Designing Decentralized Controllers for Distributed-Air-Jet MEMS-Based Micromanipulators by Reinforcement Learning

Distributed-air-jet MEMS-based systems have been proposed to manipulate small parts with high velocities and without any friction problems. The control of such distributed systems is very challenging and usual approaches for contact arrayed system don’t produce satisfactory results. In this paper, we investigate reinforcement learning control approaches in order to position and convey an object. Reinforcement learning is a popular approach to find controllers that are tailored exactly to the system without any prior model. We show how to apply reinforcement learning in a decentralized perspective and in order to address the global-local trade-off. The simulation results demonstrate that the reinforcement learning method is a promising way to design control laws for such distributed systems.     

Sintering of a quasi-crystalline powder using spark plasma sintering and hot-pressing

Sintering experiments on a granulated AlCuFeB quasi-crystalline powder were performed using hot-pressing (HP) and spark plasma sintering (SPS). By coupling the analysis of a sintering law (derived from creep rate equations) and comparative observations of almost fully dense sintered samples using mainly microprobe analysis and transmission electron microscopy, hypotheses about the densification mechanism(s) involved were expressed. Whatever sintering is performed by SPS or HP, it is proposed that two regimes exist, the difference between them lying in the fact that grain boundaries are/are not perfect sources/sinks of vacancies.     

An experimental and modeling study of 2-methyl-1-butanol oxidation in a jet-stirred reactor

In an effort to understand the oxidation chemistry of new generation biofuels, oxidation of a pentanol isomer (2-methyl-1-butanol) was investigated experimentally in a jet-stirred reactor (JSR) at a pressure of 10 atm, equivalence ratios of 0.5, 1, 2 and 4 and in a temperature range of 700–1200 K. Concentration profiles of the stable species were measured using GC and FTIR. A detailed chemical kinetic mechanism including oxidation of various hydrocarbon and oxygenated fuels was extended to include the oxidation chemistry of 2-methyl-1-butanol, the resulting mechanism was used to simulate the present experiments. In addition to the present data, recent experimental data such as ignition delay times measured in a shock tube and laminar flame speeds were also simulated with this mechanism and satisfactory results were obtained. Reaction path and sensitivity analyses were performed in order to interpret the results.     

Special Issue on Content Based Multimedia Indexing

Multimedia Tools and Applications -     

Modelling decomposition and fire behaviour of small samples of a glass‐fibre‐reinforced polyester/balsa‐cored sandwich material

This publication presents the experimental and numerical methods to model the devolatilization process of a glass‐fibre‐reinforced polyester/balsa‐cored sandwich material on small scale. The fundamental modelling of the source term in pyrolysis‐based fire simulations requires as input data the thermochemical properties of solid fuel and the kinetic parameters of the devolatilization process. First, the thermal decomposition of both elements composing the sandwich structure was studied by thermogravimetry coupled with gas analysis, in air and pure nitrogen atmospheres at several heating rates, in order to define a comprehensive multi‐step reaction pathway. A differential equation system is defined to model these decomposition processes. The kinetic parameters were then estimated by solving the system of equations by an inverse problem. Second, the fire behaviour of each element was studied separately and then combined in the sandwich structure on the cone calorimeter. In addition, numerical simulations with Fire Dynamics Simulator were performed to gradually assess the ability of the model(s) to reproduce each element composing the sandwich structure. Numerical and experimental results are compared and then discussed. Overall, the model provides a good agreement with the experimental data and encourages to model higher scales. Copyright © 2012 John Wiley & Sons, Ltd.     

C35 Hopanoid Side Chain Biosynthesis: Reduction of Ribosylhopane into Bacteriohopanetetrol by a Cell‐Free System Derived from Methylobacterium organophilum

The major bacterial triterpenoids of the hopane series each consist of a C₃₀ triterpene hopane moiety and an additional nonterpene C₅ side chain derived from D ‐ribose and linked through its C‐5 carbon atom to the hopane side chain. Bacteriohopanetetrol and aminobacteriohopanetriol are the most common representatives of this natural product series, adenosylhopane and ribosylhopane being putative precursors. Deuterium‐labelled ribosylhopane was obtained by hemisynthesis and converted into deuterium‐labelled bacteriohopanetetrol in the presence of NADPH, thus giving evidence of this as yet unknown precursor‐to‐product relationship in the bacterial hopanoid metabolic pathway.     

Calibration of flow rate in cone calorimeter tests

The cone calorimeter is one of the major fire tests. The measurement method is based on the evaluation of mass flow and oxygen concentration of fire effluents to calculate heat release rate. Different studies highlighted that the governing parameter for uncertainty at important values of heat release rate was the characteristic constant of an orifice plate used to measure mass flow (C‐factor). This parameter is usually determined each testing day by oxygen consumption calorimetry with a reference methane burner. This publication presents a calibration method for volumetric flow rate without using orifice plate and then an extension to C‐factor determination without methane burner and calorimetry. The uncertainty calculation applied to a real example highlights the fact that the method is suitable to respect tolerance of standardized test conditions for both parameters. Copyright © 2013 John Wiley & Sons, Ltd.