• 한국가구학회지
• /
• 제17권1호
• /
• pp.83-89
• /
• 2006

Because the surface of pine wood is very soft, there have been many attempts, like as surface coating, resin impregnation, and densificationt, to harden the surface of wood for its interior use. This study was carried out for surface hardening of pine wood by laminating of melamine resin sheet. The effect of laminating method on the characteristics of melamine resin sheet-laminated wood(MLW) was investigated. Flat-sawn softwoods were suitable for MLW making. And the pre-drying of wood at $105^{\circ}C$ for 30min, before laminating, was effective for preparing of defect-free MLW. The abrasion resistance of pine wood was greatly improved by laminating of melamine resin sheet on its surface. Consequently, the laminating of melamine resin sheet was proved to be a favorable method for improvement of surface abrasiveness of pine wood.

• Journal of the Korean Wood Science and Technology
• /
• 제29권2호
• /
• pp.91-99
• /
• 2001

목재산업에서 생산성과 품질 향상을 위해서는 공정의 자동화가 절실히 요구되고 있다. 화상처리시스템의 현장 적용성 개선을 위해 국산 소나무와 상수리나무 및 서어나무를 대상으로 옹이, 할렬, 수피 등과 같은 주요 결함들을 인식해내는 알고리듬을 개발하였다. 목재가공공정에서 컨베이어 위를 이동하고 있는 목재를 가상하여 제재목의 규격과 목제품의 품질에 영향을 주는 표면결함들을 대상으로 그 형태와 크기 및 위치를 인식할 수 있는 화상 처리기술을 개발하였다. 화상처리를 이용하여 결함으로 인식된 각 라벨에 대해 8가지의 특성을 개발, 결함인식의 기준으로 사용하였으며, 8가지 특성을 이용하여 수종별 결함 추출의 제한 조건을 마련한 후 필터링을 실시하여 실제 결함과 결함으로 인정되지 않는 라벨을 구분하였다. 또한 수종별로 가장 효과적인 기준의 적용절차를 밝히므로써 결함 인식의 오차를 줄이는 한편, 처리대상 수종에 대한 유연성을 확보하고자 하였다. 한편, 컨베이어를 이동하고 있는 목재의 규격이나 표면상태에 관련된 모든 자료들이 database의 형태로 작성되도록 하여 이후 실제 산업현장에 대한 적용 가능성을 타진하고자 하였다. 그 결과 소나무와 상수리나무 표변에 존재하는 옹이의 오인률은 1% 이하로 매우 우수하였으며, 상수리나무의 할렬과 서어나무의 수피에 대한 오인률도 13% 이하로 낮출 수 있었다.

• Journal of the Korean Wood Science and Technology
• /
• 제35권5호
• /
• pp.16-23
• /
• 2007

This study was performed as part of a research project aimed at developing an ultrasonic computed tomography (CT) system of wood for field application. In this reports, we investigate the variation of wave velocities on the cross section of real size wooden structural member to confirm the reason of image distortion on CT image of wood, and then proposed a new image reconstruction method by considering the velocity variation on wood cross section. First of all, the effect of wood anisotropy on ultrasonic velocities of wooden members was investigated. Based on the relationship between ultrasonic velocity and annual ring angle, which was obtained from test results of small clear specimens, ultrasonic velocities of each measuring angle were predicted. Next, they were compared with the ultrasonic velocities measured on five wood disks. There were very large differences between predicted and measured results, thought to be caused by the skewing effect of ultrasound and the presence of juvenile-wood. Based on these findings, a new method was proposed to reconstruct cross-sectional image of wood. By using this method, some distortions on reconstructed images could be removed, and defects were more easily and clearly detected. The minimum size of detectable defect was decreased remarkably, from 33 mm to 13 mm. However, the size of the detected defect was enlarged and the position somewhat shifted to the specimen surface on the CT images, which was also thought to be caused by the skewing effect of ultrasound. Additional research has been planned to solve these problems.

• Journal of the Korean Wood Science and Technology
• /
• 제31권2호
• /
• pp.31-37
• /
• 2003

목재의 결함을 탐지하는 비파괴 진단법 중에서 초음파를 이용한 비파괴 시험법은 그 경제성과 사용에서 큰 장점을 지니고 있어, 그 활용과 관련된 연구가 활발히 진행 중이다. 그러나 현재까지의 연구는 입목 또는 원목 등과 같이 비파괴 평가를 위한 측정이 비교적 적은 경우에 국한되어 있어, 목조 문화재 등을 비롯한 구조물의 부재와 같이 부재의 일부 면만을 측정할 수 있는 경우에 대한 연구가 미비한 실정이다. 따라서 본 연구에는 이러한 초음파 시험기의 측정방법의 제한을 극복하는 방법을 모색하고자 그 첫 번째 단계로 결함을 갖는 목재 내에서 초음파 전달 경로를 구명하고자 하였다. 다양한 전달 경로 모델을 가정하고 실제 실험과 비교를 통해 초음파 전달의 가장 적합한 모델을 제시하고자 한다. 실험 시편에 인위적인 결함을 만들어 실험한 결과 초음파가 직접 전달될 수 없는 부분의 경우 그 결함을 피해서 전달되고 전달 경로는 탐촉자의 반대 측면에서 일차반사를 통해 전달되는 것으로 판단되었다.

