• Transactions of the Korean Society of Mechanical Engineers
• /
• v.19 no.7
• /
• pp.1657-1664
• /
• 1995

For dynamic analysis of flexible structures, classical linear modeling has been widely used due to its several good aspects. However, it was found that the modeling often lost its accuracy. So, it is important to know the valid range of the modeling before it is used. more complicated modelings are needed to obtain reliable results only outside the valid range of the classical linear modeling. In this study, some rigid body motions of flexible structures which lead to the failure of the classical linear modeling are investigated. Hybrid deformation variable modeling, which is proved to be accurate in previous studies, is used to figure out the valid range of the classical linear modeling through numerical examples.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.31 no.2
• /
• pp.51-60
• /
• 2008

A solution method for fuzzy linear programs is proposed. A fuzzy linear program is converted to a crisp linear program with average indices being applied to the objective function and constraints. A comparative analysis between the proposed average index approach and the possibilistic approach is given. As an application example, the proposed method is applied to the linear programming model for fuzzy data envelopment analysis, and the result is compared with that of the possibilistic approach.

• Journal of applied mathematics & informatics
• /
• v.14 no.1_2
• /
• pp.377-386
• /
• 2004

A technique that provides prediction intervals based on a model called an empirical linear model is discussed. The technique, high-dimensional empirical linear prediction (HELP), involves principal component analysis, factor analysis and model selection. HELP can be viewed as a technique that provides prediction (and confidence) intervals based on a factor analysis models do not typically have justifiable theory due to nonidentifiability, we show that the intervals are justifiable asymptotically.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2004.10a
• /
• pp.326-333
• /
• 2004

This paper presents the structural characteristics analysis of a high-speed horizontal machining center with spindle speed of 50, 000rpm and feedrate of 120m/min. The spindle system is designed based on the built-in motor, angular contact ceramic ball bearings, oil-air lubrication and oil-jacket cooling method. The X-axis and Y-axis feeding systems are composed of the linear motor and linear motion guides, and the Z-axis feeding system is composed of the servo-motor, ball screw and linear motion guide. The structural analysis model of the high-speed horizontal machining center is constructed by the finite element method, and the validity of structural design is estimated based on the structural deformation of the high-speed horizontal machining center and spindle nose caused by the gravity and inertia forces.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.14 no.9 s.90
• /
• pp.868-876
• /
• 2004

The linear motion(LM) guide using ball bearing has many advantages compared with conventional sliding guides. Therefore, LM guide using ball bearing has been widely used to increase the accuracy of the position of a system. This research investigates dynamic characteristics of LM guide through mainly linear analyses. Linear analysis is accomplished by Lagrange equation and the finite element method. And another trial that performs nonlinear analysis about one mode(bouncing mode) of LM guide from Hertzian contact theory is accomplished in the latter half of this research. Through nonlinear analysis we could observe the softening characteristic due to the Hertzian contact nonlinearity.

• Journal of the Korea Society of Computer and Information
• /
• v.23 no.12
• /
• pp.21-26
• /
• 2018

As the large amount of data can be efficiently stored, the methods extracting meaningful features from big data has become important. Especially, the techniques of converting high- to low-dimensional data are crucial for the 'Data visualization'. In this study, principal component analysis (PCA; linear dimensionality reduction technique) and Isomap (non-linear dimensionality reduction technique) are introduced and applied to neural big data obtained by the functional magnetic resonance imaging (fMRI). First, we investigate how much the physical properties of stimuli are maintained after the dimensionality reduction processes. We moreover compared the amount of residual variance to quantitatively compare the amount of information that was not explained. As result, the dimensionality reduction using Isomap contains more information than the principal component analysis. Our results demonstrate that it is necessary to consider not only linear but also nonlinear characteristics in the big data analysis.

