• Structural Engineering and Mechanics
• /
• v.7 no.3
• /
• pp.319-330
• /
• 1999

The design of composite space trusses is a demanding task that involves taking several decisions on the truss depth, number of panels, member configuration, number of chord layers and concrete slab thickness and grade. The focus in this paper is on the design of top concrete slabs of composite space trusses, and in particular their thickness. Several effects must be considered in the process of designing the slab before an optimum thickness can be chosen. These effects include the inplane forces arising from shear interaction with the steel sub-truss and the flexural. and sheer effects of direct lateral slab loading. They also include a constructional consideration that the thickness must allow for sufficient cover and adequate space for placing the reinforcement. The work presented in this paper shows that the structural requirements on the concrete slab thickness are in many cases insignificant compared with the constructional requirements.

• Electronics and Telecommunications Trends
• /
• v.34 no.4
• /
• pp.32-42
• /
• 2019

The performance of machine learning algorithms significantly depends on how a configuration of hyperparameters is identified and how a neural network architecture is designed. However, this requires expert knowledge of relevant task domains and a prohibitive computation time. To optimize these two processes using minimal effort, many studies have investigated automated machine learning in recent years. This paper reviews the conventional random, grid, and Bayesian methods for hyperparameter optimization (HPO) and addresses its recent approaches, which speeds up the identification of the best set of hyperparameters. We further investigate existing neural architecture search (NAS) techniques based on evolutionary algorithms, reinforcement learning, and gradient derivatives and analyze their theoretical characteristics and performance results. Moreover, future research directions and challenges in HPO and NAS are described.

• Annual Conference of KIPS
• /
• 2022.05a
• /
• pp.648-651
• /
• 2022

Match-3 Game 은 스테이지 구성 및 난이도 설정이 중요한 게임이나 다양한 밸런스 요소로 인해 스테이지 구성에 중요한 요소인 난이도 설정에 많은 시간이 소요된다. 특히 게임을 플레이하는 유저가 재미를 느끼는 수준으로 난이도를 설정하는 것이 중요하며, 이를 자동화하기 위해 실제 유저의 플레이 데이터를 활용하여 사람과 유사한 수준의 자동 플레이 에이전트 개발이 진행되었다. 하지만 플레이 데이터의 확보는 쉽지 않기에 연구 방향은 플레이 데이터가 없는 강화학습으로 확장되고 있다. 스테이지 구성에 중요한 요소인 난이도를 설정하기 위함이라면 각 스테이지 간의 상대적인 난이도 차이를 파악하는 것으로 가능하다. 이를 위해 게임의 규칙을 학습한 강화학습 에이전트로 밸런스 요소의 변화에 따른 다양한 난이도의 스테이지를 50 회씩 플레이하여, 평균 획득 점수를 기준으로 스테이지 구성에 필요한 각 스테이지들의 난이도를 파악할 수 있었다.

• Composites Research
• /
• v.32 no.1
• /
• pp.37-44
• /
• 2019

The reinforcement effect of micro-bolt for a bonded scarf joint was investigated. Three scarf ratios of 1/10, 1/20, and 1/30 were considered to examine the effect of scarf patch configuration on joint strength. To maintain the same density of micro-bolt, 16, 32, and 48 bolts were installed in the scarf joint specimens with scarf ratios of 1/10, 1/20, and 1/30, respectively. Tests were also carried out on the joints that are bonded with only adhesive and that are fastened with only micro-bolts to obtain reference values. The average failure loads of the adhesive joints with scarf ratios of 1/10, 1/20, and 1/30 were 29.7, 39.6, and 44.8 kN, respectively. In case of micro-bolt reinforcement, the failure loads at the same scarf ratios were 28.4, 37.2, and 40.1 kN, respectively, which corresponds to 96, 94, and 90% of the pure adhesive joint failure loads. In the case of using only micro-bolts, the failure loads were only 13-25% of the average failure loads of pure adhesive joints. Fatigue test was also conducted for the joints with scarf ratio of 1/10. The results show that the fatigue strength of hybrid joints using both adhesive and microbolts together slightly increased compared to the fatigue strength of adhesive joint, but the rate of increase was small to 2-3%. Through this study, it was confirmed that the reinforcement effect of micro-bolt is negligible in the scarf joints where shear stress is dominating the failure, unlike in the structure where peel stress is dominant.

• Steel and Composite Structures
• /
• v.37 no.1
• /
• pp.15-26
• /
• 2020

The top-and-seat angles with double web angles are commonly used in the design of beam-to-column joints in Asian and North American countries. The seismic behavior analysis of these joints with large cross-section size of beam and column (often connected by four or more bolts) is a challenge due to the effects from the relatively larger size of stiffened angles and the composite action from the adjacent concrete slab. This paper presents an experimental investigation on the seismic performance of exterior composite beam-to-column joints with stiffened angles under cyclic loading. Four full-scale composite joints with different configuration (only one specimen contain top angle in concrete slab) were designed and tested. The joint specimens were designed by considering the effects of top angles, longitudinal reinforcement bars and arrangement of bolts. The behavior of the joints was carefully investigated, in terms of the failure modes, slippage, backbone curves, strength degradation, and energy dissipation abilities. It was found that the slippage between top-and-seat angles and beam flange, web angle and beam web led to a notable pinching effect, in addition, the ability of the energy dissipation was significantly reduced. The effect of anchored beams on the behavior of the joints was limited due to premature failure in concrete, the concrete slab that closes to the column flange and upper flange of beam plays an significant role when the joint subjected to the sagging moment. It is demonstrated that the ductility of the joints was significantly improved by the staggered bolts and welded longitudinal reinforcement bars.

