• Bulletin of the Korean Chemical Society
• /
• 제34권3호
• /
• pp.743-748
• /
• 2013

The temperature dependence of the partition coefficients of a neuropeptide, substance P (SP), in isotropic acidic bicelles was investigated by using a pulsed field gradient nuclear magnetic resonance diffusion technique. The addition of negatively charged dimyristoylphosphatidylserine to the neutral bicelle changed the SP partitioning a little, which implies that the hydrophobic interaction between the hydrophobic residues of SP and the acyl chains of lipid molecules is the major interaction while the electrostatic interaction is minor in SP binding in a lipid membrane. From the temperature dependence of the partition coefficients, thermodynamic functions were calculated. The partitioning of SP into the acidic bicelles is enthalpy-driven, as it is for small unilamellar vesicles and dodecylphosphocholine micelles, while peptide partitioning into a large unilamellar vesicle is entropy-driven. This may mean that the size of lipid membranes is a more important factor for peptide binding than the surface curvature and surface charge density.

• 대한원격탐사학회지
• /
• 제32권5호
• /
• pp.471-482
• /
• 2016

본 논문에서는 수계 영역의 감독 분류 성능을 향상시키기 위하여 블록 기반의 영상 분할과 수계 경계의 확장을 이용하는 수계 검출 방법을 제안한다. 초기 수계 영역을 추출하기 위하여 수계 훈련 지역의 Normalized Difference Water Index (NDWI) 및 Near Infrared (NIR) 밴드 영상의 분광 정보를 이용하여 Mahalanobis 거리 영상을 생성한다. Mahalanobis 거리 영상에 포함된 잡음 성분의 영향을 감소시키기 위해서 인접한 화소의 연결 강도에 의해 확산 계수가 제어되는 평균 곡률 확산을 적용한 후에 초기 수계 영역을 추출한다. 추출된 수계 영상을 같은 크기의 블록으로 분할한 후에 수계 경계에 속하는 수계 영역의 정보를 이용하여 수계 영역을 갱신한다. 수계 경계에 속하는 수계 영역과 수계 훈련 지역 사이의 통계적인 거리가 임계값 이하이면, 수계 영역 갱신을 반복적으로 수행한다. 제안한 알고리즘을 KOMPSAT-2 영상에 적용한 결과 블록 크기가 $11{\times}11$에서 $19{\times}19$사이인 경우에 overall accuracy는 99.47%에서 99.53%, Kappa coefficient는 95.07%에서 95.80%의 분류 정확도를 보였다.

• Nuclear Engineering and Technology
• /
• 제55권4호
• /
• pp.1400-1409
• /
• 2023

In this study, three condensation models of the CUPID code, i.e., the resolved boundary layer approach (RBLA), heat and mass transfer analogy (HMTA) model, and an empirical correlation, were tested and validated against the COPAIN and CAU tests. An improvement on HMTA model was also made to use well-known heat transfer correlations and to take geometrical effect into consideration. The RBLA was a best option for simulating the COPAIN test, having mean relative error (MRE) about 0.072, followed by the modified HMTA model (MRE about 0.18). On the other hand, benchmark against CAU test (under natural convection and occurred on a slender tube) indicated that the modified HMTA model had better accuracy (MRE about 0.149) than the RBLA (MRE about 0.314). The HMTA model with wall function and the empirical correlation underestimated significantly, having MRE about 0.787 and 0.55 respectively. When using the HMTA model, consideration of geometrical effect such as tube curvature was essential; ignoring such effect leads to significant underestimation. The HMTA and the empirical correlation required significantly less computational resources than the RBLA model. Considering that the HMTA model was reasonable accurate, it may be preferable for large-scale simulations of containment.