• Journal of Korean Vacuum Science & Technology
• /
• v.7 no.2
• /
• pp.39-44
• /
• 2003

Recent ULSI and multilevel structure trends in microelectronic devices minimize the line width down to a quarter micron and below, which results in the high current densities in thin film interconnections. Under high current densities, an EM(electromigration) induced failure becomes one of the critical problems in a microelectronic device. This study is to improve thin film interconnection materials by investigating the EM characteristics in PSG(phosphosilicate glass)/SiO$_2$ passivated Al-l%Si thin film interconnections. Straight line patterns, wide and narrow link type patterns, and meander type patterns, etc. were fabricated by a standard photholithography process. The main results are as follows. The current crowding effects result in the decrease of the lifetime in thin film interconnections. The electric field effects accelerate the decrease of lifetime in the double-layered thin film interconnections. The lifetime of interconnections also depends upon the current conditions of P.D.C.(pulsed direct current) frequencies applied at the same duty factor.

• The Journal of the Korea Contents Association
• /
• v.13 no.4
• /
• pp.399-406
• /
• 2013

Recently, studies about process mining for creating and analyzing business process models from log data have received much attention from BPM (Business Process Management) researchers. Process mining is a kind of method that extracts meaningful information and hidden rules from the event log of enterprise information systems such as ERP and BPM. In this paper, repair processes of electronic devices are analyzed using ProM which is a process mining tool. And based on the analysis of repair processes, the method for finding major failure patterns is proposed by multi-dimensional data analysis beyond simple statistics. By using the proposed method, the reliability of electronic device can be increased by providing the identified failure patterns to design team.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.17 no.10
• /
• pp.289-299
• /
• 2016

Weapon systems require reliability in the development phase for efficient combat readiness. Improved reliability in various manufacturing processes have been achieved using data analysis. However, data analysis in the development phase is difficult due to problems such as the lack of data, high cost, and the importance of security. Therefore, Post Logistics Support (PLS) data collected following integration is analyzed for long-term quality improvement of weapon systems. In this study, we propose a methodology for examining the correlation between the failure rate and PLS data as follows: First, key variables affecting reliability were identified the correlation between variables on the failure rate examined. Second, corresponding analysis was conducted for determining the correlation between patterns of categorical data. Third, extract categories with the higher contribution and quality of representation, and find the highest variable correlated with failure period through visualization. Then, after selecting patterns which have shorter failure period, the cause of decreased reliability was confirmed through frequency analysis. This study will contribute to improving reliability when developing new weapon systems and will help to strengthen the combat readiness of military.

• Proceedings of the Korean Society for Technology of Plasticity Conference
• /
• 2007.05a
• /
• pp.258-261
• /
• 2007

Failure phenomena in uni-axial tension test were experimentally and numerically investigated for AA6111-T4, AA5083-H18 and DP-Steel, which were friction-stir welded with the same and different thicknesses. Forming limit diagram(FLD) was measured using hemispherical dome stretching tests for base materials and also predicted by Hill's bifurcation and M-K theories for welded areas. Finite element simulations well predicted hardening behaviors, failure locations as well as failure patterns for the uni-axial tension tests especially utilizing very fine meshes and FLD along with stress softening.

• Computers and Concrete
• /
• v.19 no.1
• /
• pp.99-110
• /
• 2017

In this paper, the effect of normal load on the failure mechanism of echelon joint has been studied using PFC2D. In the first step, calibration of PFC was undertaken with respect to the data obtained from experimental laboratory tests. Then, six different models consisting various echelon joint were prepared and tested under two low and high normal loads. Furthermore, validation of the simulated models were cross checked with the results of direct shear tests performed on non-persistent jointed physical models. The simulations demonstrated that failure patterns were mostly influenced by normal loading, while the shear strength was linked to failure mechanism. When ligament angle is less than $90^{\circ}$, the stable crack growth length is increased by increasing the normal loading. In this condition, fish eyes failure pattern occur in rock bridge. With higher ligament angles, the rock bridge was broken under high normal loading. Applying higher normal loading increases the number of fracture sets while dilation angle and mean orientations of fracture sets with respect to ligament direction will be decreased.