• Steel and Composite Structures
• /
• v.38 no.2
• /
• pp.223-239
• /
• 2021

This paper presents the analytical modeling and finite element (FE) analysis, using ABAQUS software, of the new types of steel reinforced lightweight aggregate concrete (SRLAC) columns with cross-shaped (+shaped and X-shaped) steel section, using proposed three analytical and two FE models in total. The stress-strain material models for different components in the columns, including the confined zones of the lightweight aggregate concrete (LWAC) using three and four concrete zones divisions approaches and with and without taking into account the stirrups reaction effect, are established first. The analytical models for determining the axial load-deformation behavior of the SRLAC columns are drawn based on the materials models. The analytical and FE models' results are compared with previously reported test results of the axially loaded SRLAC columns. The proposed analytical and FE models accurately predict the axial behavior and capacities of the new types of SRLAC columns with acceptable agreements for the load-displacement curves. The LWAC strength, steel section ratio, and steel section configuration affect the contact stress between the concrete and steel sections. The average ratios of the ultimate test load to the three analytical models and FEA model loads, Put /Pa1, Put /Pa2, Put /Pa3, and Put /PFE1, for the tested specimens are 0.96, 1.004, 1.016, and 1.019, respectively. Finally, the analytical parametric studies are also studied, in terms of the effects of confinement, LWAC strength, steel section ratio, and the reinforcement ratio on the axial capacity of the SRLAC column. When concrete strength, confinements, area of steel sections, or reinforcement bars ratio increased, the axial capacities increased.

• Smart Structures and Systems
• /
• v.33 no.4
• /
• pp.291-300
• /
• 2024

Uncertainty of the model, system delay and drive dynamics can be considered as normal uncertainties, and the main source of uncertainty in the seismic control system is related to the nature of the simulated seismic error. In this case, optimizing the management strategy for one particular seismic record will not yield the best results for another. In this article, we propose a framework for online management of active structural management systems with seismic uncertainty. For this purpose, the concept of reinforcement learning is used for online optimization of active crowd management software. The controller consists of a differential controller, an unplanned gain ratio, the gain of which is enhanced using an online reinforcement learning algorithm. In addition, the proposed controller includes a dynamic status forecaster to solve the delay problem. To evaluate the performance of the proposed controllers, thousands of ground motion data sets were processed and grouped according to their spectrum using fuzzy clustering techniques with spatial hazard estimation. Finally, the controller is implemented in a laboratory scale configuration and its operation is simulated on a vibration table using cluster location and some actual seismic data. The test results show that the proposed controller effectively withstands strong seismic interference with delay. The goals of this paper are towards access to adequate, safe and affordable housing and basic services, promotion of inclusive and sustainable urbanization and participation, implementation of sustainable and disaster-resilient buildings, sustainable human settlement planning and manage. Simulation results is believed to achieved in the near future by the ongoing development of AI and control theory.

• Journal of the Korean Society of Safety
• /
• v.35 no.1
• /
• pp.53-61
• /
• 2020

This study proposes a new damper configuration for seismic performance improvement of heavy sliding facilities inside a building. For this purpose, we deal with two connection types of control system, and the parametric study has been performed to investigate their comparative seismic performances according to the variations of the control capacity. In order to simulate the seismic responses of the proposed system, we employed a recently-developed seismic response analysis method that can deal with the two-mass system with nonlinear frictional sliding behavior. The numerical results demonstrate that the typical method of diagonal bracing damper connection can exhibit effective control performance both on structure and the heavy sliding facilities, whereas the structure-facilities connection method does not show any control effect on both responses. On the other hand, the typical method has some limitations that it can adversely cause excessive sliding of the facilities, depending upon the frequency characteristics of structure and earthquake. On the contrary, the structure-facilities connection method is very effective in reducing the sliding displacement of the heavy facilities, even with small amount of control capacity. Thus, the following potential expectations can be inferred from these results: The typical diagonal bracing damper connection method will have some promising benefits in controlling the sliding facilities inside the building as well as the building itself, and the structure-facilities connection method can be a cost-effective way of protecting the internal heavy important facilities inside the structure already designed with sufficient seismic performance.

• Journal of the Korean Geotechnical Society
• /
• v.21 no.3
• /
• pp.65-77
• /
• 2005

This paper investigates the failure mechanism of geosynthetic-reinforced segmental retaining walls with tiered configuration using reduced-scale model tests. The reduced scale model test set-up was established to simulate a 5 m high full-scale wall. The geometry and material properties used in the model test were determined based on the Similitude Laws. The wall failures in the model tests were successfully generated by their self weight without any surface loading and analyzed examining the digital video recordings. The failure mechanisms was examined with respect to the various offsets between the lower and upper teres and the reinforcement length. Based on the results the appropriateness of the current design guideline was discussed.

• Journal of the Korea Convergence Society
• /
• v.6 no.4
• /
• pp.9-14
• /
• 2015

This paper aims at solving the bending problem by the applied force due to the configuration of the neck part of guitar. At reinforcing and using the existing guitar, the method to understand the area vulnerable to the load was studied. In this study, the material property of wood applied to the practical guitar was applied and the finite element analysis was carried out after the modelling. By using the result through the study of this paper, it is thought that the foundation of material about bending and damage which has been recognized as the problem of existed guitar can be obtained. This study aims at improving the guitar as the design to reinforce it is studied. The bending and the damage are prevented and the durability can be improved by applying to the practical design on the basis of the result of the improved study model. And it is possible to be grafted onto the convergence technique at design and show the esthetic sense.