IMPEDANCE MEASUREMENT SYSTEMS AND THEIR APPLICATION FOR 2D MAPPING OF CONDUCTIVITY

Youssoufa Mohamadou Kyung Hee University 대학원 2015 국내박사

Bio-impedance spectroscopy (BIS) or electrical impedance spectroscopy (EIS) is a method in which the frequency dependent impedance of loads are obtained. Current sources are generally used in impedance spectroscopy measurement systems to maximize current injection for increased signal to noise while keeping within medical safety specications. High performance current sources based on the Howland current pump with optimized impedance converters are able to minimize stray capacitance of the cables and setup. This approach is limited at high frequencies primarily due to the deteriorated output impedance of the constant current source when situated in a real measurement system. For this reason, voltage sources have been suggested, but they require a current sensing resistor and the SNR reduces at low impedance loads due to the lower current required to maintain constant voltage. In this thesis I compare the performance of a current source based BIS and voltage source based BIS which use common components. The current source BIS is based on a Howland current pump and generalized impedance converters to maintain a high output impedance of more than 1 M at 2 MHz. The voltage source BIS(EIS) is based on voltage division between an internal current sensing resistor (RS) and an external sample. To maintain high SNR, RS is varied so that the source voltage is divided more or less equally. In order to calibrate the systems, I measured the transfer function of the BIS systems with several known resistor and capacitor loads. From this I may estimate the resistance and capacitance of biological tissues using the least squares method to minimize error between the measured trans-impedance excluding the system transfer function and that from an impedance model. When tested on realistic loads including discrete resistors and capacitors, and saline and agar phantoms, the voltage source based BIS system had a wider bandwidth of 10 Hz to 2.2 MHz with less than 1 % deviation from the expected spectra compared to more than 10 % with the current source. The voltage source also showed an SNR of at least 60 dB up to 2.2 MHz in comparison to the current source based BIS system where the SNR drops below 40 dB for frequencies greater than 1 MHz. The BIS(EIS) system was used to quantitatively understand its electromechanical properties of conductive fabrics when pressure is applied to the fabrics. The aim here is to set a standard way to compare and choose conductive fabrics which will be suitable candidates for pressure sensing and mapping using electrical impedance tomography technique. I investigated electromechanical properties of two sets of three dierent sensors. The rst set comprised of three conductive fabrics. The second set comprised of a conductive fabric, fabric combined with polyurethane (PU) nanoweb, and fabric combined with polyvinylidene uoride (PVDF) nanoweb. I found that their electrical impedance spectra depend not only on the electrical properties of the conductive yarns, but also on their weaving structures, the type of nanoweb used and their size. When I apply a mechanical tension or compression, there occur structural deformations in the conductive fabrics altering their apparent electrical impedance spectra. For a stretchable conductive fabric, the impedance magnitude increased or decreased under tension or compression, respectively. Adding a sheet of nanoweb between two pieces of conducting fabrics produce a capacitive eect which increases the imaginary part of the impedance and hence increasing the sensitivity of the sensor. In the second part of this thesis I present the design of planar electrical impedance imaging (planar-EII) system and a microscopic electrical impedance imaging (micro-EII )system with planar electrodes. Both systems uses the same technique of current injection, voltage measurement and image reconstruction method. The dierence being the size of their electrode array. The planar-EII system has four current injection electrodes and 224 voltage sensing electrodes. The size of the sensing electrodes were 2mm in diameter and are 5mm apart. The micro-EII with planar electrodes on the other hand has four current injection electrodes and 112 voltage sensing electrodes. The size of the sensing electrodes were 0.4mm in diameter and an inter electrode distance of 0.8mm apart. Performance evaluation of the systems showed that both systems have high current source output impedances which is above 1 MHz and SNR of 60dB and above at all frequencies. The GREIT performance indices were computed for object at two dierent position and the results show that the planar-EII systems has Amplitude response (AR) of around 1.5, Position error (PE) of 0.04, Resolution (RES) of 0.4, Shape deformation (SD) was about 0.1 for object at the center of imaging domain and Ringing (RNG) was zero. The micro-EII system on the other hand had Amplitude response (AR) of around 1.0, Position error (PE) of 0.09, Resolution (RES) of 0.3, Shape deformation (SD) was about 0.5 for object at the center of imaging domain and Ringing (RNG) was zero. The systems where used to detect objects places inside a container with the electrodes at the bottom of the container.

