-state transition during the spray-rolling process, a novel manufacturing technique for aluminum strips. The analytical results indicate that, when the deposited materials at the specific points on one roll surface overlap their counterparts on the other roll surface, spray rolling transits from the transient state to the steady state. The specific points are the limiting deposition positions of the atomized droplets on the roll surface initially.     

Second-order stresses and strains in heterogeneous steels: Self-consistent modeling and X-ray diffraction analysis

Theoretical and experimetal methods have been developed to characterize the effect of mechanical loading on the mesoscopic and macroscopic mechanical state of polycrystalline materials. Ferritic and austenitic single-phase materials were first analyzed, then phase interaction was studied in a multiductile phase material (austeno-ferritic duplex steel) and a natural reinforced composite (pearlitic steel). The theoretical method is based on the self-consistent approach in which elastic and plastic characteristics of the phases have been applied through the micromechanical behavior of single-crystal-using slip systems and microscopic hardening. The effects of a crystallographic texture and phase interaction during loading and after unloading were studied. The elastic and plastic anisotropy of the grains having the same crystallographic orientation were assessed by diffraction strain analysis. The simulation was compared with the experiments performed using the X-ray diffraction technique. In the considered duplex and pearlitic steels, it was observed that the ferrite stress state is much lower than the austenite and cementite ones. The results of diffraction strain distribution have showed the pertinence of the models and give valuable information, for example, for the yield stress and the hardening parameters of each phase in a two-phase material.