• Journal of the Korean Geotechnical Society
• /
• v.34 no.10
• /
• pp.61-74
• /
• 2018

In this study, the pull-out behavior of tunnel type anchorage of suspension bridges was analyzed based on results from laboratory size model tests and numerical analysis. Tunnel type anchorage has found its applications occasionally in both domestic and oversea projects, therefore design method including failure mode and safety factor is yet to be clearly established. In an attempt to improve the design method, scaled model tests were conducted by employing simplified shapes and structure of the Ulsan grand bridge's anchorage which was the first case history of its like in Korea. In the model tests, the anchorage body and the surrounding rocks were made by using gypsum mixture. The pull-out behavior was investigated under plane strain conditions. The results of the model tests showed that the tunnel type anchorage underwent wedge shape failure. For the verification of the model tests, numerical analysis was carried out using ABAQUS, a finite element analysis program. The failure behavior predicted by numerical analysis was consistent with that by the model tests. The result of numerical analysis also showed that the effect of Poisson's ratio was negligible, and that a plugging type failure mode could occur only when the strength of the surrounding rocks was 10 times larger than that of anchorage body.

• Korean journal of applied entomology
• /
• v.24 no.1 s.62
• /
• pp.29-33
• /
• 1985

Every plant in $990m^2$ onion field was inspected for damages by the onion maggot. Maps were constructed every ten days to show which plants were infested and which were not from April 11 to May 21, 1984. The maps were sectioned into squares one of which contains 80 onion plants and the counts of damaged onions in each square were fitted to poisson and negative binomial distribution and tested by chi-square. We argue that the satisfactory fitness of the expected negative binomial $[P(x^2)>0.05]$ provided a useful description of the spatial distribution patterns of the damaged onions. Edge effect was tested by the differences of damage ratio and variance/mean ratio (${\sigma}^2/m$) between edge and center part. The result showed that the damage ratioes and variances of all the periods, ${\sigma}^2/m$ values after May 1 were greater in edge part than in center part. Again, the maps were sectioned into four blocks and the squares (sample units) were sectioned into quadrants. By application of the variance component technique, it was suggested that $2{\sim}8$ sample units for 5% sampling error and $1{\sim}2$ sample units for 10% error should be sampled randomly to estimate the damage ratio when $2{\sim}3$ quadrants were inspected.

• Journal of the Korea Concrete Institute
• /
• v.29 no.2
• /
• pp.149-157
• /
• 2017

As the amount of construction wastes increase, reuse of recycled materials is being considered in research areas. While there are many experimental investigations focusing on development of mortar and concrete using the recycled materials, the studies regarding the fiber-reinforced cement composites (FRCCs) using recycled materials are still limited. In this paper, an experimental attempt has been made to investigate the effect of recycled fine aggregates and fly ash on the mechanical properties of PVA FRCCs. The cement and natural sand were respectively replaced by fly ash and recycled fine aggregates at two content levels, 25% and 50%. Ten types of PVA FRCCs mixes were fabricated and tested to investigate the flexural, compressive and direct tensile behaviors. The test results show that flexural, compressive and direct tensile strength were decreased with increase in fly ash content at all ages. In particular, flexural, compressive and direct tensile strengths of specimens, containing 50% recycled fine aggregates and 50% fly ash, showed the lowest values. The modulus of elasticity of specimens showed similar trend to the 28-day compressive strength. Poisson's ratio was increased with increase in fly ash and recycled fine aggregates content.

• Journal of the Korean Ceramic Society
• /
• v.50 no.3
• /
• pp.206-211
• /
• 2013

The effect of a $WO_3$ or $Ga_2O_3$ addition on the densification, phase evolution, optical reflectance, and elastic and dielectric properties of $Y_2O_3$-doped AlN ceramics sintered at $1800^{\circ}C$ for 3 h is investigated. The investigated compositions of the additives are 4.5 wt% $Y_2O_3$ (YA), 3.5 wt% $Y_2O_3$-1.0 wt% $Ga_2O_3$ (YGA), and 3.5 wt% $Y_2O_3$-1.0 wt% $WO_3$ (YWA). $YAlO_3$ and $Y_4Al_2O_9$ form as the secondary phases in all of the investigated compositions, whereas $W_2B$ appears additionally in the YWA. In the YGA, Ga is detected in the AlN grains, indicating that the dissolution of $Ga_2O_3$ into the AlN lattice occurs. The addition of $WO_3$ blackens the specimen more significantly than that of $Ga_2O_3$ does. In all of the investigated specimens, the linear shrinkage and the apparent density are above 20 percent and in the range of 3.34-3.37 $g/cm^3$, respectively. The elastic modulus, Poisson's ratio, the dielectric constant, and the dielectric loss are in the ranges of 335-368 GPa, 0.146-0.237, 8.60-8.63, $2.65-3.95{\times}10^{-3}$, respectively. The sinterability and the properties of $Y_2O_3$-doped AlN ceramics are not much altered by the addition of $WO_3$ or $Ga_2O_3$.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.37 no.9
• /
• pp.15-22
• /
• 2000

