No description available.

Standard eddy-viscosity models lack curvature and system rotation sensitized terms in their formulation. Hence they fail to capture the effects of curvature and system rotation on turbulence anisotropy. As part of this effort, an algebraic expression for a characteristic rotation term is developed and tuned with the help of rotating homogeneous shear flow. This formulation is primarily based upon the rotation and curvature sensitized eddy-viscosity coefficient developed by York et al. (2009). A new scalar transport equation loosely based on Durbin’s wall normal turbulent velocity scale (Durbin, 1991) is introduced to account for the modification in turbulence structure due to system rotation and curvature effects. The added transport equation also introduces history effects and stability in the solution with small increase in computational cost. The eddy-viscosity is redefined based on new turbulent velocity scale and hence the effects of rotation and streamline curvature are introduced into the mean momentum equation. A number of canonical test cases with significant curvature and rotation effects along with a cyclone flow, a representative of complex industrial flows, are considered for model validation. Hybrid modeling framework combines the strength of RANS in boundary layers and LES in separated shear layers to alleviate the weaknesses of RANS and limitations of LES model in some complex flows. A recently proposed hybrid RANS-LES modeling framework uses a weighing parameter that dynamically determines the RANS and LES regions based on solution statistics. The hybrid modeling methodology is implemented on a normal jet in crossflow, and a film cooling case for the purpose of model validation and evaluation. The final goal of the proposed effort is to combine advanced RANS modeling capability with LES using the new hybrid modeling framework. Specifically, the curvature and rotation sensitive RANS model developed here is coupled with commonly used LES models to produce a novel model for complex turbulent flows with the potential to improve accuracy of CFD predictions (versus existing RANS models) as well as significantly reduce the computational expense (versus existing LES models). Performance of the model form hence developed is evaluated on a cyclone flow case.

Electrically conductive membranes (ECMs) are of significant research interest for their ability to mitigate fouling, enhance separation capacity, and induce electrochemical degradation of contaminants. Most ECM development has been in flat sheet format suitable for laboratory studies; in industrial applications, formats such as hollow fiber (HF) are preferred for their high packing density. While ECMs in HF format are emerging in research, these techniques typically employ the same methods proven for flat sheet, often involving direct deposition of conductive material onto a support membrane with no further investigation into how the deposition process affects ECM properties. This is a significant challenge for long (~1 m) HF membranes where coating uniformity is essential to ensure consistent performance. The goal of this project was to fabricate conductive HF membranes, ensuring uniform performance along the fiber. In this work, we have developed a “crossflow deposition” technique to deposit a uniform layer of single walled/ double walled carbon nanotubes (SW/DWCNTs) onto the interior surface of commercial polyether sulfone HF membranes. In a design-of-experiments model, feed pressure and crossflow velocity were shown to directly impact composite membrane conductivity and permeability. The highest permeability (~2900 LMH/bar) and conductivity (~670 S/m) were both achieved at the high pressure (0.2 bar) and high crossflow velocity (1.06 cm/s) condition. An inverse relationship was identified between conductivity and permeability for 29 different HF membranes coated under various flow and particle loading conditions. Similar trends were evident in ECM literature when comparing 80 membranes across 38 papers, covering various conductive materials, separation types, configurations, and applications. Metallic-based ECMs outperformed graphitic nanomaterial or conductive polymer-based ECMs with conductivities three orders of magnitude higher. This review also revealed a wide variation in performance testing with 35 unique pollutants in 63 total tests, indicating a need for standardization to accurately compare ECMs and a need for testing with more realistic feed sources. Finally, electrochemical degradation of methyl orange using the CNT-coated HF membranes was evaluated in batch and continuous removal experiments. Although no significant MO removal was detected in either configuration, these modules can be used for further optimization in terms of targeted conductivity, contact time, and electrochemical parameters such as applied voltage. This work highlights the existence of a conductivity/ permeability trade-off in ECM development and how manipulation of flow parameters during deposition can impact this trade-off in HF membrane development. / Thesis / Master of Applied Science (MASc) / Membrane separation technologies are a common purification strategy in many fields due to their simplicity and low energy requirements. Membranes operate by rejecting particles from feed water based on their chemical or physical properties such as size or charge. Long-term membrane operations are limited by fouling, incurring large operating costs for frequent cleaning cycles and downtime. Furthermore, traditional membrane separations only physically remove particles, presenting a risk for contaminant re-introduction into the environment. Electrically conductive membranes are an emerging strategy for addressing these concerns due to their demonstrated antifouling, enhanced selectivity, and redox capabilities. To date, these membranes have almost exclusively been developed as flat sheets with limited research into other membrane formats. Hollow fiber membranes resemble thin tubes ~1 mm in diameter and up to ~1 m in length where filtration occurs through the tubular wall of the fiber; the small diameter allows for hundreds of fibers to pack into an individual module, thus maximizing throughput. In this thesis, several issues with hollow fiber conductive membrane fabrication are addressed to ensure consistent performance along the length of the fiber. A key trade-off between membrane surface conductivity and throughput was found to exist universally in the conductive membrane field. This knowledge can be used to select fabrication methods and parameters to target certain performance ranges.