Taxonomic Characterization of Three Novel Bacteria

Eul Kon Kim Kyung Hee University 대학원 2015 국내박사

Panax ginseng Meyer, belonging to the Araliceae family, which mainly cultivated in China, Korea and Japan. It has been used as traditional medicine in Asian countries a thousand years ago. Korean Panax ginseng have pharmacological effects such as memory improvement, protects memory, anti-cancer, anti-tumor, anti-stress, blood pressure lowering, anti-diabetic, improving liver functions, adjusting blood pressure, improving male sexual dysfunctions, and inhibition of AIDS virus (HIV) growth. During the long time cultivation ginseng soil is one of the best habitats for bacteria, yeast and fungi growth. The ginseng soil sample was collected from near the rhizophere of the ginseng root in Yeoncheon, Hwacheon and Gochang Republic of Korea. Three novel bacteria was found, designated as strains DCY60T, DCY90T, DCY100T were investigated for taxonomic study. All novel strains were Gram-staining positive, rod-shaped bacteria. The novelty of strain was confirmed by 16S rRNA gene sequence. The novel strain must be have 16S rRNA gene sequence similarity less than 97% or whole genome no more than 70% DNA relatedness with whole genome. The polyphasic characteristic was conducted for three novel strain based on the physiological characteristics, biochemical characteristics, and chemotaxonomic characteristics. Base on the polyphasic taxanomic characterization, three novel bacterial strains were isolated from ginseng soil in Yeoncheon and Hwacheon Republic of Korea. One novel strain belonging to the phylum Actinobacteria for which name Microbacterium rhizomatis sp. nov. Two novel strains belonging to the phylum Bacteroidetes for which name Humibacter ginsengiterrae sp. nov., and Humibacter ginsengisoli sp. nov.

Investigation on Synthesis, Characterization and Electrochemical Behavior of Efficient Electrode Active Materials for Rechargeable Lithium Batteries

