• Membrane and Water Treatment
• /
• v.8 no.2
• /
• pp.113-123
• /
• 2017

This study investigated effects of physical and chemical cleaning methods on the initial flux recovery of fouled membrane in membrane distillation process. A laboratory scale direct contact membrane distillation (DCMD) experiment was performed to treat digested livestock wastewater with 3.89 mg/L suspended solids, 874.7 mg/L COD, 543.7 mg/L nitrogen, 15.6 mg/L total phosphorus, and pH of 8.6. A hydrophobic PVDF membrane with an average pore size of $0.22{\mu}m$ and a porosity of 75 % was installed inside a direct contact type membrane distillation module. The temperature difference between feed and permeate side was maintained at $40^{\circ}C$ with the feed and permeate stream velocity of 0.18 m/s. The results showed that the permeate flux decreased from $22.1L{\cdot}m^{-2}{\cdot}hr^{-1}$ to $19.0L{\cdot}m^{-2}{\cdot}hr^{-1}$ after 75 hours of distillation. The fouled membrane was cleaned first by physical flushing and consecutively by chemicals with NaOCl and citric acid. After the physical cleaning the flux was recovered to 92 % as compared with the initial clean water flux of the virgin membrane. Then 94 % of the flux was recovered after cleaning by 2,000 ppm NaOCl for 90 minutes and finally 97 % of flux recovered after 3 % citric acid for 90 minutes. SEM-EDS and FT-IR analysis results presented that the foulants on the membrane surface were removed effectively after each cleaning step. The contact angle measurement showed that the hydrophobicity of the membrane surface was also restored gradually after each cleaning step to reach nearly the same hydrophobicity level as the virgin membrane.

• Proceedings of the Korean Radioactive Waste Society Conference
• /
• 2017.05a
• /
• pp.83-83
• /
• 2017

Electrorefining is a key step in pyroprocessing. The electrorefining process is generally composed of two recovery steps - the deposit of uranium onto a solid cathode and the recovery of the remaining uranium and TRU elements simultaneously by a liquid cadmium cathode. The solid cathode processing is necessary to separate the salt from the cathode since the uranium deposit in a solid cathode contains electrolyte salt. Distillation process was employed for the cathode processing. It is very important to increase the throughput of the salt separation system due to the high uranium content of spent nuclear fuel and high salt fraction of uranium dendrites. In this study, a mesh-covered crucible was investigated for the sat distillation of electrorefiner uranium deposits. A liquid salt separation step and a vacuum distillation step were combined for salt separation. The adhered salt in uranium deposits was efficiently removed in the mesh-covered crucible. The salt distiller was operated simply since repeated cooling - heating step was not necessary for the change of the crucible. The operation time could be reduced by the use of the mesh-covered crucible and the combined operation of the two steps. A method to preserve a vacuum level was proposed by double O-rings during the operation of the distiller with the mesh-covered crucible. After the salt distillation, the salt content was measured and was below 0.1wt% after the salt distillation. The residual salt after the salt distillation can be removed further during melting of uranium metal.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.4
• /
• pp.31-39
• /
• 2023

In this paper, we propose an Internal/External Knowledge Distillation (IEKD), which utilizes both external correlations between feature maps of heterogeneous models and internal correlations between feature maps of the same model for transferring knowledge from a teacher model to a student model. To achieve this, we transform feature maps into a sequence format and extract new feature maps suitable for knowledge distillation by considering internal and external correlations through a transformer. We can learn both internal and external correlations by distilling the extracted feature maps and improve the accuracy of the student model by utilizing the extracted feature maps with feature matching. To demonstrate the effectiveness of our proposed knowledge distillation method, we achieved 76.23% Top-1 image classification accuracy on the CIFAR-100 dataset with the "ResNet-32×4/VGG-8" teacher and student combination and outperformed the state-of-the-art KD methods.

• Journal of the Korea Convergence Society
• /
• v.12 no.9
• /
• pp.185-191
• /
• 2021

In our study, we evaluated possibility of distillation process introduction for Korean Rubus coreanus balsamic vinegar production development, has a less complicated and more rapid process leading to higher production volumes and cost savings. In order to investigate the applicability of concentrated vinegar fraction (CVF) and distilled vinegar fraction (DVF) obtained from the distillation process as balsamic and distillation vinegar, acidity, pH, and specific gravity (S.G.) were measured for the vinegar fraction obtained from the distillation process. The acidity of final CVF (60% of initial volume) obtained 4 types of vinegar (acidity; 4.42, 4.84, 5.61, and 6.35) were increased mean 38.1%, compared to initial acidities. Especially, acidity and pH of the final CVF was similar to balsamic vinegar. In addition, acidity of DVF also increased and it of the final fraction was 4.41%, 4.95%, 5.88%, and 6.72%, respectively. And it was suggested that the final DVF could be used as distilled vinegar. In this study, it was confirmed that vinegar can be effectively concentrated through the distillation process, and distilled vinegar can be produced at the same time. However, it is thought that the introduction of the distillation process can be decided only after additional biochemical characteristics such as organic acid content and antioxidant activity measurement for each vinegar fraction are analyzed.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.10
• /
• pp.29-41
• /
• 2022

Recently, with the development of deep learning technology, a variety of huge models with excellent performance have been devised by pre-training massive amounts of text data. However, in order for such a model to be applied to real-life services, the inference speed must be fast and the amount of computation must be low, so the technology for model compression is attracting attention. Knowledge distillation, a representative model compression, is attracting attention as it can be used in a variety of ways as a method of transferring the knowledge already learned by the teacher model to a relatively small-sized student model. However, knowledge distillation has a limitation in that it is difficult to solve problems with low similarity to previously learned data because only knowledge necessary for solving a given problem is learned in a teacher model and knowledge distillation to a student model is performed from the same point of view. Therefore, we propose a heterogeneous knowledge distillation method in which the teacher model learns a higher-level concept rather than the knowledge required for the task that the student model needs to solve, and the teacher model distills this knowledge to the student model. In addition, through classification experiments on about 18,000 documents, we confirmed that the heterogeneous knowledge distillation method showed superior performance in all aspects of learning efficiency and accuracy compared to the traditional knowledge distillation.