In order to follow increasingly strict regulation of pollutant emissions, a new concept of Lean Premixed pre-vaporized (LPP) combustion has been proposed for turbines. In LPP combustion, controlled atomization, dispersion and vaporization of different types of liquid fuel in the premixer are the key factors required to stabilize the combustion process and improve the efficiency. A numerical study is conducted for the fundamental understanding of the liquid fuel dispersion and vaporization process in pre-mixers using both cross-flow and co-flow injection methods. First, the vaporization model is validated by comparing the numerical data to existing experiments of single droplet vaporization under both low and high convective air temperatures. Next, the dispersion and vaporization process for biofuels and conventional fuels injected transversely into a typical simplified version of rectangular pre-mixer are simulated and results are analyzed with respect to vaporization performance, degree of mixedness and homogeneity. Finally, collision model has been incorporated to predict more realistic vaporization performance. Four fuels, Ethanol, Rapeseed Methyl Esters (RME), gasoline and jet-A have been investigated. For mono-disperse spray with no collision model, the droplet diameter reduction and surface temperature rise were found to be strongly dependent on the fuel properties. The diameter histogram near the pre-mixer exit showed a wide droplet diameter distribution for all the fuels. In general, pre-heating of the fuels before injection improved the vaporization performance. An improvement in the drag model with Stefan flow correction showed that a low speed injection and high cone angle improved performance. All fuels achieved complete vaporization under a iv spray cone angle of 140°. In general, it was found that cross-flow injection achieved better vaporization performance than co-flow injection. A correlation is derived for jet-A‟s total vaporization performance as a function of non-dimensional inlet air temperature and fuel/air momentum flux ratio. This is achieved by curve-fitting the simulated results for a broad range of inlet air temperatures and fuel/air momentum flux ratios. The collision model, based on no-time-counter method (NTC) proposed by Schmidt and Rutland, was implemented to replace O‟Rourke‟s collision algorithm to improve the results such that the unphysical numerical artifact in a Cartesian grid was removed and the results were found to be grid-independent. The dispersion and vaporization processes for liquid fuel sprays were simulated in a cylindrical pre-mixer using co-flow injection method. Results for jet-A and Rapeseed Methyl Esters (RME) showed acceptable grid independence. At relatively low spray cone angle and injection velocity, it was found that the collision effect on the average droplet size and the vaporization performance were very high due to relatively high coalescence rate induced by droplet collisions. It was also found that the vaporization performance and the level of homogeneity of fuel-air mixture could be significantly improved when the dispersion level is high, which can be achieved by increasing the spray cone angle and injection velocity. In order to compare the performance between co-flow and cross-flow injection methods, the fuels were injected at an angle of 40° with respect to the stream wise direction to avoid impacting on the wall. The cross-flow injection achieved similar vaporization performance as co-flow because a higher coalescence rate induced by droplet collisions cancelled off its higher heat transfer efficiency between two phases for cross-flow injections.

A better understanding of several mineral processing devices and applications was gained through studies conducted with mathematical analysis tools. Linear circuit analysis and population balance modeling were utilized to remedy inefficiencies found in a number of popular mineral processing water-based unit operations. Improvements were made in areas, including unit capacity and separation efficiency. One process-engineering tool, known as linear circuit analysis, identified an alternative coal spiral circuit configuration that offered improved performance while maintaining a reasonable circulating load. In light of this finding, a full-scale test circuit was installed and evaluated at an existing coal preparation facility. Data obtained from the plant tests indicate that the new spiral circuit can simultaneously reduce cut-point and improve separation efficiency. A mathematical population balance model has also been developed which accurately simulates a novel hindered-bed separator. This device utilizes a tangential feed presentation system to improve the performance of conventional teeter-bed separators. Investigations utilizing the mathematical model were carried out and have predicted solid feed rates of up to 71 tph/m² (6 tph/ft²) can be achieved at acceptable efficiencies. The model also predicts that the unfavorable impact of operating at low feed percent solids is severely reduced by the innovative feed presentation design. Tracer studies have verified that this system allows excess feed water to cross over the top of the separator without entering the separation chamber, thereby reducing turbulence. A hindered-bed separator population balance model was also developed whose results were utilized to improve the efficiencies encountered when using a teeter-bed separator as a mineral concentrator. It was found that by altering the apparent density of one of the feed components, the efficiency of the gravity separation could be greatly improved. These results led to the development of a new separator which segregates particles based on differences in mass after the selective attachment of air bubbles to the hydrophobic component of the feed stream. Proof-of-concept and in-plant testing indicate that significant improvements in separation efficiency can be achieved using this air-assisted teeter-bed system. The in-plant test data suggest that in some cases, recoveries of the plus 35 mesh plant feed material can be increased by more than 40% through the application of this new technology. / Ph. D.