The effects of carrier transport and input power on the extinction ratio was theoretically analyzed in a 1.55${\mu}m$ InGaAsP/InGaAsP multiple-quantum-well(MQW) electroabsorption(EA) modulator. Poisson's equation, current continuity equations for electrons and holes, and optical field distribution were self-consistently solved by considering electric field dependent absorption coefficients. The field screening effect due to the carrier accumulation in heterointerface and the space-charge region occurred more seriously at the input side of modulator as input optical intensity increased. It was revealed that extinction ratio could be steeply degraded for modulator with the length of 200${\mu}m$ when an input power exceeds 10mW. A degradation of extinction ratio due to the field screening effect would be more significantly at high-performance devices such as a 1.55${\mu}m$DFB-LD/EA-modulator integrated source where optical coupling efficiency is almost complete or a very high-speed modulator with its length as short as a few tens ${\mu}m$.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.19 no.4 s.74
• /
• pp.389-398
• /
• 2006

Structural optimization design has been developed with finite element analysis using effective and fast computational technology. Especially topology optimization design has been recently often used since it yields an optimal topology as well as an optimal shape under satisfied constraints. In general in finite element analysis, it is assumed that the structural material properties such as Young's modulus and Poisson's ratio and the variable of applied loading are fixed with obvious values in structure. However practically these values may take uncertainties because of environmental effect or manufactural error of structures. Therefore static or dynamic analysis of the structures may make an error, then finally it may have an influence on qualify of optimal design. In this study, the topology optimization design of structure is carried out using so called the interval finite element method, and the analysis method Is proposed. The results are also validated by comparing with conventional topology optimization results of density distribution method and finite element analysis results. The present method can be used to predict the optimal topology of linear elastostatic structures with respect to structural uncertainty of behavior.

• Tunnel and Underground Space
• /
• v.34 no.5
• /
• pp.421-432
• /
• 2024

Pre-fracturing, a type of pre-conditioning blast, is a method used to weaken rock masses prior to mechanical excavation. In this study, various laboratory tests were conducted on rock core samples obtained from the field to verify the effectiveness of pre-fracturing in controlling the BIDZ (Blast-Induced Damaged Zone) by measuring the physical properties of rock cores obtained from a test site and assessing changes in these properties. In the EDZ (Excavation Damaged Zone) caused by combined excavation using blasting and mechanical excavation, the effect of blasting is generally more significant than that of mechanical excavation, so BIDZ control directly leads to EDZ control. In terms of Poisson's ratio, elastic wave velocity, porosity, density, thermal conductivity, tensile strength and hydraulic conductivity, the BIDZ size in pre-fracturing was smaller than that of conventional blasting. However, no clear reduction in the BIDZ was observed in the case of Young's modulus and uniaxial compressive strength. By applying a theoretical formula predicting the range of tensile cracks caused by blasting and comparing it with the laboratory results, the BIDZ was reasonably predicted. Nonetheless, limitations in accurately predicting BIDZ size were identified due to assumptions regarding ground conditions and charge density in the formula.

• The Korean Journal of Nuclear Medicine
• /
• v.31 no.3
• /
• pp.330-338
• /
• 1997

