京都大学 / 新制・課程博士 / 博士(理学) / 甲第24870号 / 理博第4980号 / 新制||理||1711(附属図書館) / 京都大学大学院理学研究科地球惑星科学専攻 / (主査)准教授 伊藤 正一, 教授 野口 高明, 准教授 三宅 亮 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DGAM

Considerable jump and slip phenomena are observed at the non-equilibrium phase interface in microflows. Hence, accurate modelling of the liquid-vapour interface transport mechanisms that matches the observations is required, e.g. in applications such as micro/nanotechnology and micro fuel cells. In the sharp interface model, the classical Navier-Stokes-Fourier (NSF) equations can be used in the liquid and vapour phases, while the interface resistivities describe the jump and slip phenomena at the interface. However, resistivities are challenging to find from the measurements, and most of the classical kinetic theories consider them as constants. One possible approach is to determine them from a model that resolves the phase interface. In order to resolve the interface and the transport processes at and in front of the interface in high resolutions, there are two ways in general, microscopic or macroscopic. The microscopic studies are based either on molecular dynamics (MD) or kinetic models, such as the Enskog-Vlasov (EV) equation. The EV equation modifies the Boltzmann equation by considering dense gas effects, such as the interaction forces between the particles and their finite size. It can be solved by the Direct Simulation Monte Carlo (DSMC) method, which considers sample particles that stand in for thousands to hundred thousands of particles and determine most likely collisions based on interaction probabilities, but it is time-consuming and costly. Here, a closed set of 26-moment equations is numerically solved to resolve the liquid-vapour interface macroscopically while considering the dense gas and phase change effects. The 26-moment set of equations is derived by Struchtrup & Frezzotti as an approximation of the EV equation using Grad's moment method. The macroscopic moment equations resolve the phase interface in a high resolution competitive to the microscopic studies. The resolved interface visualizes the interface structure and the changes of the system variables between the two phases at the interface. The 26-moment equations are solved for a one-dimensional steady-state system for non-equilibrium evaporation/condensation process. Then, solutions are used to find the jump and slip conditions at the interface, which leads to determining the interface resistivities at different interface temperatures and non-equilibrium strengths from the Linear Irreversible Thermodynamics (LIT). The interface resistivities show their dependence on the temperature of the liquid at the interface as well as the strength of the non-equilibrium process. As a result, in further studies, similar systems can be modelled using the sharp interface method with the appropriate jump conditions at the phase interface that can be found from the determined EV interface resistivities. / Graduate

There exist large number of techniques for the preparation of nanostructured materials. Among them the preparation of nanopowders by gas/vapour condensation is a popular one. Because of very high level of surface to volume ratio, powders of metals which may or may not be reactive in the bulk form undergo vigorous oxidation. Oxidation once initiated continues in an auto catalytic fashion leading to a rise in temperature further increasing the oxidation rate. Therefore, the nanopowders are consolidated in situ under high vacuum. Alternatively a thin passivating oxide layer of few nm can be produced by slow exposure to air. Such powders lend themselves to be handled in further processing in ambient atmosphere. The main objective of the present research is to understand the various scientific and technological issues involved in preparing such passivated nanopowders by levitational gas condensation (LGC) technique, a relatively less explored vapour condensation process and their subsequent consolidation by the powder metallurgical route of compaction and sintering. The nanopowders systems studied are Fe and Fe-Cu (4 wt. %Cu). In chapter 1 a brief review of the gas condensation process and the consolidation behavior of nanopowders produced by this method were carried out. Existing knowledge on various topics relevant for the present study like formation of nanoparticles, agglomeration during gas condensation, physical, structural and chemical nature of the passive layer formed during passivation, compaction and sintering behaviour of this passivated nanopowders were discussed. Chapter 2 details the synthesis of Fe nanopowders by levitational gas condensation process and its physical and structural characteristics. The nanopowders in the as synthesized condition showed extremely low packing density due to loosely packed weakly interlocked agglomerates. The nanoparticles manifest as three dimensional reticulated spongy structure composed of chains of these nanoparticles. Heat transfer calculation carried out to determine the particle temperature at