• Journal of Korean Association for Spatial Structures
• /
• v.23 no.2
• /
• pp.29-36
• /
• 2023

Reinforcement learning (RL) is widely applied to various engineering fields. Especially, RL has shown successful performance for control problems, such as vehicles, robotics, and active structural control system. However, little research on application of RL to optimal structural design has conducted to date. In this study, the possibility of application of RL to structural design of reinforced concrete (RC) beam was investigated. The example of RC beam structural design problem introduced in previous study was used for comparative study. Deep q-network (DQN) is a famous RL algorithm presenting good performance in the discrete action space and thus it was used in this study. The action of DQN agent is required to represent design variables of RC beam. However, the number of design variables of RC beam is too many to represent by the action of conventional DQN. To solve this problem, multi-agent DQN was used in this study. For more effective reinforcement learning process, DDQN (Double Q-Learning) that is an advanced version of a conventional DQN was employed. The multi-agent of DDQN was trained for optimal structural design of RC beam to satisfy American Concrete Institute (318) without any hand-labeled dataset. Five agents of DDQN provides actions for beam with, beam depth, main rebar size, number of main rebar, and shear stirrup size, respectively. Five agents of DDQN were trained for 10,000 episodes and the performance of the multi-agent of DDQN was evaluated with 100 test design cases. This study shows that the multi-agent DDQN algorithm can provide successfully structural design results of RC beam.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.28 no.4
• /
• pp.193-203
• /
• 2024

Existing reinforced concrete (RC) building frames constructed before the seismic design was applied have seismically deficient structural details, and buildings with such structural details show brittle behavior that is destroyed early due to low shear performance. Various reinforcement systems, such as fiber-reinforced polymer (FRP) jacketing systems, are being studied to reinforce the seismically deficient RC frames. Due to the step-by-step modeling and interpretation process, existing seismic performance assessment and reinforcement design of buildings consume an enormous amount of workforce and time. Various machine learning (ML) models were developed using input and output datasets for seismic loads and reinforcement details built through the finite element (FE) model developed in previous studies to overcome these shortcomings. To assess the performance of the seismic performance prediction models developed in this study, the mean squared error (MSE), R-square (R²), and residual of each model were compared. Overall, the applied ML was found to rapidly and effectively predict the seismic performance of buildings according to changes in load and reinforcement details without overfitting. In addition, the best-fit model for each seismic performance class was selected by analyzing the performance by class of the ML models.

• Computers and Concrete
• /
• v.2 no.1
• /
• pp.79-96
• /
• 2005

The use of high-strength concrete (HSC) has significantly increased over the last decade, especially in offshore structures, long-span bridges, and tall buildings. The behavior of such concrete is noticeably different from that of normal-strength concrete (NSC) due to its different microstructure and mode of failure. In particular, the shear capacity of structural members made of HSC is a concern and must be carefully evaluated. The shear fracture surface in HSC members is usually trans-granular (propagates across coarse aggregates) and is therefore smoother than that in NSC members, which reduces the effect of shear transfer mechanisms through aggregate interlock across cracks, thus reducing the ultimate shear strength. Current code provisions for shear design are mainly based on experimental results obtained on NSC members having compressive strength of up to 50MPa. The validity of such methods to calculate the shear strength of HSC members is still questionable. In this study, a new approach based on artificial neural networks (ANNs) was used to predict the shear capacity of NSC and HSC beams without shear reinforcement. Shear capacities predicted by the ANN model were compared to those of five other methods commonly used in shear investigations: the ACI method, the CSA simplified method, Response 2000, Eurocode-2, and Zsutty's method. A sensitivity analysis was conducted to evaluate the ability of ANNs to capture the effect of main shear design parameters (concrete compressive strength, amount of longitudinal reinforcement, beam size, and shear span to depth ratio) on the shear capacity of reinforced NSC and HSC beams. It was found that the ANN model outperformed all other considered methods, providing more accurate results of shear capacity, and better capturing the effect of basic shear design parameters. Therefore, it offers an efficient alternative to evaluate the shear capacity of NSC and HSC members without stirrups.