ThiruvengadamSubburaj Kyung Hee University 대학원 2015 국내박사

In recent years, incredible consideration has been paid to the development of high-power lithium-ion batteries for electric vehicles (EVs) and hybrid electric vehicles (HEVs) to satisfy the energy and environmental concerns. Unfortunately, the presently used materials cannot meet the high-power demand due to their (i) expensive synthesis, (ii) old chemistry, (iii) thermal instability, (iv) safety hazards, and (v) poor rate performances. My works have been focused on finding the solution of above mentioned problems. High capacity layered Li[Lix(Ni0.3Co0.1Mn0.6)1-x]O2(x = 0.11) cathode and high voltage Li4Ti5O12 spinel anode for the full cell systems have been studied. With combined this cathode and anode materials in an advanced lithium ion battery, a new electrochemical reaction has been demonstrated. It exhibited the initial discharge capacity of 173 mAh g-1 and also was maintained at the average specific capacity of 170 mAh g-1 and more than 90% capacity retention. In addition, the battery can cycle at C/10 rate up to 30 cycles with a very stable capacity delivery. Taking an average voltage of 2.5 V, a specific energy density value of 400 Wh kg-1 was obtained. These environmentally benign Mn-rich layered cathode and spinel Li4Ti5O12 anode materials are going to be future electrodes for Li-ion rechargeable batteries in hybrid electric vehicle applications. The conventional electrolyte system has been compared with the ionic liquid (IL) additive containing electrolyte system at room temperature as well as elevated temperature. In this work, two types of monocationic ILs such as 1-butyl-3-methylpyrrolidinium hexafluorophosphate (Pyr IL) and 1-ethyl-3-methylimidazolium hexafluorophosphate (IMI IL) are added as an additive at two different weight ratios in 1.15 M LiPF6 (EC/EMC=3/7 v/v) electrolyte solution, the structural, electrochemical and thermal characteristics of LiNi0.80Co0.15Al0.05O2 (NCA)/carbon full-cell in different electrolyte formulations have been reconnoitered. X-ray diffraction (XRD) studies have proved that IL as an electrolyte additive does not alter the structural stability of cathode materials after cycling. Under room temperature, Pyr IL additives at 1wt% and 3wt% deliver better cycleability than others, with the retention ratios of 93.62% and 92.8%, respectively. At elevated temperature, only 1wt% Pyr IL additive is giving stable capacity retention ratio of 80.74%. Ionic conductivity and self-extinguishing time (SET) values are increasing with respect to the amount of additive added to the electrolyte. Thermal studies reveal that 3wt% Pyr IL is favorable regarding the safety of the battery as it shows shifting of peak to higher temperature of 272.10°C. Among the IL additives evaluated in this study, addition of 1wt% Pyr IL is the most desirable additive for achieving the best cycling performance as well as thermal behavior of Li-ion batteries. TiO2 nanofibers (NFs) are prepared by electrospinning method and they are decorated over the surafce of Ni-rich Li[Ni0.8Co0.15Al0.05]O2 (LNCA) cathodes at three different ratios namely 0.5 wt%, 1 wt%, and 1.5 wt%, respectively. The structrual, morphological, electrochemical, and thermal characteristics of TiO2 NFs decorated LNCAs are compared with pristine LNCA. From the X-ray diffraction (XRD) and field emission-scanning electron microscopy (FE-SEM) investigation, it is evident that the TiO2 NFs are decorated over the surface of cathode powder particles and the NFs are acting like a connecting bridge between the LNCA particles. The TiO2 NFs decorated LNCA electrodes are giving better cycleability and charge-discharge capacity than the pristine LNCA. Under room temperature at C/10 rate, 1 wt% TiO2 NFs decorated LNCAs is giving the capacity retention of 89.2%. Whereas at elevated temperature, it is delivering 81.4% of capacity retention. The onset of thermal decomposition temperature is also shifted towards the higher temperature for TiO2 NFs decorated LNCA electrodes than pristine LNCA electrodes. The improved electrochemical and thermal behaviors of the Ni-rich LNCA is achieved by decorating TiO2 NFs over the LNCA surface. Micro-sized Li4Ti5-xBixO12 (0 ≤ x ≤ 0.15) materials are synthesized using a simple solid state method in air. The structural, morphological, and electrochemical characteristics of Bi-doped lithium titanates and pristine samples are methodically analyzed by X-ray diffraction (XRD), Raman spectroscopy, field emission-scanning electron microscopy (FE-SEM), and electrochemical impedance spectroscopy (EIS). The XRD and Raman spectroscopy results demonstrate that bismuth-doping do not alter the spinel structure and good crystalline materials are synthesized. The FE-SEM images show that all samples possess the same morphological characteristics, with a particle size distribution of 0.5-1 μm. The electrochemical cycling testing reveals that the Li4Ti4.9Bi0.10O12 sample exhibits discharge capacities of 205.4 mAh g-1, 160.8 mAh g-1, and 135.4 mAh g-1 after 50 cycles at 1C, 5C, and 10C-rates, respectively. The differential capacity curves suggest that the Li4Ti4.9Bi0.10O12 sample has a weaker polarization effect than the other samples. The EIS measurements imply that the Li4Ti4.9Bi0.10O12 sample possesses a high electronic conductivity and lithium ion diffusivity, which demonstrate that this new Li4Ti4.9Bi0.10O12 material would be a good candidate as an anode for lithium ion batteries. A nano-ribbon-like Cu/Cu0.85V2O5 hybrid composite material with a width of 100 nm has been synthesized via a simple, one-step hydrothermal method. The structural and morphological characteristics of the hybrid composite confirm the phase arrangement, composition, and nano-ribbon-like morphology. The formation reaction mechanism has been investigated via hydrothermal reaction time dependent variables, and the electrochemical characterization demonstrates good cyclical behavior. The hybrid composite delivered approximately ~190 mAh g-1 of the specific discharge capacity at a current density of 50 mA g-1 as a cathode, and approximately ~600 mAh g-1 of the specific discharge capacity at a current density of 100 mA g-1 as an anode. These enhanced electrochemical properties can be ascribed to an increase in electronic conductivity via the presence of metallic copper in the hybrid structure and the facilitation of better intercalation-deintercalation of lithium ions due to the porous nano-ribbon-like structure.