different distance from the levitating drop indicates that the nanoparticles can be ferromagnetic at a distance of less than 2 mm away from the levitation drop and hence the magnetic nature of the materials plays an important role in the formation of nanoparticle chains and spongy agglomerates. Passivation of the nanopowders by slow exposure to air produces 3-4 nm thick oxide layer (Fe3O4) over α-Fe and the volume of these oxides was around 45%. The 3rd chapter presents and discusses the results of Fe-Cu alloy nanopowder synthesized by levitating gas condensation process. While synthesis of elemental nanopowders by gas condensation is straight forward as the operating conditions only influence the particle size, alloys require careful control of the levitating drop composition. Although initially we start the process with levitated drop of required composition, the vapour generated will be richer in more volatile element (Cu in our case). Thus the composition of the levitated drop progressively becomes deficient in Cu which in turn reduces Cu in the vapour. Composition of the drop can be stabilised by continuous feeding of the alloy of required composition that can be estimated from the knowledge of equilibrium relation. To establish the equilibrium relationship between composition of the liquid and vapour in evaporation and condensation, phase diagram in the liquid and vapour region was calculated and was validated by determining composition of the drops. Good agreement between the drop composition and the composition predicted by the phase diagram were observed. Various physical, chemical and structural properties of the Fe-Cu nanopowders are characterized in detail using scanning electron microscopy, transmission electron microscopy, X-ray diffraction, X-ray photoelectron spectroscopy (XPS) and thermogravimetry (TG) analysis. The overall chemical composition of the Fe-Cu alloy nanopowders and of the individual agglomerates is same as the composition of feed materials used. However, Cu was found to segregate to the surface of the nanopowders which is attributed to minimization of surface energy with Cu at the surface. The total weight loss observed in TG in flowing hydrogen indicates that the surface passive layer of Fe-Cu appears to be very thin compared to Fe. The consolidation behaviour of both Fe and Fe-Cu nanopowders was studied by both conventional and spark plasma sintering (SPS) and are discussed in chapter 4. The as collected nanopowders from the apparatus have extremely low apparent density. The powders were further subjected to soft milling in a ball mill under ethanol to disentangle the agglomerates there by improving the pack density. A tenfold improvement was achieved thus making it suitable for consolidation. Uniaxial compaction of these powders for conventional sintering at pressure below 200 MPa yielded compacts free from defects. However, at higher pressure the compacts cracks and delaminates during ejection stage. Analysis of the compaction curves helped us to understand various processes involved during compaction as well as providing explanation for lower green density of Fe-Cu powder compared to Fe. Conventional sintering of the nanopowders compacts were carried out in the flowing hydrogen atmosphere in a laboratory vacuum furnace over wide range of temperatures. Instrumented sintering experiments were also carried out in a dilatometer under hydrogen atmosphere to evaluate shrinkage rate at different temperatures. SPS were carried out under 10 Pa vacuum at a compaction pressure of 250 MPa in WC-Co die cavity. The stability, density and residual oxygen content of the sintered compacts were quantified. Detailed microstuctural analysis of the sintered samples were also carried out using optical microscopy, atomic force microscopy, scanning electron microscopy, transmission electron microscopy, scanning transmission electron microscopy and quantitative composition analysis by EDS. Conventionally sintered compacts of both the powders showed stability only when sintered at 700°C and beyond. The maximum shrinkage/densification occured around 450°C for both the powders and the densification rate observed in Fe-Cu is three times higher than Fe. This enhanced densification in Fe-Cu is attributed to an enhanced diffusivity of Fe atoms in the transient liquid Cu layer formed at the interfaces. The microstructure of Fe-Cu is completely free from any separate oxide phase unlike in samples of Fe sintered at 700°C that contain ultrafine oxide grains. This was explained on the basis of role played by acrawax, a lubricant, admixed to increase the green density. Based on the sintered densities of both SPS and conventional sintered compacts, residual oxide content of the compacts sintered at different temperature and experimentally observed shrinkage rate a phenomenological model has been proposed for the possible sequence of processes occurring during sintering of these nanopowders. The major findings of this work are summarized in chapter 6 and chapter 7 details the scope for the future work.