• Structural Engineering and Mechanics
• /
• v.72 no.1
• /
• pp.31-41
• /
• 2019

Structural integrity assessment of piping components is of paramount important for remaining life prediction, residual strength evaluation and for in-service inspection planning. For accurate prediction of these, a reliable fracture parameter is essential. One of the fracture parameters is stress intensity factor (SIF), which is generally preferred for high strength materials, can be evaluated by using linear elastic fracture mechanics principles. To employ available analytical and numerical procedures for fracture analysis of piping components, it takes considerable amount of time and effort. In view of this, an alternative approach to analytical and finite element analysis, a model based on relevance vector machine (RVM) is developed to predict SIF of part through crack of a piping component under fatigue loading. RVM is based on probabilistic approach and regression and it is established based on Bayesian formulation of a linear model with an appropriate prior that results in a sparse representation. Model for SIF prediction is developed by using MATLAB software wherein 70% of the data has been used for the development of RVM model and rest of the data is used for validation. The predicted SIF is found to be in good agreement with the corresponding analytical solution, and can be used for damage tolerant analysis of structural components.

• Advances in concrete construction
• /
• v.17 no.1
• /
• pp.27-36
• /
• 2024

Reinforced concrete (RC) flat slabs should be designed based on punching shear strength. As part of this study, machine learning (ML) algorithms were developed to accurately predict the punching shear strength of RC flat slabs without shear reinforcement. It is based on Bayesian optimization (BO), combined with four standard algorithms (Support vector regression, Decision trees, Random forests, Extreme gradient boosting) on 446 datasets that contain six design parameters. Furthermore, an analysis of feature importance is carried out by Shapley additive explanation (SHAP), in order to quantify the effect of design parameters on punching shear strength. According to the results, the BO method produces high prediction accuracy by selecting the optimal hyperparameters for each model. With R² = 0.985, MAE = 0.0155 MN, RMSE = 0.0244 MN, the BO-XGBoost model performed better than the original XGBoost prediction, which had R² = 0.917, MAE = 0.064 MN, RMSE = 0.121 MN in total dataset. Additionally, recommendations are provided on how to select factors that will influence punching shear resistance of RC flat slabs without shear reinforcement.

• Nuclear Engineering and Technology
• /
• v.54 no.8
• /
• pp.3027-3033
• /
• 2022

Background: In this study, various types of deep-learning models for predicting in vitro radiosensitivity from gene-expression profiling were compared. Methods: The clonogenic surviving fractions at 2 Gy from previous publications and microarray gene-expression data from the National Cancer Institute-60 cell lines were used to measure the radiosensitivity. Seven different prediction models including three distinct multi-layered perceptrons (MLP), four different convolutional neural networks (CNN) were compared. Folded cross-validation was applied to train and evaluate model performance. The criteria for correct prediction were absolute error < 0.02 or relative error < 10%. The models were compared in terms of prediction accuracy, training time per epoch, training fluctuations, and required calculation resources. Results: The strength of MLP-based models was their fast initial convergence and short training time per epoch. They represented significantly different prediction accuracy depending on the model configuration. The CNN-based models showed relatively high prediction accuracy, low training fluctuations, and a relatively small increase in the memory requirement as the model deepens. Conclusion: Our findings suggest that a CNN-based model with moderate depth would be appropriate when the prediction accuracy is important, and a shallow MLP-based model can be recommended when either the training resources or time are limited.

• Proceedings of the Korea Concrete Institute Conference
• /
• 2001.05a
• /
• pp.189-194
• /
• 2001

In this study, ultimate strength of concrete deep beams with an opening is predicted by using Strut-and-Tie Model with a new effective compressive strength. First crack occurs around an opening by stress concentration due to geometric discontinuity. This results in decreasing ultimate strength of deep beams with an opening compared with general deep beams. With fundamental notion that ultimate strength of deep beam with an opening decreases as a result of reduction in effective compressive strength of a concrete strut, an equivalent effective compressive strength formula is proposed in order to reflect ultimate strength reduction due to an opening located in a concrete strut. An equivalent effective compressive strength formula which can reflect opening size and position is added to a testified algorithm of predicting ultimate strength of concrete deep beams. Therefore, ultimate strength of concrete deep beam with an opening is predicted by using a simple and rational STM algorithm including an equivalent effective compressive strength formula, not by finite element analysis or a former complex Strut-and-Tie Model