• Journal of Minimally Invasive Surgery
• /
• v.21 no.4
• /
• pp.177-179
• /
• 2018

A rectocele with a weakened rectovaginal septum can be repaired with various surgical techniques. We performed laparoscopic posterior vaginal wall repair and rectovaginal septal reinforcement without mesh using a modified transperineal approach. A 63-year-old woman with outlet dysfunction constipation complained of lower pelvic pressure and sense of heaviness for 30 years. Initial defecography showed an anterior rectocele with a 45-mm anterior bulge and perineal descent. Laparoscopic procedures included peritoneal and rectovaginal septal dissection directed toward the perineal body, rectovaginal septal suturing, and peritoneal closure. The patient started a soft diet the following day and was discharged on the 5th postoperative day without any complications. The patient had no dyschezia or dyspareunia, and no problem with bowel function; 3-month follow-up defecography showed a decrease in bulging to 18 mm. Laparoscopic posterior vaginal wall and rectovaginal septal repair is safe and feasible for treatment of a rectocele, and enables early recovery.

• Advances in concrete construction
• /
• v.13 no.1
• /
• pp.1-10
• /
• 2022

Concrete is a brittle material and weak in tension. Traditionally, web reinforcement in the form of vertical stirrups is used in reinforced concrete (RC) beams to take care of principal stresses that may cause failure when they are subjected to shear stresses. In recent decades, the potential of various types of fibers for improving post-cracking behavior of RC beams and replacing stirrups completely or partially have been studied. It has been shown that the use of steel fibers randomly dispersed and oriented in concrete has a significant potential for enhancing mechanical properties of RC beams. However, the studies on deep steel fiber reinforced concrete (SFRC) beams are limited when compared to those focusing on slender beams. An experimental program consisting of three RC and nine SFRC deep beams without stirrups were conducted in this study. Besides, various models developed for predicting the ultimate shear strength and diagonal cracking strength of SFRC deep beams without stirrups were applied to experimental data obtained from the literature and this study.

• Journal of the Korea Concrete Institute
• /
• v.17 no.5 s.89
• /
• pp.717-726
• /
• 2005

Shear test data have been extracted from previous experimental research and compiled into a database that may be the largest ever made. In this paper, the shear database (SDB) was used for evaluating shear design provisions for both reinforced concrete (RC) beams and prestressd concrete (PSC) beams. A discussion on the use of the results of this evaluation related to calibration and strength reduction factor for the shear design provisions was also provided. It was observed that the shear design provisions did not provide good predictions for RC members and gave very poor predictions especially for RC members without shear reinforcement. On the other hand, the limit on shear strength contributed by transverse reinforcement was observed to be lower than necessary. The shear design provisions gave very unconservative results for the large RC members (d>700mm) without shear reinforcement having light amount of longitudinal reinforcement $(\rho_w<1.0\%)$. However, for PSC members the shear design provisions gave a good estimation of ultimate shear strength with a reasonable margin of safety. Despite of a large difference of accuracy in prediction of shear strength for RC members and PSC members, the shear design provisions used a same shear strength reduction factor for these members. As a result, the shear design provisions did not provide a uniform factor of safety against shear failure for different types of members.

• Journal of the Korea Concrete Institute
• /
• v.15 no.6
• /
• pp.820-828
• /
• 2003

Recent research has indicated that the current ACI shear provision provides unconservative predictions for large slender beams and beams with low level of longitudinal reinforcement, and conservative results for deep beams. To modify some problems of ACI shear provision, ultimate shear strength equation considering size effect and arch action to compute shear strength in high-strength concrete beams without stirrups is presented in this research. Three basic equations, namely size reduction factor, rho factor, and arch action factor, are derived from crack band model of fracture mechanics, analysis of previous some shear equations for longitudinal reinforcement ratio, and concrete strut described as linear prism in strut-tie model deep beams. Constants of basic equations are determined using statistical analysis of previous shear testing data. To verify proposed shear equation for each variable, effective depth, longitudinal reinforcement ratio, concrete compressive strength and shear span-to-depth ratio, about 300 experimental data are used and proposed shear equation is compared with ACI 318-99 code, CEB-FIP Model code, Kim &Park's equation and Zsutty's equation. The proposed shear equation is not only simpler than other shear equations, it is but also shown to be economical predictions and reasonable safety margin. Hence proposed shear strength equation is expected to be applied to practical shear design.