• Proceedings of the Membrane Society of Korea Conference
• /
• 2008.05a
• /
• pp.238-241
• /
• 2008

• Korean Chemical Engineering Research
• /
• v.46 no.2
• /
• pp.348-355
• /
• 2008

Recently, an understanding of new sources of liquid hydrocarbons such as bio-ethanol is economically very important. The present dissertation is also designed with purpose of developing the energy-saving process for the separation of bio-ethanol. In order to illustrate the predictability of proposed process for the separation of bio-ethanol, the experimental data from literatures and real plant data are used. Application of the thermodynamics of multicomponent mixtures and phase equilibria to the extractive distillation process with syntheses of heat exchanger network has enabled the development of energy-saving process for different separating agents. Developed process is capable of minimizing the energy usage and the environmental effect. This extractive process is also able to properly describe the effect of impurities, the choice of separating agent. Simulation results of extractive distillation using ethylene glycol show that impurities do not affect to extractive distillation operation and agent, ethylene glycol, was recycled without any loss. It is possible that extraction distillation has various heat network for anhydride ethanol and recovery of ethanol is maximized. Ethylene glycol as separating agent has a high boiling point to eliminate azeotropic point and on the contrary solubility of agent is low to be almost completed recovered. Proposed process is also the energy efficient process configuration in which 99.85mole% anhydride ethanol can be produced with low energy of 1.37198 (kg steam/kg anhydride ethanol).

• Korean Journal of Food Science and Technology
• /
• v.35 no.6
• /
• pp.1028-1032
• /
• 2003

To assess accurate methods for measuring sulfur dioxide residue in pickles, the acid distillation HPLC-UV method and Monnier Willams modified method were examined. By the acid distillation HPLC-UV method, sulfites released from pickles by acid distillation were absorbed in 1% triethanolamine solution and detected as sulfite ion by HPLC with UV monitoring at 240nm. An anion exchange column was employed with 1.8mM $Ma_2CO_3-1.7mM\;NaHCO_3$ solution as a mobile phase, $84.0{\sim}91.7%$ of sulfite added to pickled radish were recovered. Total sulfite levels from 48 kinds of pickles analyzed by acid distillation HPLC-UV was compared with those analyzed by the Monnier Williams modified method. The Monnier Williams modified method showed higher levels of sulfur dioxide than the acid distillation HPLC-UV method due to the presence of volatile acids in pickles. The concentration of sulfur dioxide was in the range of $N.D{\sim}173.05ppm$ in pickled radish and over 30ppm of sulfur dioxide from 3 samples by the acid distillation-HPLC-UV method.

• Resources Recycling
• /
• v.11 no.6
• /
• pp.38-46
• /
• 2002

Fundamental investigations were conducted far the recovery process of waste organic solvent by fractional distillation in the aspects of equilibrium and kinetics. Mixture of toluene and xylene, which were both being used in the largest amount as industrial organic solvent, was taken as the artificial waste organic solvent and their distillation behaviors were studied. The purity of recovered solvent was investigated by Cir Chromatography and shown to be in the range of 94～98%. Based upon equilibrium calculations, the changes in the Gibbs free energy, standard enthalpy, and standard entropy for distillation reaction have been estimated. The standard enthalpy changes for toluene and xylene were shown to be 44.833 and 47.044 kJ $mol^{-1}$ respectively, which were similar to their molar heats of evaporation. The activation energies of distillation fur toluene and xylene obtained from kinetic studies were 3.281 and 2.699 kJ $mol^{-1}$ and they were about one tenths of the standard enthalpy changes of distillation reaction. The highness of the purity of recovered organic solvents suggested the possibility that the recovered waste organic sol-vent could partly replace the original solvent.

• Korean Chemical Engineering Research
• /
• v.46 no.4
• /
• pp.739-746
• /
• 2008

The aim of this study is finding the optimal design parameters and the optimal operation variables of a reactive distillation column. Different from steady state optimization, dynamic optimization makes it possible considering operation ability as well as design problems at process design step. For performing dynamic optimization, dynamic simulation should be done first. If dynamic simulation is already finished, dynamic optimization can be performed with less effort than that of dynamic simulation.Reactive distillation systems involving reaction and separation in a single unit have the potential to reduce capital and operating costs, particularly when reaction have conversion constraint or when azeotropes exist making conventional separation difficult and expensive. This study here present work on the continuous distillation process, the homogeneous catalyzed esterification of methanol and acetic acid, the synthesis of methyl acetate. Based on an equilibrium stage model of a reactive distillation column a dynamic optimization problem was formulated and solved. And the results were verified by performing dynamic simulation and showing the variation of conversion and purity as the variation of the operation variables. As the results of dynamic optimization, this study found optimal feed ratio, reflux ratio and reboiler duty of this system. And as this study applied it to dynamic simulations the dynamic characteristics of a reactive distillation column are showed under optimal operating condition.