• Geomechanics and Engineering
• /
• v.33 no.1
• /
• pp.1-9
• /
• 2023

The unconfined compression strength (UCS) of soils is commonly used either before or during the construction of geo-structures. In the pre-design stage, UCS as a mechanical property is obtained through a laboratory test that requires cumbersome procedures and high costs from in-situ sampling and sample preparation. As an alternative way, the empirical model established from limited testing cases is used to economically estimate the UCS. However, many parameters affecting the 1D soil compression response hinder employing the traditional statistical analysis. In this study, gene expression programming (GEP) is adopted to develop a prediction model of UCS with common affecting soil properties. A total of 79 undisturbed soil samples are collected, of which 54 samples are utilized for the generation of a predictive model and 25 samples are used to validate the proposed model. Experimental studies are conducted to measure the unconfined compression strength and basic soil index properties. A performance assessment of the prediction model is carried out using statistical checks including the correlation coefficient (R), the root mean square error (RMSE), the mean absolute error (MAE), the relatively squared error (RSE), and external criteria checks. The prediction model has achieved excellent accuracy with values of R, RMSE, MAE, and RSE of 0.98, 10.01, 7.94, and 0.03, respectively for the training data and 0.92, 19.82, 14.56, and 0.15, respectively for the testing data. From the sensitivity analysis and parametric study, the liquid limit and fine content are found to be the most sensitive parameters whereas the sand content is the least critical parameter.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2004.05a
• /
• pp.442-445
• /
• 2004

The acoustic target strength (TS) of submarine is associated with its active detection, positioning and classification. That is, the survivability of submarine depends on its target strength. So it should be managed with all possible means. An anechoic coating to existing submarine or changing of curvature can be considered as major measures to reduce the TS of submarine. It is mainly based on the prediction of its TS. Under this circumstances, a study on the more accurate numerical methods becomes big topic for submarine design. In this paper, Kirchhoff approximation method was adopted as a numerical tool for the physical optics region. Secondly, the scaled models of submarine were built and tested in order to verify its performance. Through the comparison, it was found out that the Kirchhoff approximation method could be good design tool for the prediction of TS of submarine.

• Computers and Concrete
• /
• v.14 no.3
• /
• pp.233-245
• /
• 2014

In this paper, a rule based Mamdani type fuzzy logic model for prediction of slippage at maximum tensile strength and slippage at rupture of structural lightweight concretes were discussed. In the model steel rebar diameters and development lengths were used as inputs. The FL model and experimental results, the coefficient of determination R2, the Root Mean Square Error were used as evaluation criteria for comparison. It was concluded that FL was practical method for predicting slippage at maximum tensile strength and slippage at rupture of structural lightweight concretes.

• Proceedings of the Korean Society For Composite Materials Conference
• /
• 2002.10a
• /
• pp.139-142
• /
• 2002

This study presents a mathematical model predicting the stress-strain behavior of fiber reinforced (FMMCs) and fiber/particle reinforced metal matrix composites (F/P MMCs). MMCs were fabricated by squeeze casting method using Al2O3 short fiber and particle as reinforcement, and A356 aluminum alloy as matrix. The fiber/particle ratios of F/P MMCs were 2:1, 1:1, 1:2 with the total reinforcement volume fraction of 20 vol.%, and the FMMCs were reinforced with 10 vol,%, 15 vol. %, 20 vol. % of fibers. Tensile tests were conducted and compared with predictions which were derived using laminate analogy theory and multi-failure model of reinforcements. Results show that the tensile strength of FMMCs with 10 vol.% of fiber was well matched with prediction, and as the fiber volume increases, predictions become larger than experimental results. The difference between the prediction and experiment is considered to be a result of matrix allowance of fiber damage in tensile loading. As the fiber volume fraction in FMMCs increases, the fiber damage increases and so that the tensile strength is reduced. The strength of F/P MMCs approaches more closely to the prediction than FMMCs reinforced with 20 vol.% of fibers because F/P MMCs contains small quantity of fibers and thus has a positive effect in fiber strengthening.

• ETRI Journal
• /
• v.29 no.4
• /
• pp.527-529
• /
• 2007

In this paper, a decision-tree-based Markov model for phrase break prediction is proposed. The model takes advantage of the non-homogeneous-features-based classification ability of decision tree and temporal break sequence modeling based on the Markov process. For this experiment, a text corpus tagged with parts-of-speech and three break strength levels is prepared and evaluated. The complex feature set, textual conditions, and prior knowledge are utilized; and chunking rules are applied to the search results. The proposed model shows an error reduction rate of about 11.6% compared to the conventional classification model.

• International Journal of Concrete Structures and Materials
• /
• v.5 no.1
• /
• pp.57-64
• /
• 2011

An extensive database has been constructed of reinforced concrete (RC) beam-column connection tests subjected to cyclic lateral loading. All cases within the database experienced joint shear failure, either in conjunction with or without yielding of longitudinal beam reinforcement. Using the experimental database, envelope curves of joint shear stress vs. joint shear strain behavior have been created by connecting key points such as cracking, yielding, and peak loading. Various prediction approaches for RC joint shear behavior are discussed using the constructed experimental database. RC joint shear strength and deformation models are first presented using the database in conjunction with a Bayesian parameter estimation method, and then a complete model applicable to the full range of RC joint shear behavior is suggested. An RC joint shear prediction model following a U.S. standard is next summarized and evaluated. Finally, a particular joint shear prediction model using basic joint shear resistance mechanisms is described and for the first time critically assessed.