http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Dahai Wang,Xinzhong Chen,Jie Li,Hao Cheng 한국풍공학회 2016 Wind and Structures, An International Journal (WAS Vol.22 No.6
This paper presents a wind tunnel study of wind loads of the large billboard structures with two-plate and three-plate configurations. Synchronous dynamic pressures on the surfaces of plates are measured, and the characteristics of local pressures, integrated forces on each individual plate and on the overall structures are investigated. The influences of wind direction and plate configuration on wind load characteristics, and the contributions of overall crosswind load and torque to the stress responses are examined. The results showed that the wind load characteristics of windward plate in both two- and three-plate configurations are very similar. The contribution of overall crosswind load makes the total resultant force from both alongwind and crosswind loads less sensitive to wind direction in the case of three-plate configuration. The overall torque is lower than the value specified in current codes and standards, and its contribution is less significant in both two-plate and three-plate configurations.
Dahai Zhou,Vinodh Kannappan,Xiang Chen,Jingqin Li,Xuefeng Leng,Jinping Zhang,Shiying Xuan 생화학분자생물학회 2016 Experimental and molecular medicine Vol.48 No.-
Renal cell carcinoma (RCC), one of the most common kidney cancers, has a poor prognosis. Epithelial to mesenchymal transition (EMT) is a hallmark of carcinoma invasion and metastasis. Several studies have examined the molecular regulation of EMT, but the relationship between histone demethylases and EMT is little understood. In this study, we investigated the role of retinoblastoma-binding protein-2 (RBP2), a histone demethylase that is highly expressed in RCC and is positively correlated with poor RCC prognosis in the regulation of EMT. We found that ectopic overexpression of RBP2 can induce cancer stem cell-like (CSC) phenotypes through EMT in RCC cells by converting them to a more mesenchymal phenotype. This results in increased resistance to apoptosis, which leads to enhanced tumor growth in xenograft models. Together, our data show that RBP2 is an epigenetic regulator that has an important role in the initiation of CSC phenotypes through EMT, leading to tumor progression. RBP2 is also a novel biomolecule for RCC diagnosis, and prognosis and may be a therapeutic target.
Assessment of wind-induced fragility of transmission towers under quasi-static wind load
Dahai Wang,Sen Li,Chao Sun,Guoqing Huang,Qing-shan Yang 한국풍공학회 2021 Wind and Structures, An International Journal (WAS Vol.33 No.4
Overhead power transmission line systems consisting of long-span conductors and high-rise towers are windsensitive structures featured with significant structural nonlinearity and fragility under wind hazards. To assess wind-induced structural fragility of a transmission tower, a novel efficient quasi-static approach, which is based on the analytical probability distribution of extreme wind effect in frequency domain and the probabilistic wind-resistant capacity, is developed in the present study. The 90-degree wind direction (perpendicular to conductors), which is always the worst scenario, is considered in this paper. The structural nonlinearity and failure modes are captured using a nonlinear static push-over analysis method, which simulates the failure process of the tower structure with random initial geometric defects. Wind-resistance performance of the tower is quantified based on the principle of energy equivalence. Damage of the tower is classified into three levels including slight damage, severe damage, and collapse. The tower fragility curve, which predicts damage of the tower as a function of wind speed, is presented and discussed.
A Strain Based Method for Determining the Crack Closure and Initiation Stress in Compression Tests
Dahai Wang,Shaohui He,Dwayne D. Tannant 대한토목학회 2019 KSCE JOURNAL OF CIVIL ENGINEERING Vol.23 No.4
The pre-peak loading stages of rock in compression tests are divided into four stages (i.e., crack closure, elastic deformation, stable crack growth and unstable crack growth) by identifying the Crack Closure stress (CC), Crack Initiation stress (CI), and crack damage stress. A new method for determining the CC and CI is presented in this paper and compared with previous methods. The new method is called “Continuous Strain Deviation” (CSD), and it solves two problems associated with other methods: 1) determining the limits for the elastic range in laboratory data, and 2) identifying where crack closure or initiation occurs from the subtle changes in the stress-strain data. Starting from an initial point corresponding to 30% to 40% UCS, the proposed algorithm provides a distinct indicator for CC and CI. The CC and CI for Badaling granite and Äspö diorite are determined with the proposed method, results from which are similar to other methods. Sensitivity analyses of the CSD method show that stable CC and CI values could be estimated using any initial point from 30% to 40% UCS. Comparison studies show that the CSD method predicts a smaller stress range and gives a more distinct indicator for both CC and CI.
Semiparametric mixture of experts with unspecified gate network
Dahai Jung,Byungtae Seo 한국데이터정보과학회 2017 한국데이터정보과학회지 Vol.28 No.3
The traditional mixture of experts (ME) modeled the gate network using a certain parametric function. However, if the assumed parametric function does not properly reflect the true nature, the prediction strength of ME would become weak. For example, the parametric ME often uses logistic or multinomial logistic models for the network model. However, this could be very misleading if the true nature of the data is quite different from those models. Although, in this case, we may develop more flexible parametric models by extending the model at hand, we will never be free from such misspecification problems. In order to alleviate such weakness of the parametric ME, we propose to use the semi-parametric mixture of experts (SME) in which the gate network is estimated in a non-parametrical way. Based on this, we compared the performance of the SME with those of ME and neural networks via several simulation experiments and real data examples.
Dahai Luo 대한기계학회 2019 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.33 No.5
This paper presents some key features of partially averaged Navier-Stokes (PANS) method in the simulation of turbulent flow over a backward facing step at Re h = 5100. Both fixed and spatially variable values for the unresolved-to-total ratio of kinetic energy (f k ) are adopted in the present PANS simulations. Different model parameters relevant to PANS have been assessed. Detached eddy simulation (DES) is also adopted and the results from DES are compared in detail with PANS and the available experimental data. Variable f k PANS gives overall acceptable predictions but it is slightly less accurate than DES. The defects of PANS simulation with fixed f k throughout the entire computational domain are highlighted in this investigation. An improved step is to introduce variable f k to ensure the near-wall Reynolds averaged Navier-Stokes (RANS) solution. This paper provides encouraging results for PANS simulations of this wall-bounded flow involving separation.
Semiparametric mixture of experts with unspecified gate network
Jung, Dahai,Seo, Byungtae The Korean Data and Information Science Society 2017 한국데이터정보과학회지 Vol.28 No.3
The traditional mixture of experts (ME) modeled the gate network using a certain parametric function. However, if the assumed parametric function does not properly reflect the true nature, the prediction strength of ME would become weak. For example, the parametric ME often uses logistic or multinomial logistic models for the network model. However, this could be very misleading if the true nature of the data is quite different from those models. Although, in this case, we may develop more flexible parametric models by extending the model at hand, we will never be free from such misspecification problems. In order to alleviate such weakness of the parametric ME, we propose to use the semi-parametric mixture of experts (SME) in which the gate network is estimated in a non-parametrical way. Based on this, we compared the performance of the SME with those of ME and neural networks via several simulation experiments and real data examples.