• Geomechanics and Engineering
• /
• v.28 no.3
• /
• pp.313-324
• /
• 2022

Rock properties are important in the design of mines and civil engineering excavations to prevent the imminent failure of slopes and collapse of underground excavations. However, the time, cost, and expertise required to perform experiments to determine those properties are high. Therefore, empirical models have been developed for estimating the mechanical properties of rock that are difficult to determine experimentally from properties that are less difficult to measure. However, the inherent variability in rock properties makes the accurate performance of the empirical models unrealistic and therefore necessitate the use of soft computing models. In this study, Gaussian process regression (GPR), artificial neural network (ANN) and response surface method (RSM) have been proposed to predict the static and dynamic rock properties from the P-wave and rock density. The outcome of the study showed that GPR produced more accurate results than the ANN and RSM models. GPR gave the correlation coefficient of above 99% for all the three properties predicted and RMSE of less than 5. The detailed sensitivity analysis is also conducted using the RSM and the P-wave velocity is found to be the most influencing parameter in the rock mechanical properties predictions. The proposed models can give reasonable predictions of important mechanical properties of sedimentary rock.

• Journal of the Korean Society of Safety
• /
• v.30 no.3
• /
• pp.1-6
• /
• 2015

Self-piercing riveting (SPR) is receiving more recognition as a possible and effective solution for joining automotive body panels and structures, particularly for aluminum parts and dissimilar parts. In this study, static strength and fatigue tests were conducted using coach-peel and cross-tension specimens with Al-5052 plates for evaluation of fatigue strength of the SPR joints. For the static experiment results, the fracture modes are classified into pull-out fracture due to influence of plastic deformation of joining area. During the fatigue tests for the coach-peel and cross-tension specimens with Al-5052, interface failure mode occurred on the top substrate close to the rivet head in the most cycle region. There were relationship between applied load amplitude $P_{amp}$ and life time of cycle N, $P_{amp}=715.5{\times}N^{-0.166}$ and $P_{amp}=1967.3{\times}N^{-0.162}$ were for the coach-peel and cross- tension specimens, respectively. The finite element analysis results for specimens were adopted for the parameters of fatigue lifetime prediction. The relation between SWT fatigue parameter and number of cycles was found to be $SWT=192.8N_f^{-0.44}$.

• Proceedings of the Korean Society of Tribologists and Lubrication Engineers Conference
• /
• 2002.10b
• /
• pp.227-228
• /
• 2002

One of the serious challenges in developing rotary compressor with HFC refrigerant is the prediction of scuffing times and wear amounts between vane and roller surface. In this study, the tribological characteristics of sliding surfaces using roller-vane geometry of rotary compressor were investigated. The sliding tests were carried out under various sliding speeds, normal loads and surface roughness. During the tests, friction force, wear scar width, time to failure, surface temperature, and surface roughness were monitored. Because severe wear was occurred on vane surface, TiN coating was applied on sliding surfaces to prolong the wear-life of vane-roller interfaces. From the sliding tests, it was found that there was the optimum initial surface roughness to break in and to prolong the wear life of sliding surfaces. Depending on load and speed, the protective layers, which were composed of metallic oxide and organic compound, were formed on sliding surfaces. Those would play an important role in the amount of friction and wear between roller and vane surfaces.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.7
• /
• pp.33-38
• /
• 2017

The number of damaged lifeline structures have been increasing with urban acceleration under earthquakes. To predict the damage, damage mitigation technology of lifeline structures should be analyzed using damage prediction technology. Therefore, in this paper, the degree of the fragility of structures under an earthquake was evaluated stochastically through an evaluation of the seismic fragility. The aim was to develop damage prediction technology of sewage pipes among the lifeline facilities. The site response was performed using the data from 158 boreholes in Seoul and 7 real earthquake waves to determine the responses in real urban areas. The seismic fragility was deduced through a total of 29822 time history analysis. In addition, sewer pipes were evaluated and the persisting period was passed by applying the research results of strength reduction which is due to sulphate erosion. As a result, the difference in failure probability between 300 and 800 with the smaller diameter of the representative pipes was approximately double and the size of the pipes has a significant effect on the seismic fragility function. Moreover, the failure probability of a seismic load increases by up to 10 fold as the strength reduction rate increases. The results of this study can be used as a means of predicting the damage and countermeasures of sewer pipes and might be reflected in the seismic design of underground facilities.

• Journal of Ocean Engineering and Technology
• /
• v.13 no.3 s.33
• /
• pp.36-48
• /
• 1999

Fatigue life and crack retardation behavior after penetration were experimentally examined using surface pre-cracked specimens of aluminium alloy 5083. The Wheeler model retardation parameter was used successfully to predict crack growth behavior after penetration. By using a crack propagation rule, the change in crack shape after penetration can be evaluated quantitatively. Advanced, waveform-based acoustic emission (AE) techniques have been successfully used to evaluate signal characteristics obtained form fatigue crack propagation and penetratin behavior in 6061 aluminum plate with surface crack under fatigue stress. Surface defects in the structural members are apt to be origins of fatigue crack growth, which may cause serious failure of the whole structure. The nondestructive analysis on the crack growth and penetration from these defects may, therefore, be one of the most important subjects on the reliability of the leak before break (LBB) design. The goal of the present study is to determine if different sources of the AE could be identified by characteristics of the waveforms produced from the crack growth and penetration. AE signals detected in four stages were found to have different signal per stage. With analysis of waveform and power spectrum in 6061 aluminum alloys with a surface crack, it is found to be capabilities on real-time monitoring for the crack propagation and penetration behavior of various damages and defects in structural members.

