Computing steady-state mean time to failure for non-coherent repairable systems

Mean time to failure (MTTF) is an important reliability measure. Previous research is mainly concerned with the MTTF computation of coherent systems. In this paper, we derive equations to calculate the steady-state MTTF for noncoherent systems. Based on the equations, we extend the BDD by adding an intersection edge in each BDD node to efficiently store additional information for MTTF computation of noncoherent systems. A recursive algorithm is developed for MTTF computation using the extended BDD. To accelerate building the extended BDD, a method is proposed to avoid calculating the intersection edge for some nodes by keeping node monotonicity during the BDD construction. We show the efficiency of our algorithm by applying it to some example fault trees, real-life applications, and large fault tree benchmarks.     

The development of solidification microstructures in the presence of lateral constraints

The effect of a sudden change in cross section on the microstructural development has been investigated by the directional solidification technique. Experiments have been carried out in a transparent model system of succinonitrile-acetone so that the dynamical changes in the interface shapes can be monitoredin situ as the interface travels from a region of uniform cross section to a region of sharply reduced cross section. Different experimental conditions which give rise to initial steady-state planar, cellular, and dendritic interfaces have been investigated. Significant changes in microstructures have been observed as the interface approaches a sharply reduced cross section. The planar interface undergoes transitions to cellular and dendritic morphologies as the cross section is reduced, and reverse transitions are observed as the cross section is then increased gradually to its original value. An initial cellular interface is found to become dendritic as the cross section is reduced, and again it becomes cellular as the cross section is increased. When the experimental conditions are designed to give initial dendritic structures, the change in microstructure is found to occur only when the reduced cross section is of the order of primary dendrite spacing. When the reduced cross section is more than about 5 times the primary spacing, no appreciable change is observed in the dendritic array which travels across the cross-sectional change. The dynamical changes in the interface shape and the microstructural transitions that occur with the change in cross section have been examined quantitatively and discussed. Formerly with Ames Laboratory     

Flux enhancement of thin film composite RO membrane by controlled chlorine treatment

A major stumbling block in polyamide thin film composite RO membrane performance is its incompatibility with chlorine and oxidizing agents. The amide bond of the membrane is highly vulnerable to chlorine attack. Two reactions are possible with chlorine exposure i.e. N-H bond chlorination and/or aromatic ring chlorination. In this way, chlorine may cause degradation/modification in the membrane leading to deterioration in performance.However, low concentration of chlorine up to a certain time may give synergistic effect on membrane and improve its performance. Chlorine solution, if exposed to membrane for certain time gives enhancement in trans-membrane flux of the membrane.The same solution if exposed for more time deteriorates ultra-thin polyamide layer of TFC membrane. Conspicuously, the membrane with poor salt rejection and flux benefited more as compared to the membrane with better performance. In the present study, membranes with different salt rejection and flux were taken and exposed to the inorganic chlorine solution. The inorganic chlorine solutions were made by dissolving sodium hypochlorite in pH buffer. The different solutions were made by varying pH to investigate the pH dependence. The membrane samples were kept in solution for different time durations. The exposure time was monitored and the exposure level was taken in terms of ppm h (ppm chlorine solution exposed to membrane for a fixed time in h). With the same chlorine concentration, effect of varying pH was studied. Spiral wound TFC membrane modules were also subjected to chlorine solution to study its effect.     

Lead sorption onto ferrihydrite. 3. Multistage contacting

Few studies have demonstrated the practical application of surface complexation models, calibrated with fundamental macroscopic and spectroscopic metal sorption data, in helping to solve industrial trace metal emissions problems. In this work, multistage ferrihydrite sorption systems are evaluated for their effectiveness in reducing single-solute lead(II) [Pb(II)] concentrations in contaminated water streams to very low levels. Experimental data and modeling results indicate that a multistage sorption system can significantly reduce Pb(II) effluent concentrations for the same total amount of sorbent or, alternatively, dramatically lower total sorbent consumption for the same effluent Pb(II) concentration. Model predictions were generated using a steady-state, multistage, equilibrium adsorber model that was specifically developed for and integrated into 0LI Systems' Environmental Simulation Program. The modified triple-layer model was used to simulate Pb(II) surface-liquid equilibria within the adsorber model. Engineering screening evaluations indicate that a 2-3-stage sorption process can provide significant economic savings when compared to a 1-stage process operating with the same target effluent Pb(II) concentration. Additional equilibrium stages beyond 2 or 3 provide diminishing economic returns. The major economic driver for multiple contacting stages is reduced capital investment and operating costs for sludge handling, dewatering, and disposal.