• Earthquakes and Structures
• /
• v.1 no.1
• /
• pp.21-41
• /
• 2010

Multi-span steel-concrete composite (SCC) bridges are very sensitive to earthquake loading. Extensive damage may occur not only in the substructures (piers), which are expected to yield, but also in the other components (e.g., deck, abutments) involved in carrying the seismic loads. Current seismic codes allow the design of regular bridges by means of linear elastic analysis based on inelastic design spectra. In bridges with superstructure transverse motion restrained at the abutments, a dual load path behavior is observed. The sequential yielding of the piers can lead to a substantial change in the stiffness distribution. Thus, force distributions and displacement demand can significantly differ from linear elastic analysis predictions. The objectives of this study are assessing the influence of piers-deck stiffness ratio and of soil-structure interaction effects on the seismic behavior of continuous SCC bridges with dual load path, and evaluating the suitability of linear elastic analysis in predicting the actual seismic behavior of these bridges. Parametric analysis results are presented and discussed for a common bridge typology. The response dependence on the parameters is studied by nonlinear multi-record incremental dynamic analysis (IDA). Comparisons are made with linear time history analysis results. The results presented suggest that simplified linear elastic analysis based on inelastic design spectra could produce very inaccurate estimates of the structural behavior of SCC bridges with dual load path.

• Proceedings of the Earthquake Engineering Society of Korea Conference
• /
• 2003.03a
• /
• pp.422-429
• /
• 2003

A comparative study was performed for a suspension bridge to grasp the possible differences in seismic responses evaluated by several analytical methods. The items mainly investigated are the linear vs. nonlinear response the response spectrum method vs. the linear dynamic analysis method and the damping ratio and it's implementation into analysis procedures. According to the numerical example, it is found that the seismic responses are considerably affected by the damping-related parameters even though slight differences are shown depending on the response quantities md the exciting directions. On the other hand, it is also confirmed that the seismic responses are less affected by the analysis method-related parameters such as the response spectrum method vs. the linear dynamic analysis method, and the linear and nonlinear analysis method. The response spectrum method is expected to give conservative results for the examined bridge, provided that the design response spectrum in the Korean Highway Design Specification is modified according to the proper damping ratio.

• Steel and Composite Structures
• /
• v.19 no.1
• /
• pp.59-74
• /
• 2015

Since there are several places located in active seismic zones in the world, serious damages and losses have happened due to major scaled earthquakes. Especially, structures having different irregularities have been severely damaged or collapsed during these seismic events. Behavior of existing structures under several loading conditions is not completely determined due to some uncertainties. This situation reveals the importance of design and analysis of structures under seismic effects. Several non-linear static procedures have been developed in recent years. Determination of the seismic safety of the existing structures and strengthening techniques are significant civil engineering problems Non-linear methods are defined in codes to determine the performance levels of structures more accurately. However, displacement based ones give more realistic results. These methods provide more reliable evaluation possibilities for existing structures with developing computer technology. In this study, non-linear performance analysis of existing and strengthened steel structures by X shaped bracing members with 3, 5 and 7 stories which have soft story irregularity is performed according to FEMA-356 and Turkish Earthquake Code-2007. Damage ratios of the structural members and global performance levels are determined as well as modal properties and story drift ratios after non-linear finite elements analysis for each structure.

• Proceedings of the KSME Conference
• /
• 2001.06d
• /
• pp.396-401
• /
• 2001

Heat transfer analysis of a iron core type linear motor for surface mounting device applications was considered in this study. In order to avoid the complex conjugate problem a fluid flow regime and a solid regime were considered separately. First, film coefficients of the moving parts were evaluated from computational fluid dynamic analysis and those of the stationary parts from the existing empirical or analytic correlations. And then, by applying them, internal and external temperatures of the linear motor pal1s were computed through finite element analysis. Both computation and measurement were carried out with respect to motor driving power. The measurement did not exhibit a linear temperature variation trend with respect to motor power while the computation revealed a linear correlation. Nonetheless, the computations agreed with the measurements within an error range of 20%. It indicates that an adequate heat transfer model for the reciprocative coil assembly may help more exact prediction.