• Journal of the Korean Society for Nondestructive Testing
• /
• v.17 no.3
• /
• pp.156-161
• /
• 1997

The eddy current probe was designed for the test of fin tubes that have uneven outer and inner surfaces to enhance the efficiency of heat emission. Because of the surface roughness of the fin tube, it needs much care to detect flaws in the tube employing eddy current test(ECT). We made ECT probes with different coil lengths and performed eddy current test using these probes for artificially flawed specimens. By the fast Fourier transform and digital filtering, signals from these probes were processed to characterize frequency spectra. From the analysis of eddy current signals and their frequency spectra, we concluded that, for the effective testing of the tubes with the fins of 1mm pitch, 4mm coil length gave the highest S/N ratio.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.9 no.6
• /
• pp.1379-1386
• /
• 2005

For efficient welding it is necessary to maintain stability of the welding process and control the shape of the welding bead. The welding quality can be controlled by monitoring important parameters, such as, the Arc Voltage, Welding Current and Welding Speed during the welding process. Welding systems use either a vision sensor or an Arc sensor, both of which are unable to control these parameters directly. Therefore, it is difficult to obtain necessary bead geometry without automatically controlling the welding parameters through the sensors. In this paper we propose a novel approach using fuzzy logic and neural networks for improving welding qualify and maintaining the desired weld bead shape. Through experiments we demonstrate that the proposed system can be used for real welding processes. The results demonstrate that the system can efficiently estimate the weld bead shape and remove the welding detects.

• Tunnel and Underground Space
• /
• v.11 no.3
• /
• pp.217-224
• /
• 2001

Rock masses are usually discontinuous in nature due to various geological processes and contain rock joints and bridges. Crack propagation and coalescence processes in rock bridge mainly cause rock failures in slopes, foundations, and tunnels. In this study, we focused on the crack initiation, propagation and coalescence process of rock materials containing two pre-existing open cracks arranged in different geometries. Specimens of 120${\times}$60${\times}$25 mm in size, which were made of Yeoman Marble, were prepared. In the specimens, two artificial cracks were cut with pre-existing crack angle ${\alpha}$, bridge angle ${\beta}$, pre-existing crack length 2c and bridge length 2b. Wing crack initiation stress, wing crack propagation angle, and crack coalescence stress were measured and crack initiation, propagation and coalescence processes were observed during uniaxial compression. Crack coalescence types were classified and analytical study using Ashby and Hallam model (1986) was performed to be compared with the experimental results.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.24 no.4
• /
• pp.307-314
• /
• 2004

A circumferential guided wave method was developed to detect the axial crack on the bent feeder pipe. Dispersion curves of circumferential guided waves were calculated as a function of curvature of the pipe. In the case of thin plate, i.e. infinite curvature, as the frequency increases, the $S_0$ and $A_0$ mode coincide and eventually become Rayleigh wave mode. In the case of pipe, however, as the curvature increases, the lowest modes do not coincide even in the high frequencies. Based on the analysis, a rocking technique using angle beam transducer was applied to detect an axial defect in the bent region of PHWR feeder pipe. Based on the analysis of experimenal data for artificial notches, the vibration modes of each signal were identified. It was found that the notches with the depth of )0% of wall thickness can be detected with the method.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.463-465
• /
• 2021

Currently, the inspection method printed on food packages and boxes is to sample only a few products and inspect them with human eyes. Such a sampling inspection has the limitation that only a small number of products can be inspected. Therefore, accurate inspection using a camera is required. This paper proposes a deep learning object recognition technology model, which is an artificial intelligence technology, as a method for detecting the defects of expiration date printed on the product packaging. Using the Faster R-CNN (region convolution neural network) model, the color images, converted gray images, and converted binary images of the printed expiration date are trained and then tested, and each detection rates are compared. The detection performance of expiration date printed on the package by the proposed method showed the same detection performance as that of conventional vision-based inspection system.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.3
• /
• pp.130-139
• /
• 2024

Artificial intelligence models are being used to detect facility anomalies using physics data such as vibration, current, and temperature for predictive maintenance in the manufacturing industry. Since the types of facility anomalies, such as facility defects and failures, anomaly detection methods using autoencoder-based unsupervised learning models have been mainly applied. Normal or abnormal facility conditions can be effectively classified using the reconstruction error of the autoencoder, but there is a limit to identifying facility anomalies specifically. When facility anomalies such as unbalance, misalignment, and looseness occur, the facility vibration frequency shows a pattern different from the normal state in a specific frequency range. This paper presents an N-segmentation anomaly detection method that performs anomaly detection by dividing the entire vibration frequency range into N regions. Experiments on nine kinds of anomaly data with different frequencies and amplitudes using vibration data from a compressor showed better performance when N-segmentation was applied. The proposed method helps materialize them after detecting facility anomalies.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.30 no.4
• /
• pp.386-395
• /
• 2010

