조인호 교수

A new high-efficient phase-shifted full-bridge (PSFB) converter is proposed in this paper. The conventional PSFB converter with an external inductor and clamping diodes is widely used in server power systems due to limited voltage stress on switches and its zero-voltage switching (ZVS) characteristics. However, a hold-up time regulation rule in server systems limits its operating duty ratio in nominal state and it increases the conduction loss of the converter. With simple modification, the prop

This paper presents a half-bridge LLC resonant converter having a boost pulse width modulation (PWM) converter characteristic for hold-up state operation. The proposed converter is based on a half-bridge LLC resonant converter structure and a single auxiliary switch is added at the primary side. The converter has two different operational characteristics. It shows the same operational characteristic with the conventional LLC resonant converters during nominal state, which is frequency modulation

Decades-long experimental databases become accessible in global earthquake engineering community. Yet, complex interactions of a multitude of variables pose formidable challenges to data-driven research. We embarked upon developing an advanced statistical learning and prediction framework with the generalized additive model (GAM). We showed promising performance of GAM with applications to existing RC shear wall databases. Without any prejudice, GAM can predict structural responses accurately us

SUMMARY Multiscale analysis technique became a successful remedy to complicated problems in which nonlinear behavior is linked with microscopic damage mechanisms. For efficient parallel multiscale analyses, hierarchical grouping algorithms (e.g., the two‐level ‘coarse‐grained’ method) have been suggested and proved superior over a simple parallelization. Here, we expanded the two‐level algorithm to give rise to a multilayered grouping parallel algorithm suitable for large‐scale multiple‐level mu

Parallel computing in civil engineering has been restricted to monotonic shock or blast loading with explicit algorithm which is characteristically feasible to be parallelized. In the present paper, efficient parallelization strategies for the highly demanded implicit nonlinear finite-element analysis (FEA) program for real scale reinforced concrete (RC) structures under cyclic loading are proposed. Quantitative comparison of state-of-the-art parallel strategies in terms of factorization were ca

Summary There exists a deep chasm between machine learning (ML) and high‐fidelity computational material models in science and engineering. Due to the complex interaction of internal physics, ML methods hardly conquer or innovate them. To fill the chasm, this paper finds an answer from the central notions of deep learning (DL) and proposes information index and link functions, which are essential to infuse principles of physics into ML. Like the convolution process of DL, the proposed informatio

We embarked upon developing a novel parallel simulation platform that is rooted in microphysical mechanisms. Primarily aiming at large‐scale reinforced‐concrete structures exposed to cyclic loading, we sought to settle the question as to how to capture nonlinear shear, localized damage and progressive buckling of reinforcing bar. We proposed a tribology‐inspired three‐dimensional (3‐D) interlocking mechanism in the well‐established framework of multidirec‐tional smeared crack models. Strong corr

Abstract Attempts to use machine learning to discover hidden physical rules are in their infancy, and such attempts confront more challenges when experiments involve multifaceted measurements over three-dimensional objects. Here we propose a framework that can infuse scientists’ basic knowledge into a glass-box rule learner to extract hidden physical rules behind complex physics phenomena. A “convolved information index” is proposed to handle physical measurements over three-dimensional nano-sca

The significant difference in the values obtained using the Osstell™ and Osstell™ Mentor between the first and second stages of implant surgery indicates that these values can be a convenient and precise way for evaluating the implant stability in clinical practice.

The confinement effect has been of significant importance for improving the resilience against extreme compression loadings such as seismic excitations. Notwithstanding the accuracy of previous confinement models, some challenges remain regarding their applicability. The previous approaches often build on structure-dependent parameters necessitating intractable calibrations, and their formulations are defined on an integration point or a small portion of the structure, thereby precluding general

Large‐scale earthquake risk assessment necessitates models of the seismic performance of building classes. This work addresses how to depict a class with only a few (index) buildings whose designs span the attributes that most impact the seismic behavior of the class. We propose a general numerical moment matching (MM) technique to represent those seismic attributes of index buildings, which can then be individually analyzed by second‐generation performance‐based earthquake engineering methods (