Deflection Assessment of Prestressed Concrete One-Way Slabs

Ha Eun Park Kyung Hee University 대학원 2015 국내석사

ABSTRACT Prestress concrete is a method for overcoming concrete's weakness in tension. Prestressing tendons are used to produce a compressive stress that balances the tensile stress. Thus, economic design can be realized because as the span length increases, the cross-sectional area decreases, and the structure becomes stronger than the existing reinforced concrete structures. But, since members of prestressed concrete are thinner and have longer span lengths than those of reinforced concrete structures, the serviceability assessment on deflection should receive greater focus. Hence, the present study examined the one-way slab deflection of prestressed concrete by considering span length, concrete compression strength, eccentricity, live load, prestressing force, and tendon layout as variables. SAP2000, the commonly used computer program for finite element analysis, was used for analysis, and the validity of the analysis method was verified by comparing the analysis results with experiment results. Deflection aspects according to variables, this study verified that the deflection is lesser as the span length is lesser, concrete compressive strength is greater, eccentricity is greater, live load is lesser, and prestressing force is greater. Also, the straight tendon showed the least deflection and the bending tendon showed the greatest deflection. Regression analysis was conducted to analyze how camber was influenced by the span length, concrete compressive strength, eccentricity, and prestressing force. The result of comparing the regression analysis with those of calculation, calculation overestimated the camber compared to the regression analysis.

Anisotropic Conductivity and Current Density Imaging in MREIT

Saurav Zaman Khan Sajib Kyung Hee University 대학원 2015 국내박사

A new method for visualizing the anisotropic conductivity distribution has proposed based on the cross property relationship between the eigen values of the conductivity and diffusivity tensor. The proposed method, however, determines the scale factor or the effective conductivity to diffusivity ratio (ECDR) from the data obtained in the MREIT and DT-MRI experiment. Since MREIT technique is widely used to determine the internal conductivity distribution at each pixel position by measuring the induced magnetic flux density data therefore, by incorporating with the measured diffusion tensor information the position dependent scale factor can be determined using the proposed DT-MREIT technique. Visualizing the anisotropic conductivity distribution using the proposed DT-MREIT technique requires to recover the current density information from the measured magnetic flux density data. Since in MREIT, only one component of the magnetic flux density data is available, the method projected current density is adopted. However, for the animal and human imaging experiment the target imaging slice often contains the poor SNR region such as skull and/or air filled lung etc. Therefore, inclusion of the defective region causes instability in the recovered projected current density. In order to alleviate such problem a new technique called regionally projected current density is proposed. This proposed stably determines the current density inside the local ROI region. These two proposed methods are first validated by the numerical simulation and carefully designed phantom experiment. By combining the proposed local projected current density with the DT-MRIET technique the anisotropic conductivity distribution inside the canine head was evaluated. Transcarinel direct current stimulation (tDCS) is a neuro-modulatory technique used for treating the patient for a certain neurological disorder. In tDCS typically 1 » 3 mA current is injected through the attached pair of surface electrodes for the duration of 10 » 20min on the important anatomical brain location. The projected current density method in MREIT cannot effectively handle the non-transversal current flow caused by the tDCS current inside the anisotropic brain region. Therefore, a new procedure for improving the projected current density method is presented by incorporating the cross property relationship. Using a realistic head model the numerical simulation result shows the new method can reliably determine the current density caused by the tDCS current flow inside the brain region. MR driven individual modeling study had been widely used to understand the effect of electrode positing for the tDCS treatment. However, these methods do not consider the individual conductivity distribution which further affects the current flow inside the brain region. In addition to that, no successful method had not been reported yet for monitoring the stimulation effect during tDCS. These newly proposed techniques by utilizing the cross property relationship in MREIT, are expected to be useful for the personalized treatment planing as well as monitoring during tDCS.

The monitoring of food web and coastal zooplankton community in a eutrophic-brackish reservoir, Saemangeum