This paper aims to study how varying crossflow burning temperatures from 1100 C to 1800 C affect the liquid droplet breakup, size distribution, and atomization of a liquid water jet injected into a vitiated crossflow. The LJIC injection mechanism was implemented using the high-pressure axially staged combustion facility at the University of Central Florida. The measurement devices used to gather particle data from the exhaust plume were the TSI Aerodynamic Particle Sizer (APS), which measures particles between 0.523 µm and 20 µm, and the Sensirion SPS30 (SPS30), which measures particles between 0.3 µm and 10 µm. Both measurement devices were placed 3 ft away from the choked exit. Table 3 shows that the 1800 C crossflow temperature behaved as predicted by having the largest particle distribution of 67.97% and the largest particle count of 19,301 at 0.523 µm. The 1100 C crossflow produced the second-largest normalized particle count of 66.69% and raw particle count of 20,209 at 0.523 µm. This result is contrary to the original hypothesis because it shows that the relationship between temperature and particle count is non-linear and that many other factors must be at play in the atomization process, such as the droplet distribution at the nano level. The SPS30 was used to compare the particle size distributions between a 1500 C and 1800 C crossflow. Acquiring number concentration data for particles up to 10 µm in size, the 1800 C crossflow had a distribution peak at 802.76416 N/cm3, and the 1500 C crossflow had a peak of 867.28272 N/cm3. For the 0.5 µm peak, The 1800 C had a 10 µm particle size distribution peak at 674.27.76416 N/cm3, and the 1500C crossflow had a peak of 730.501 N/cm3. The decreased number concentration from 1500 C to 1800 C case grants the water particles in the 1800 C crossflow increased surface area, which allows for increased heat exposure from the vitiated crossflow [7]. Despite some nonlinear particle count results, the highest crossflow temperature of 1800 C produces the best atomization results by reducing the total particle count and having the largest collection of particles at the lowest detectable particle size of 0.523 µm.

京都大学 / 新制・課程博士 / 博士(工学) / 甲第23614号 / 工博第4935号 / 新制||工||1771(附属図書館) / 京都大学大学院工学研究科機械理工学専攻 / (主査)教授 黒瀬 良一, 教授 花崎 秀史, 教授 岩井 裕 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

This dissertation investigates potential applications for Machine Learning (ML) and real-time fouling monitors in Reverse Osmosis (RO) desalination. The main objective was to develop a framework that minimizes the cost of membrane fouling by deploying AI-generated cleaning patterns and real-time fouling monitoring. Membrane manufacturers and researchers typically recommend cleaning (standard operating procedure – SOP) when normalized permeate flow, a performance metric tracking the decline of permeate flow/output from its initial baseline with respect to operating pressure, reaches 0.85-0.90 of baseline values. This study used estimates of production cost, internal profitability metrics, and permeate volume output to evaluate and compare the impact of time selection for cleaning intervention. The cleanings initiated when the normalized permeate flow reached 0.85 represented the control for cleaning intervention times. In deciding optimal times for cleaning intervention, a Deep Reinforcement Learning (RL) agent was trained to signal cleaning between 0.85-0.90 normalized with a cost-based reward system. A laboratory-scale RO flat membrane desalination system platform was developed as a model plant, and data from the platform and used to train the model and examine both simulated and actual control of when to trigger membrane cleaning, replacing the control operator's 0.85 cleaning threshold. Compared to SOP, the intelligent operator showed consistent savings in production costs at the expense of total permeate volume output. The simulated operation using the RL initiated yielded 9% less permeate water but reduced the cost per unit volume ($/m3) by 12.3%. When the RL agent was used to initiate cleaning on the laboratory-scale RO desalination system platform, the system produced 21% less permeate water but reduced production cost ($/m3) by 16.0%. These results are consistent with an RL agent that prioritizes production cost savings over product volume output. / Doctor of Philosophy / The decreasing supply of freshwater sources has made desalination technology an attractive solution. Desalination—or the removal of salt from water—provides an opportunity to produce more freshwater by treating saline sources and recycled water. One prominent form of desalination is Reverse Osmosis (RO), an energy intensive process in which freshwater is forced from a pressurized feed through a semipermeable membrane. A significant limiting cost factor for RO desalination is the maintenance and replacement of semipermeable RO membranes. Over time, unwanted particles accumulate on the membrane surface in a process known as membrane fouling. Significant levels of fouling can drive up costs, negatively affect product quality (permeate water), and decrease the useful lifetime of the membrane. As a result, operators employ various fouling control techniques, such as membrane cleaning, to mitigate its effects on production and minimize damage to the membrane. This dissertation investigates potential applications for Machine Learning (ML) and real-time fouling monitors in Reverse Osmosis (RO) desalination. The main objective was to develop a framework that minimizes the cost of membrane fouling by deploying AI-generated cleaning patterns and real-time fouling monitoring. Membrane manufacturers and researchers typically recommend cleaning (standard operating procedure – SOP) when normalized permeate flow, a performance metric tracking the decline of permeate flow/output from its initial baseline with respect to operating pressure, reaches 0.85-0.90 of baseline values. This study used estimates of production cost, internal profitability metrics, and permeate volume output to evaluate and compare the impact of time selection for cleaning intervention. The cleanings initiated when the normalized permeate flow reached 0.85 represented the control for cleaning intervention times. In deciding optimal times for cleaning intervention, a Deep Reinforcement Learning (RL) agent was trained to signal cleaning between 0.85-0.90 normalized with a cost-based reward system. A laboratory-scale RO flat membrane desalination system platform was developed as a model plant, and data from the platform and used to train the model and examine both simulated and actual control of when to trigger membrane cleaning, replacing the control operator's 0.85 cleaning threshold. Compared to SOP, the intelligent operator showed consistent savings in production costs at the expense of total permeate volume output. The simulated operation using the RL initiated yielded 9% less permeate water but reduced the cost per unit volume ($/m3) by 12.3%. When the RL agent was used to initiate cleaning on the laboratory-scale RO desalination system platform, the system produced 21% less permeate water but reduced production cost ($/m3) by 16.0%. These results are consistent with an RL agent that prioritizes production cost savings over product volume output.

