• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.3
• /
• pp.82-88
• /
• 2014

We study the effect of mesh reordering on Laplacian smoothing for parallel high performance computing systems. Specifically, we use the Reverse-Cuthill McKee algorithm to reorder meshes and use Laplacian Smoothing to improve the mesh quality on Nonuniform memory access architecture-based parallel high performance computing systems. First, we investigate the effect of using mesh reordering on Laplacian smoothing for a single core system and extend the idea to NUMA-based high performance computing systems.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.01a
• /
• pp.369-372
• /
• 2023

본 논문에서는 삼각형 구조로 구성된 3차원 메쉬(Mesh)에서 합성곱 신경망(Convolution Neural Network, CNN)을 응용하여 정확도가 높은 새로운 학습 표현 기법을 제시한다. 우리는 메쉬를 구성하고 있는 폴리곤의 edge와 face의 로컬 특징을 기반으로 학습을 진행한다. 일반적으로 딥러닝은 인공신경망을 수많은 계층 형태로 연결한 기법을 말하며, 주요 처리 대상은 1, 2차원 데이터 형태인 오디오 파일과 이미지였다. 인공지능에 대한 연구가 지속되면서 3차원 딥러닝이 도입되었지만, 기존의 학습과는 달리 3차원 딥러닝은 데이터의 확보가 쉽지 않다. 혼합현실과 메타버스 시장의 확대로 인해 3차원 모델링 시장이 증가하고, 기술의 발전으로 데이터를 획득할 수 있는 방법이 생겼지만, 3차원 데이터를 직접적으로 학습에 이용하는 방식으로 적용하는 것은 쉽지 않다. 그렇게 때문에 본 논문에서는 산업 현장에서 이용되는 데이터인 메쉬 구조를 폴리곤의 최소 단위인 삼각형 형태로 구성하여 학습 데이터를 구성해 기존의 방법보다 정확도가 높은 학습 기법을 제안한다.

• International Journal of Naval Architecture and Ocean Engineering
• /
• v.2 no.3
• /
• pp.139-145
• /
• 2010

In the present study, we conducted resistance test, propeller open water test and self-propulsion test for a ship's resistance and propulsion performance, using computational fluid dynamics techniques, where a Reynolds-averaged Navier-Stokes equations solver was employed. For convenience of mesh generation, unstructured meshes were used in the bow and stern region of a ship, where the hull shape is formed of delicate curved surfaces. On the other hand, structured meshes were generated for the middle part of the hull and the rest of the domain, i.e., the region of relatively simple geometry. To facilitate the rotating propeller for propeller open water test and self-propulsion test, a sliding mesh technique was adopted. Free-surface effects were included by employing the volume of fluid method for multi-phase flows. The computational results were validated by comparing with the existing experimental data.

• Interaction and multiscale mechanics
• /
• v.2 no.3
• /
• pp.277-293
• /
• 2009

The Enriched Free Mesh Method (EFMM) is a patch-wise procedure in which both a displacement field on an element and a stress/strain field on a cluster of elements connected to a node can be defined. On the other hand, the Superconvergent Patch Recovery (SPR) is known to be an efficient post-processing procedure of the finite element method to estimate the error norm at a node. In this paper, we discuss the relationship between solutions of the EFMM and those of the SPR through several convergence studies. In addition, in order to solve the demerit of the smoothing effect on the fracture mechanics fields, we implement a singular stress field to a local patch in the EFMM, and its effectiveness is investigated.

• Journal of the Society of Naval Architects of Korea
• /
• v.53 no.3
• /
• pp.210-216
• /
• 2016

Direct numerical simulations of a spatially developing turbulent boundary layer on a flat plate have been performed to verify the applicability of OpenFOAM and adapted mesh with prism layers to turbulent numerical simulation with high fidelity as well as provide a guideline on numerical schemes and parameters of OpenFOAM. Reynolds number based on a momentum thickness at inlet and a free-stream velocity was Re_θ=300. Time dependent inflow fields with near-wall turbulent structures were generated by a method of Lund et al. (1998), which was to extract instantaneous velocity fields from an auxiliary simulation with rescaled and recycled velocities at inlet. To ascertain the statistical characteristics of turbulent boundary layer, the mean profiles of streamwise velocity and turbulent intensities obtained from structured and adapted meshes were compared with the previous data.