Sho Nakano Kyung Hee University 대학원 2015 국내석사

The Saemangeum dyke construction created huge brackish reservoir having strong eutrophication impacts from the two rivers (Mangyung and Dongjin River) as well as limited seawater inputs from the open sea through the sluice gates of the dyke simultaneously. The Saemangeum coast provides chances to study the relation of aquatic ecosystem and environmental changes in coastal area particularly the impacts of eutrophication and salinity changes In the present study, dynamics of zooplankton community, of which composition and distribution shows significant responding to environmental changes sensitively was analyzed in accordance with environmental changes. The distribution of zooplankton community with special emphasis on copepods community which is dominant species in Saemangeum coast, and the relationship with environmental factors were analyzed by Canonical Correspondence Analysis (CCA) and Generalized liner model (GLM). Further, the food web analysis using staple isotope ratios (δ13C and δ15N) was carried out for quantitative estimation of the ecosystem structure and function of Saemangeum coast. Dual δ13C and δ15N composition indicate the origin of organic matter and nutrients flows (i.e. distribution of food sources and pollutants) and trophic relation. The sample collection including particulate organic matter (POM), zooplankton, fish and sediment for food web analysis were conducted in each season (July, October 2013, February, April, August, October 2014 and February 2015) at 17 sites in the Saemangeum coast including two river water input points with water quality monitoring. As results, the brackish and coastal copepod species, Acartia spp., Oithona spp. and Paracalanus spp. appeared as dominant species during the study period and their distribution pattern showed strong seasonality. Especially, frequent appearance and high density of Oithona spp. which prefer the eutrophic environment can suggest eutrophication is progressing in Saemangeum coast. In addition, the results of CCA analysis suggested that copepod species responded to salinity, COD and water temperature significantly. As a result of food web analysis, the isotope ratios indicated complicated food web structure of the Saemangeum coast during survey period. Especially, the δ13C signals of POM showed bifacial patterns of environmental changes, desalination by freshwater accumulation in summer season and restoration of original oceanic status in winter season. Zooplankton and fish communities showed seasonally different distribution patterns and consequently different food web structure considering the feeding patterns according to environmental changes. Therefore, it can be suggested that desalination and eutrophication are main environmental factors determining the species composition and distribution patterns of zooplankton and fish communities.

Approaches to Improve the Electrical and Ionic Conduction in Li-ion Batteries

Kadirvelayutham Prasanna Kyung Hee University 대학원 2015 국내박사

The present technology of Li-ion batteries supports the increased usage of portable electronic devices and electric vehicles, but the fast growing energy society demands more power and high energy. The thirst by the society for higher power and energy persuades the researchers to focus more on the electrode materials used during Li-ion batteries fabrication. In this thesis alternative electrode fabrication material, electrode material, and separator have been examined with a main focus on the improvement ionic and electronic conduction. For the cathode well known LiFePO4 is chosen due to its environmental friendly nature, material abundance, reduced cost, and increased specific capacity. The commercially used polyvinylidene ?uoride (PVDF) binder and the blending solvent N-Methylpyrolidone during the fabrication of LiFePO4 electrode were replaced using environmental friendly biopolymer chitosan and water, respectively, to overcome its foremost drawback of poor electronic conduction. On the anode side, silicon anodes have been investigated due its high theoretical capacity. Facile technique has been used to overcome its main challenge of volume expansion upon cycling. In accession, its electronic and ionic conduction is also amended for better electrochemical behavior. With regard to the separators physical and electrochemical properties of the stretched polyethylene separator is characterized and it is modified with SiO2 ceramic materials to improve its ionic conduction and safety properties. During LiFePO4 electrode fabrication, the