• Proceedings of the Korea Concrete Institute Conference
• /
• 2006.05a
• /
• pp.282-285
• /
• 2006

This study is to examine bond strength of beam reinforced with GFRP rebar under 4-point bending test by adopting BRITISH STANDARD. The variables were made to have bonding length of 5times$(5d_b)$, 10times$(10d_b)$ and 15times$(15d_b)$ of the nominal diameter of GFRP rebar and were done to analyze the relationship between the bonding strength and the slip. In the result of the test, pull-out failure was dominant in the $5d_b$ and $10d_b$ specimen, both patterns of the pull-out failure and concrete splitting failure appeared in the $10d_b$. On the other hand, the $15d_b$ specimen showed only concrete splitting failure at the end of bonding length. Therefore, it was prove that available bonding length of the GFRP rebar under bending condition on static test is over $15d_b$ then farther research such as fatigue bending test, development of bonding model, FEM parameter study should be performed.

• Radiation Oncology Journal
• /
• v.7 no.1
• /
• pp.45-50
• /
• 1989

From Feb.1985 to Feb.1988,76 patients with squamous cell carcinoma of the lung treated at the Department of Therapeutic Radiology in Kyungpook National University Hospital were available for the analysis of this study. All patients received radiation of 4000cGy-6600cGy with curative aim. The overall rate of complete response was 25.0% and partial response was 452.6% The complete and partial regression of tumor was 14.3% in patients treated with dose below 5000cGy and 84.1% in the group treated with dose above 5000cGy (p<0.01). The complete response was seen only in the group of patients received radiation at least 6000cGy. The patterns of failure were as follows. The rate of initial intrathoracic recurrence was 52.6% in patients with complete response. The overall rate of failure was 68.8%. Distant metastasis was found in 47.4% of patients. Bone, contralateral lung, and brain were common metastatic sites in decreasing order All of the distant metastases and 80% of local recurrences were found within the first year after treatment.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.21 no.2
• /
• pp.146-153
• /
• 2011

Model based machine condition diagnosis methods are generally using a normal and many failure models which need sufficient data to train the models. However, data, especially for failure modes of interest, is very hard to get in real applications. So their industrial applications are either severely limited or impossible when the failure models cannot be trained. In this paper, continuous hidden Markov model(CHMM) with only a normal model has been suggested as a very promising machine condition diagnosis method which can be easily used for industrial applications. Generally hidden Markov model also uses many pattern models to recognize specific patterns and the recognition results of CHMM show the likelihood trend of models. By observing this likelihood trend of a normal model, it is possible to detect failures. This method has been successively applied to arc weld defect diagnosis. The result shows CHMM's big potential as a machine condition monitoring method.

• Transactions of Materials Processing
• /
• v.16 no.4 s.94
• /
• pp.304-308
• /
• 2007

Failure phenomena in uni-axial tension test were experimentally and numerically investigated for AA6111-T4, AA5083-H18 and DP-Steel, which were friction-stir welded with the same and different thicknesses. Forming limit diagram(FLD) was measured using hemispherical dome stretching tests for base materials and also predicted by Hill's bifurcation and M-K theories for welded areas. Finite element simulations well predicted hardening behaviors, failure locations as well as failure patterns for the uni-axial tension tests especially utilizing very fine meshes and FLD along with stress softening.

• Convergence Security Journal
• /
• v.10 no.1
• /
• pp.55-61
• /
• 2010

It is of great practical interest to deciding when to stop testing a software system in development phase and transfer it to the user. This decision problem called an optimal release policies. In this paper, because of the possibility of introducing new faults when correcting or modifying the software, we were researched release comparative policies which based on infinite failure NHPP model and types of interval failure times. The policies which minimize a total average software cost of development and maintenance under the constraint of satisfying a software reliability requirement can optimal software release times. In a numerical example, applied data which were patterns, if intensity function constant or increasing, decreasing, estimated software optimal release time.