The steam generator management program(SGMP) has recently defined the procedures for the qualification of eddy current hardware and technique. These procedures provide two basic methods for qualification. The first way is to qualify the equipment or the probe by using the flaw mechanism and method of the pulled tubes from the heat exchangers or the artificial flawed tubes. The second way is to verify the equivalency with the characteristics of the qualified equipment or probe. In this case, the qualified equipment or probe may be modified to substitute or replace instruments or probes without re-qualification provided that the range of essential variables defined in the examination technique specification sheet are met. This study is to describe the result of the comparative performance evaluation of bobbin coil eddy current probes manufactured by KEPCO Research Institute and probes manufactured by a foreign manufacturer. As a result of this study, although there were minor differences between the two kinds of probes, it was evaluated that the two kinds of probes were almost identical in the significant performance characteristics described in the KEPCO Research Institute guideline.

• Journal of Internet Computing and Services
• /
• v.24 no.6
• /
• pp.91-98
• /
• 2023

Recently, according to development of artificial intelligence, a wide range of industry being automatic and optimized. Also we can find out some research of using supervised learning for deteceting defect of railway in domestic rail industry. However, there are structures other than rails on the track, and the fastener is a device that binds the rail to other structures, and periodic inspections are required to prevent safety accidents. In this paper, we present a method of reducing cost for labeling using semi-supervised and transfer model trained on rail fastener data. We use Resnet50 as the backbone network pretrained on ImageNet. At first we randomly take training data from unlabeled data and then labeled that data to train model. After predict unlabeled data by trained model, we adopted a method of adding the data with the highest probability for each class to the training data by a predetermined size. Futhermore, we also conducted some experiments to investigate the influence of the number of initially labeled data. As a result of the experiment, model reaches 92% accuracy which has a performance difference of around 5% compared to supervised learning. This is expected to improve the performance of the classifier by using relatively few labels without additional labeling processes through the proposed method.

• Journal of Biomedical Engineering Research
• /
• v.19 no.2
• /
• pp.163-170
• /
• 1998

The ultimate goal of total artificial heart is permanent substitute for a failed heart in a patient without any other therapeutic modality. Until now, infection has been the main problem related to the mechanical circulatory support system. The best way to solve this catastrophic complication and to improve the quality of life of TAH patients in terms of tethering must be implantation of TAH totally. The EH-TAH has been developed in NCVC from 1987 for this purpose. The system consists of an energy converter and pumps, which are designed to be placed in abdomen and pericardial space separately for a good anatomical fit. To evaluate the anatomical fit and hemodynamic performance of the EH-TAH, in vivo test was done. General condition of the animal and hemodynamic status had been stable until the TAH stopped on the 11th pumping day. The estimated cardiac output was about 7.7L/min. The values of mean aortic pressure, left and right atrial pressure were 93$\pm$10, 19$\pm$3 and 15$\pm$4 mmHg, respectively. The correlation coefficient between left and right atrial pressure was 0.96, which represents the dynamic function of the interatrial shunt in controlling left-right imbalance of cardiac output. During pumping days, the temperature on the surface of actuator had been maintained at 39.7$\pm$0.4$^{\circ}C$, less than 1$^{\circ}C$ higher than the rectal temperature. The TAH stopped on the 11th day due to mechanical problems. We concluded that the EH-TAH possessed satisfactory basic performance including anatomic fit and hemodynamic adequacy, although there were several mechanical problems to be solved yet.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.36 no.5
• /
• pp.370-376
• /
• 2016

Corrosion on the surface of a structure can generate cracks or cause walls to thin. This can lead to fracturing, which can eventually lead to fatalities and property loss. In an effort to prevent this, laser imaging technology has been used over the last ten years to detect thin-plate structure, or relatively thin piping. The most common laser imaging was used to develop a new technology for inspecting and imaging a desired area in order to scan various structures for thin-plate structure and thin piping. However, this method builds images by measuring waves reflected from defects, and subsequently has a considerable time delay of a few milliseconds at each scanning point. In addition, the complexity of the system is high, due to additional required components, such as laser-focusing parts. This paper proposes a laser imaging method with an increased scanning speed, based on excitation and the measurement of standing waves in structures. The wavenumber of standing waves changes at sections with a geometrical discontinuity, such as thickness. Therefore, it is possible to detect defects in a structure by generating standing waves with a single frequency and scanning the waves at each point by with the laser scanning system. The proposed technique is demonstrated on a wall-thinned plate with a linear thickness variation.