biopolymer chitosan has been investigated as a potential binder. Chitosan is compared to the conventional binder PVDF. Dispersion of the active material, LiFePO4, and conductive agent, Super P carbon black, is tested using a viscosity analysis. The enhanced structural and morphological properties of chitosan are compared to the PVDF binder using X-Ray diffraction analysis (XRD) and field emission scanning electron microscopy (FE-SEM). Using an electrochemical impedance spectroscopy (EIS) analysis, the LiFePO4 electrode with the chitosan binder is observed to have a high ionic conductivity and a smaller increase in charge transfer resistance based on time compared to the LiFePO4 electrode with the PVDF binder. The electrode with the chitosan binder also attains a higher discharge capacity of 159.4 mAh g-1 with an excellent capacity retention ratio of 98.38% compared to the electrode with the PVDF binder, which had a discharge capacity of 127.9 mAh g-1 and a capacity retention ratio of 85.13%. Further, the cycling behavior of the chitosan-based electrode is supported by scrutinizing its charge-discharge behavior at specified intervals and by a plot of dQ/dV. A facile strategy is reported here to entrap the milled silicon (m-Si) particles using nitrogen doped carbon (N-C@m-Si) to overcome its drawback of vivid volume change during intercalation of lithium ions and also to improve its poor electronic conductivity behavior. The nitrogen containing biopolymer chitosan is used as a carbon source. Simple hydrothermal technique followed by a subsequent carbonization process is used to synthesis N-C@m-Si particles. The N-C@m-Si samples used as an anode in Li-ion batteries exhibits a significantly improved electrochemical performance compared to bare the m-Si samples, as confirmed by the obtained discharge capacity and columbic efficiency of 942.4 mAh g-1 and 97%, respectively after 50th cycle at 0.1 C rate. The solid electrolyte interface layer (SEI) formed over the m-Si and N-C@m-Si electrodes are well studied using X-Ray photoelectron spectroscopy (XPS) analysis to support the proposed hypothesis. In comparison to the SEI layer formed over m-Si electrode after 10 charge-discharge cycles, the N-C@m-Si electrode has a significantly less amount of lithium fluoride (LiF) and fluorinated C and Si species. This is due to the difference in surface chemistry between the m-Si and N-C@m-Si samples. Brief reaction mechanism representing the formation of different species in the SEI layer is derived, to interpret its behavior during electrochemical process. The single layered microporous polyethylene separator is prepared by dry process and has been stretched in uni-axial direction to two different ratios namely 180% and 300% in order to create high-performance and cost effective separator for practical application in lithium-ion batteries. In this study, the structures of 180% and 300% stretched polyethylene separators were characterized. The physical structure of the stretched separator is characterized by key factors such as thickness, mean pore size, porosity, Gurley value, ionic resistivity, MacMullin number, and tortuosity. The thermal behavior of the stretched separator is explained by using differential scanning calorimeter (DSC). DSC explains the melting and shutdown behavior of the separator. Electrochemical property is studied by linear sweep voltammetry, electrochemical impedance spectroscopy (EIS) and cyclic performance. EIS is performed to explain, in elaborate, the resistance of the separator and the specific discharge capacity is observed using the cyclic performance. 300% stretched separator is observed to have comparatively less resistance and higher discharge capacity than the 180% stretched separator. The tested 180% and 300% stretched separators have been coated with a SiO2 ceramic substance to increase ionic conductivity and thermal stability without sacrificing mechanical properties. To test the thermal and tensile properties, thermomechanical analyzer (TMA) is employed. CR 2032-type coin cells are prepared by sandwiching pristine and coated stretched separators, respectively, between the Li anode and Li[Ni1/3Co1/3Mn1/3]O2 cathode to evaluate the AC impedance and cyclic performance. The coated separators are observed with superior ionic conductivity, thermal and tensile properties. The cells prepared with coated separator have slightly higher discharge capacity and a better capacity retention ratio than the cells with pristine separators. These results suggest that the coated separator is a better option for lithium ion batteries.