• Nuclear Engineering and Technology
• /
• v.48 no.1
• /
• pp.236-245
• /
• 2016

In a reactor coolant system of a nuclear power plant (NPP), an overpressure protection system keeps pressure in the loop within 110% of design pressure. However if the system does not work properly, pressure in the loop could elevate hugely in a short time. It would be seriously disastrous if a weak point in the pressure boundary component bursts and releases radioactive material within the containment; and it may lead to a leak outside the containment. In this study, a gross deformation that leads to a burst of pressure boundary components was investigated. Major components in the primary pressure boundary that is structurally important were selected based on structural mechanics, then, they were used to study the burst pressure of components by finite element method (FEM) analysis and by number of closed forms of theoretical relations. The burst pressure was also used as a metric of design optimization. It revealed which component was the weakest and which component had the highest margin to bursting failure. This information is valuable in severe accident progression prediction. The burst pressures of APR-1400, AP1000 and VVER-1000 reactor coolant systems were evaluated and compared to give relative margins of safety.

• Transactions of Materials Processing
• /
• v.31 no.4
• /
• pp.173-178
• /
• 2022

The present study employed two different machine-learning approaches, the extreme gradient boosting (XGB) and light gradient boosting machine (LGBM), to predict a compressive deformation behavior of additively manufactured Ti-6Al-4V. Such approaches have rarely been verified in the field of metallurgy in contrast to artificial neural network and its variants. XGB and LGBM provided a good prediction for elongation to failure under an extrapolated condition of processing parameters. The predicting accuracy of these methods was better than that of response surface method. Furthermore, XGB and LGBM with optimum hyperparameters well predicted a deformation behavior of Ti-6Al-4V additively manufactured under the extrapolated condition. Although the predicting capability of two methods was comparable, LGBM was superior to XGB in light of six-fold higher rate of machine learning. It is also noted this work has verified the LGBM approach in solving the metallurgical problem for the first time.

• Transactions of Materials Processing
• /
• v.23 no.4
• /
• pp.238-243
• /
• 2014

Recently, magnesium alloys have been widely used in the automotive, aerospace and electronics industries with the advantages of high specific strength, excellent machinability, high electrical conductivity, and high thermal conductivity. Deep drawn magnesium alloys not only meet the demands environmentally and the need for lighter products, but also can lead to remarkably improved productivity and more rapid qualification of the product The current study reports on a failure prediction procedure using finite element modeling (FEM) and a ductile fracture criterion and applies this procedure to the design of a deep drawing process. Critical damage values were determined from a series of uniaxial tensile tests and FEM simulations. They were then expressed as a function of strain rate and temperature. Based on the plastic deformation histories obtained from the FEM analyses of the warm drawing process and the critical damage value curves, the initiation time and location of fracture were predicted. The proposed method was applied to the process design for fabrication of a Mg automotive compressor case and verified with experimental results. The final results indicate that a Mg case part 39% lighter than an Al die casting part can be produced without any defects.

• Tribology and Lubricants
• /
• v.20 no.6
• /
• pp.337-342
• /
• 2004

One of the serious challenges in developing rotary compressor with HFC refrigerant is the prediction of scuffing times and wear amounts between vane and roller surfaces. In this study, the tribological characteristics of sliding surfaces using vane-roller geometry of rotary compressor were investigated. The sliding tests were carried out under various sliding speeds, normal loads and surface roughness. During the test, friction force, wear depth, time to failure and surface temperature were monitored. Because severe wear occurred on vane surface, TiN coating was applied on sliding surfaces to prolong the wear life of vane-roller interfaces. From the sliding test it was found that there was the optimum initial surface roughness to break in and to prolong the wear life of sliding surfaces. Depending on the load and speed, the protective layers, which were composed of metallic oxide and organic compound, were formed on sliding surfaces. Those would play an important role in the amounts of friction and wear between roller and vane surfaces.

• Geomechanics and Engineering
• /
• v.37 no.6
• /
• pp.629-641
• /
• 2024

The presence of excavations or cavities beneath the foundations of a building can have a significant impact on their stability and cause extensive damage. Traditional methods for calculating the bearing capacity and subsidence of foundations over cavities can be complex and time-consuming, particularly when dealing with conditions that vary. In such situations, machine learning (ML) and deep learning (DL) techniques provide effective alternatives. This study concentrates on constructing a prediction model based on the performance of ML and DL algorithms that can be applied in real-world settings. The efficacy of eight algorithms, including Regression Analysis, k-Nearest Neighbor, Decision Tree, Random Forest, Multivariate Regression Spline, Artificial Neural Network, and Deep Neural Network, was evaluated. Using a Python-assisted automation technique integrated with the PLAXIS 2D platform, a dataset containing 272 cases with eight input parameters and one target variable was generated. In general, the DL model performed better than the ML models, and all models, except the regression models, attained outstanding results with an R² greater than 0.90. These models can also be used as surrogate models in reliability analysis to evaluate failure risks and probabilities.