• Magazine of the Korea Concrete Institute
• /
• v.8 no.6
• /
• pp.173-182
• /
• 1996

An experimental research was carried out to investigate the effect of thc compressive strength of concrete on the stirrup effectiveness in shear behavior of concrete beams. For this purpose. total 24 beams of section dimension of $300{\times}600mm$ were tested: 4 specimens without web reinforcement and 20 specimens with web reinforcement in the form of vertical stirrups. Main variables were two levels(norma1 and high strength) of the compressive strength of concrete and six types of t h e shear rcinfor.cement ratios. Prior to experiment, for given sections and assumed material constants, the reference shear reinforcement ratio(${\rho}_vACI$) which leads to the flexure failure using the provisions of the ACI Building Code(AC1 318-95) was calculated. and the shear reinforcement ratios were relatively selected from the value of ${\rho}_vACI$. From test results, it was shown that thc safety factor of ACI eyuation for p1,ediction of shear strength was decreased with increasing the compressive strength of concrete in beams without stirrups. However. it was observed that as the amount of' stirrup is increased, the safety factor for high strength conci,ete beams with high stirrup ratio is ensured more than that for normal strength concrete beams. Therefore i t appears that the stirrup effectiveness of high strength concrete beams is greater than that of normal strength concrete beams.

• Journal of the Korean Geosynthetics Society
• /
• v.12 no.4
• /
• pp.77-86
• /
• 2013

Laboratory pullout tests were conducted to evaluate pullout performance of steel strip reinforcements with deformed steel bars as transverse members. The steel strip reinforcement has an installation hole to assemble a deformed steel bar. Jumunjin standard sand is used to form a relative density of ground model to 80%. Frictional resistance of steel strip reinforcement without transverse member increases sharply at the initial displacement and quickly decreases with displacement. Maximum frictional resistance increases linearly as normal pressure increasing, and soil-reinforcement interaction friction angle(${\rho}_{peak}$) of a steel strip reinforcement is estimated to $14.64^{\circ}$. Passive resistance increases with displacement and converge into maximum passive resistance in most cases. Maximum passive resistance increases linearly as normal pressure increasing irrespective of shape of the steel reinforcement. Pullout force of steel strip reinforcements with installation holes or transverse members largely increases about 4 to 7 times compared to frictional resistance force of steel strip reinforcements when embedment length($L_e$) of steel strip reinforcements is 500 mm. In the case of using 2 transverse members, interference effect is observed due to the spacing of 2 transverse members and location of assembly holes and transverse members.

• Journal of the Korean Geotechnical Society
• /
• v.21 no.10
• /
• pp.5-16
• /
• 2005

Geosynthetic-reinforced and pile-supported embankments have been increasingly used and researched around the world. The inclusion of one or multiple geosynthetic reinforcements over the pile is intended to enhance the efficiency of load transfer from soft ground to piles, to reduce total and differential settlement and increase global or local stability. In this paper, the reinforcement effectiveness and arching effect of the geogrid-reinforced and pile-supported embankments have been studied in terms of field model tests and numerical analysis with varying the space between piles and reinforcement. 2-dimensional numerical analysis has been conducted using the FLAC (Fast Lagrangian Analysis of Continua) program. And load transfer mechanisms between soil-piles-geogrid were investigated. The mechanisms of load transfer can be considered as a combination of embankment soil arching, tension geogrid, and stress concentration due to the stiffness difference between pile and soft ground. Based on the field model test and numerical analysis results, it was found that the geosynthetic reinforcement slightly interferes with soil arching, and helps reduce differential settlement of the soft ground. Also. at the D/b=3 (D: spacing of pile cap, b: diameter of pile), the total settlement is reduced by about $40\%$ compared to that without reinforcement. For $D/b{\ge}6$, the effectiveness of geogrid reinforcement in reducing settlement is negligible.