Magnetic Capture of Iron-containing Particles - Numerical Simulation Study

Shan Huang Graduate School, Kyung Hee University 대학원 2015 국내박사

Due to the adverse health effects, metro-subway dust with much higher concentration than outdoor atmosphere has aroused increasing attention. Mechanical wear and attrition of rails and wheels in underground tunnels are the main emission sources. The subway dust is collected as circulating mechanical ventilation chambers of the station, but some currently are discharged to outdoor through the ceiling duct without any control facilities. It has been known that the dust contains a fair amount of iron compounds which have magnetic properties. Thus several works attempted to collect the subway dust through magnetic filtration, and this study includes close examination of dynamic behavior of iron-containing dust in magnetic fields by numerical simulation. Characterization of subway dust showed the content of iron compounds; 54.5% and 35.9% from Daecheong and Jegi station, respectively. XRD analysis indicated SiO2, Ca(SO4)(H2O)2 and magnetite (Fe3O4) and/or maghemite(γ-Fe2O3) as major chemical species. Amongst these major ferric chemical species, magnetite and/or maghemite involve obvious magnetic properties. Magnetic property of the sample dust was analyzed by vibrating sample magnetometer (VSM). Dynamic behavior of subway dust particles in magnetic fields was investigated in terms of particle size, magnetic field intensity and flow velocity. Oscillation of fine particles in the magnetic field of the test channel could be seen owing to magnetic interforces. Particle deflection to the high gradient field was also affirmed. A high gradient magnetic filtration (HGMF) unit was designed for a lab scale. With the application of magnetic field (0.165 T-0.3 T), the capture efficiency for PM1.0, PM2.5 and PM10 were increased 10%-63%, 50%-80%, and 32%-87%, respectively for a medium filter. Two-step simulation was used to trace the particle motion in the HGMF system. Flow field and magnetic field were first calculated focusing on a macroscopic channel, then particle tracing proceeded in the microscopic subdomain in terms of flow velocity and magnetic field intensity was investigated. Particle collection in a subdomain was evaluated, and the efficiency distribution across the entire filter was deduced. The results showed that the capture efficiency depended strongly on particle size and magnetic field intensity as well as the magnetic field gradient. High magnetic field and high flow velocity facilitated high capture efficiency for particles larger than 2.5 μm while low velocity led to high capture efficiency for 1μm particle due to magnetic diffusional effect. Regression formula from the obtained efficiency was given as: . The coefficients are respective functions of particle size, magnetic flux density, wire diameter, wire opening, flow velocity and the magnet displacement. Under the turbulent condition, direct interception, turbulence dispersion and inertial impaction are potentially high on a fine filter. Thus, it was noticed that for fine filter, the velocity was directly proportional to the efficiency while medium and coarse filters showed the inverse relation. Larger particles were more effectively trapped on the convex part of wire along the flow direction. Main capturing regions of the filter existed in the lip shape site around the middle permanent magnet and the narrow circumferences between channel wall and magnets. The mean overall capture efficiencies for particles of 1 ?m, 2.5 ?m, 5 ?m and 10 ?m were 4.97%, 8.51%, 24.02%, and 35.31%, respectively on medium filter. Addition of magnetic field (0.3 T) to the mesh screen contributed to increment of the efficiency in 3.95%, 4.78%, 6.58%, and 8.23% for 1 ?m, 2.5 ?m, 5 ?m and 10 ?m particle. Submicron particle capture, in addition, was conducted in a microchannel magnetic system. Four configurations according to the magnet pieces and with or without mesh screen filter were prepared. Magnets were arranged in an attractive array to form strong magnetic field. Insertion of a few magnets formed larger effective magnetic field in the mesh screen than a single piece. The magnetized ferromagnetic mesh screen enhanced the magnetic gradient across the flow path; thus particle separation from the main stream was improved more than 80% for 1 ?m particle. The thinner the wire, the higher magnetic field gradients were formed promoting particle capture. A concise particle velocity model was presented in this study and enabled to understand the motion and collection of submicron particles from 0.1 ?m to 0.5 ?m in size. The repulsive (positive particle velocity) and attractive (negative particle velocity) zones were formed evenly in front and backside of the wire without external magnetic field. However, magnetic field altered the shape of particle velocity zones, and the attractive zones appeared in the front and enhanced more at the back side of the wire. Meanwhile, the particle velocity magnitude was also found to be significantly depending on wire thickness, filter mesh packing density and flow feeding velocity. Large flow velocity led to larger repulsive regions occurring mainly in the front side of wire, hindered the particles’ approach to the front attractive zones and finally aided to be captured by the magnetized wires. Thin wire produced strong attractive zones both in front and behind of the wires, with repulsive sections by the sides between the attractive zones. High packing density caused high capture efficiency due to the dense wire interception. On the other hand, capturing at the backside was frequently found for fine particles of 0.1 ?m and 0.2 ?m. It was valid to evaluate the capture efficiency based on the critical capture distance at the symmetry boundaries. Conclusively, the present study has provided an alternative design of magnetic filtration with core operation factors and thereby estimated the particle capture efficiency through mathematical models. Collection efficiency was directly proportional to the particle size, magnetic flux density, while is inversely or directly proportional to the velocity depending on the flow pattern. The obtained results could optimize the design configuration and operation condition of the magnetic pre-filtration.

Analysis and Experiment of High-Efficiency, Free-Positioning, Power Division Mid-Range Wireless Power Transfer System

DUONG PHI THUC Kyung Hee University, Graduate School 대학원 2015 국내박사

Recent years, with the revolution in term of number of portable devices such as mobile phones, laptops, digital cameras, robots, electric vehicles…which are mainly powered by baterries, the need of recharging baterries often and often becomes very vital. This problem motivates the development of wireless power transmission, where energy from power source is transferred to electrical load without connecting wires. In fact, wireless powering is not a new technology, its intial background theory was proposed by Nikola Tesla in years of 1890. However, in today’s society wireless power technology has not been used widely due to some limitations such as inefficiency at far distances, free-positioning and multiple receivers issue, and human safety condition, etc. Wireless power works are principally directed into two types: inductive coupling for short-range applications and magnetically resonant coupling for mid-range applications. The principle of resonant coupling is based on the physical concept that two resonators with the same resonant frequency are able to effectively exchange power. In practical systems, the transmission distance and position is usually changed and that makes the efficiency of system drop rapidly. Therefore, it is essential to analyze these effects in order to find the condition to maintain high power delivery efficiency in these systems. In this work, a method of improving efficiency when changing the distance between resonators is proposed. This technique does not require additional lossy matching network, and is based on optimally adjusting the coupling factors between source (load) and the internal resonator.