• International Journal of Naval Architecture and Ocean Engineering
• /
• v.4 no.3
• /
• pp.228-240
• /
• 2012

Morphing is a geometric interpolation technique that is often used by the animation industry to transform one form into another seemingly seamlessly. It does this by producing a large number of 'intermediate' forms between the two 'extreme' or 'parent' forms. It has already been shown that morphing technique can be a powerful tool for form design and as such can be a useful addition to the armoury of product designers. Morphing procedure itself is simple and consists of straightforward linear interpolation. However, establishing the correspondence between vertices of the parent models is one of the most difficult and important tasks during a morphing process. This paper discusses the mesh-merging method employed for this process as against the already established mesh-regularising method. It has been found that the merging method minimises the need for manual manipulation, allowing automation to a large extent.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.16 no.5
• /
• pp.215-223
• /
• 2021

3D data can be categorized into two parts : Euclidean data and non-Euclidean data. In general, 3D data exists in the form of non-Euclidean data. Due to irregularities in non-Euclidean data such as mesh and point cloud, early 3D deep learning studies transformed these data into regular forms of Euclidean data to utilize them. This approach, however, cannot use memory efficiently and causes loses of essential information on objects. Thus, various approaches that can directly apply deep learning architecture to non-Euclidean 3D data have emerged. In this survey, we introduce various deep learning methods for mesh and point cloud data. After analyzing the operating principles of these methods designed for irregular data, we compare the performance of existing methods for shape classification and segmentation tasks.

• Journal of the Society of Naval Architects of Korea
• /
• v.48 no.3
• /
• pp.260-266
• /
• 2011

For long slender offshore structures, such as cables and pipe lines, their interaction with surrounding fluid flow becomes an important issue for global design of ocean systems. We employ a long circular cylinder as a representative case of slender offshore structure. A flexibly mounted cylinder in cross-flow generates complex vortex shedding and results in oscillation of the structure. In this paper, flow behind a circular cylinder at Re=100 is simulated. The vortex shedding pattern and flow induced motion are examined in the cross flow configuration as well as with various yaw-angled configurations. The "Lock-in" phenomenon is also observed when reduced velocity is approximately 4.0. The MAC Grid system, which is the typical grid system for Cartesian mesh and pressure correction methods, are used for solving the incompressible Navier-Stokes equations. Predictor/Corrector method is applied for obtaining a non-linear response of structure at the flexibly mounted. The existance and motion of the body is represented by the immersed boundary technique.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.11
• /
• pp.91-96
• /
• 2009

In this paper, interoperability in heterogeneous wireless mesh network, and mesh nodes for providing efficient IP mobility technique offers multi-hop ARP. Heterogeneous wireless mesh networks to MANETs based on a wireless mesh network backbone and non-MANET architecture is based on a client wireless mesh network and the two mobile networks, combined with a hybrid wireless mesh network are separate. In two different hybrid wireless mesh network routing protocols used to connect the two protocols in the protocol conversion at the gateway to parallel processing problems seriously overload occurs. All of the network reliability and stability are factors that reduce. Therefore, for efficient integration with L3 routing protocols, design techniques to build ARP multi-hop go through the experiment to increase the number of mesh nodes, the packet forwarding rate and an increased hop number of the node was to ensure reliability and stability.

• Journal of the Semiconductor & Display Technology
• /
• v.20 no.4
• /
• pp.6-9
• /
• 2021

The 3D industry is drawing attention for its applications in various markets, including architecture, media, VR/AR, metaverse, imperial broadcast, and etc.. The current feature of the architecture we are introducing is to make 3D models more easily created and modified than conventional ones. Existing methods for generating 3D models mainly obtain values using specialized equipment such as RGB-D cameras and Lidar cameras, through which 3D models are constructed and used. This requires the purchase of equipment and allows the generated 3D model to be verified by the computer. However, our framework allows users to collect data in an easier and cheaper manner using cell phone cameras instead of specialized equipment, and uses 2D data to proceed with 3D modeling on the server and output it to cell phone application screens. This gives users a more accessible environment. In addition, in the 3D modeling process, object classification is attempted through deep learning without user intervention, and mesh and texture suitable for the object can be applied to obtain a lively 3D model. It also allows users to modify mesh and texture through requests, allowing them to obtain sophisticated 3D models.