No description available.

The phenomenon of spray in crossflow is of relevance in gas turbine combustor development. The current work focuses on spray in crossflow rather than liquid jet in crossflow from the standpoint of enhancing fuel dispersion and mixing. Specifically, the first part of the work involves study of spray structure, droplet sizing, and velocimetry for sprays of water and ethanol in a crossflow under ambient conditions. Laser-based diagnostic techniques such as Particle/Droplet Image Analysis (PDIA) and Particle Tracking Velocimetry (PTV) are utilized. Using spray structure images, trajectory equations are derived by multi-variable regression. It is found that the spray trajectory depends only on the two-phase momentum ratio and is independent of other flow parameters. A generalized correlation for the spray trajectory is proposed incorporating the liquid surface tension, which is found to be effective for our data, with water and ethanol, as well as data on Jet-A from the literature for a wide variety of operating conditions. An interesting phenomenon of spatial bifurcation of the spray is observed at low Gas-to-Liquid ratios (GLRs). The reason for this phenomenon is attributed to the co-existence of large and highly deformed ligaments along with much smaller droplets at low GLR conditions. The smaller droplets lose their vertical momentum rapidly leading to lower penetration, whereas the larger ligaments/droplets penetrate much more due to their larger momentum leading to a spatial separation of the two streams. The second part of the study focuses on evaporating sprays in preheated crossflow. Experiments are conducted using ethanol, decane, Jet-A1 fuel, and a two-component surrogate for Jet-A1 fuel. The crossflow air is heated up to 418 K and the effect of evaporation is studied on spray trajectory and droplet sizes. Measured droplet sizes and velocities at two successive locations are used to estimate droplet evaporation lifetimes. Evaporation constant for the d2 law derived from the droplet lifetimes represents the first-ever data for the above-mentioned liquids under forced convective conditions. This data can be used to validate multi-component droplet evaporation models. The last part of the study focuses on Large Eddy Simulations (LES) of the spray in crossflow. The near-nozzle spray structure is investigated experimentally to obtain droplet size and velocity distributions that are used as inputs to the computational model. For the spray in crossflow under ambient conditions, trajectory and droplet sizes at different locations are compared with experimental results. While the predicted trajectory is found to be in good agreement with data, the predicted droplet sizes are larger than the measured values. This is attributed to the implicit assumption in the secondary breakup model that the droplets are spherical, whereas the experimental data in the near-nozzle region clearly shows presence of mostly ligaments and non-spherical droplets, especially for the low GLR cases. A modified breakup model is found to lead to improved agreement in droplet sizes between predictions and measurements. Overall, the experiments and computations have provided significant insight into spray in crossflow phenomenon, and have yielded useful results in terms of validated spray trajectory correlations, droplet evaporation lifetimes under forced convective conditions, and a methodology for simulation of airblast sprays.