This study investigated the method to adjust acquisition time(a) and injection dose (i) to make the best basal and subtraction images in consecutive SPECT. Image quality was assumed to be mainly affected by signal to noise ratio(S/N). Basal image was subtracted from the second image consecutively acquired at the same position. We calculated S/N ratio in basal SPECT images($S_1/N_1$) and subtraction SPECT images(Ss/Ns) to find a(time) and i(dose) to maximize S/N of both images at the same time. From phantom images, we drew the relation of image counts and a(time) and i(dose) in our system using fanbeam-high-resolution collimated triple head SPECT. Noise by imaging process depended on Poisson distribution. We took maximum tolerable duration of consecutive acquisition as 30 minutes and maximum injectible dose as 1,850MBq(50 mCi)(sum of two injections) per study. Counts of second-acquired image($S_2$), counts($S_s$) and noise($N_s$) of subtraction SPECT were as follows. $C_1$ was the coefficient of measurement with our system. $$S_2=S_1{\cdot}(\frac{30-a}{a})+background{\cdot}(1-\frac{30-a}{a})+C_1{\cdot}(30-a){\cdot}{\epsilon}{\cdot}(50-i)$$ $$Ss=S_2-\{S_1{\cdot}(\frac{30-a}{a})+background{\cdot}(1-\frac{(30-a)}{a})\}$$ $$Ns={\sqrt{N_2^2+N_1^2{\cdot}\frac{(30-a)^2}{a^2}}={\sqrt{S_2+S_1{\cdot}\frac{(30-a)^2}{a^2}}$$ In case of rest/acetazolamide study, effect(${\epsilon}$) of acetazolamide to increase global brain uptake of Tc-99m-HMPAO could be 1.5 or less. Varying ${\epsilon}$ from 1 to 1.5, a(time) and i(dose) pair to maximize both $S_1/N_l$ and Ss/Ns was determined. 15 mCi/17 min and 35mCi/13min was the best a(time) and i(dose) pair for rest/acetazolamide study(when ${\epsilon}$ were 1.2) and came to be used for our clinical routine after this study. We developed simple method to maximize S/N ratios of basal and subtraction SPECT from consecutive acquisition. This method could be applied to ECD/HMPAO and brain activation studies as well as rest/acetazolamide studies.

• Computational Structural Engineering
• /
• v.8 no.4
• /
• pp.87-97
• /
• 1995

For the same configuration of two-dimensional finite element models, 6-node element exhibits stiffer bending stiffness than 8-node element. This is true in the relation between 16-node element and 20-node element for three-dimensional model. This stiffening phenomenon comes from the elimination of several mid nodes from full-node elements. Therefore, this may be called 'relative stiffness stiffening phenomenon'. It seems that there are a couple of ways to correct the stiffening effect, however, we could find only one effective method-the method of modification of Gauss sampling points-which passes the patch test and does not alter other kinds of stiffness, such as extensional stiffness. The quantity of modification is a function of Poisson's ratios of the constituent materials. We could obtain two modification equations, one for plane stress case and the other for plane strain case. This method can be extended to 3-dimensional solid elements. Except the exact plane strain cases, most 3-dimensional plates could be modeled successfully with 16-node element modified by the equation for the plane stress case. The effectiveness of the modification method is checked by applying it to several examples with excellent improvements. In numerical examples, beams with various boundary conditions are subjected to static and time-dependent loads. Free and forced motion analyses of beams and plates are also tested. The beam and plate may be composed of isotropic multilayers as well as a single layer.

• Journal of KIISE:Software and Applications
• /
• v.29 no.4
• /
• pp.270-277
• /
• 2002

Text filtering is a task of deciding whether a document has relevance to a specified topic. As Internet and Web becomes wide-spread and the number of documents delivered by e-mail explosively grows the importance of text filtering increases as well. The aim of this paper is to improve the accuracy of text filtering systems by using machine learning techniques. We apply AdaBoost algorithms to the filtering task. An AdaBoost algorithm generates and combines a series of simple hypotheses. Each of the hypotheses decides the relevance of a document to a topic on the basis of whether or not the document includes a certain word. We begin with an existing AdaBoost algorithm which uses weak hypotheses with their output of 1 or -1. Then we extend the algorithm to use weak hypotheses with real-valued outputs which was proposed recently to improve error reduction rates and final filtering performance. Next, we attempt to achieve further improvement in the AdaBoost's performance by first setting weights randomly according to the continuous Poisson distribution, executing AdaBoost, repeating these steps several times, and then combining all the hypotheses learned. This has the effect of mitigating the ovefitting problem which may occur when learning from a small number of data. Experiments have been performed on the real document collections used in TREC-8, a well-established text retrieval contest. This dataset includes Financial Times articles from 1992 to 1994. The experimental results show that AdaBoost with real-valued hypotheses outperforms AdaBoost with binary-valued hypotheses, and that AdaBoost iterated with random weights further improves filtering accuracy. Comparison results of all the participants of the TREC-8 filtering task are also provided.