• 한국건축시공학회:학술대회논문집
• /
• 한국건축시공학회 2020년도 가을 학술논문 발표대회
• /
• pp.161-162
• /
• 2020

In this study, as part of securing related data, water droplets were dropped on the upper part of the wooden board for flooring in an environment such as floor heating, and the degree of absorption according to the surface temperature change was tested. The test results showed that the contact angle of the surface was low (25℃→40℃) or the droplet was absorbed into the small plate and disappeared. The contact angle of the OSB and MDF was decreased within 30 minutes, but the surface water droplet was maintained longer than the plywood. This is because the surface is coated with hydrophobicity unlike the plywood, but moisture absorption in the cross section after the second processing will not be prevented and it will lead to defect occurrence problem.

• 공업화학
• /
• 제18권5호
• /
• pp.527-530
• /
• 2007

방염도료로 일액형 염화고무(Chlorinated Rubber) 백색도료가 많이 사용되고 있는데 이는 탄화시 검은 연기가 많이 발생되고 목재에 대한 표면강도와 접착력 등의 미흡한 단점이 있어 이러한 점 등을 고려한 Polyurethane Resin계(Alkyd Resin) 이액형 방염우레탄라카 백색도료를 개발하였다. 본 실험에서 개발한 목재용 Polyurethane Resin계 방염도료 조성의 물성은 기존의 방염백색도료보다 목재와의 친화력이 향상되고 다량의 연기발생은 억제하는 결과를 얻었다.

• Journal of the Korean Wood Science and Technology
• /
• 제14권4호
• /
• pp.29-39
• /
• 1986

To investigate the effect of presurfacing, both 25mm rough and pre surfaced oak boards (Quercus grosseserrata BI.) were dried in the same dry kiln condition. Drying curves, drying strains and drying defects of rough and presurfaced boards were compared. The results obatained are as follows. 1. Average drying rate of rough and presurfaced boards from green to 10% M.C. was 0.276%/hr. and 0.284%/hr., respectively. 2. At the early stage of drying, in case of rough boards, maximum tensile strain of outer slices was $-24.2{\times}10^{-4}$mm/mm and maximum compressive strain of innermost slices was $13.0{\times}10^{-4}$mm/mm, and in case of pre surfaced boards, maximum tensile strain of outer slices was $-14.5{\times}10^{-4}$mm/mm and maximum compressive strain of innermost slices was $28.1{\times}10^{-4}$mm/mm. And in both cases, stress reversal occurred at about 40% M.C.. 3. Maximum surface checking appeared at about 40% M.C.. Of the 10 rough boards. 6 hoards contained surface checks, but presurfaced boards did not contained surface checks after drying. And the results of end checking were similar to those of surface checking. But, honeycomb was not found in both cases. 4. Board shrinkage. warp and casehardening of presurfaced boards were similar to those of rough boards. But, collapse of prsurfaced boards was less than that of rough boards.

• Journal of the Korean Wood Science and Technology
• /
• 제22권2호
• /
• pp.61-70
• /
• 1994

This study was executed to find the applicability of press drying of tree disk by investigating the shrinkage and drying defect and to form appropriate model by comparing the actual moisture content(MC) and internal temperature in respect of drying time with calculated values based computer simulation to which was applied finite difference method. In press drying disk, heating period, constant drying rate period maintained plateau temperature at 100$^{\circ}C$ and falling drying rate period were significantly distinguished. Actual MC and internal temperature were analogous to those calculated at comparing points. Heat transfer model formed by Fourier's law using specific heat of moist wood and conduction coefficient considering fractional volume of each element of wood cell wall, bound water, free water and air showed applicability as basic data to developing heat expansion, shrinkage and drying stress during press drying. Also mass transfer model formed by Fick's diffusion law using water vapor diffusion coefficient showed applicability. Longitudinal shrinkage was developed by pressure of hot press and tangential shrinkage was restrained by hygrothermal recovery. The heart check, surface check and ring failure were occurred differently in species, but V-shaped crack didn't develop.

• Journal of the Korean Wood Science and Technology
• /
• 제18권1호
• /
• pp.39-51
• /
• 1990

This research was carried out to offer the basis data for development of optimum drying schedule for a domestic oak species (Quercus grosseserrata B1.) by investigating drying rate. stress, defect, and moisuture gradient with board thicknesses and drying schedules (code number T4-C2 and T3-B1). The results were obtained as follows: 1. Average drying rate and total drying time from 52.2% to 5.8% were 0.105%/hr and 486 hours for drying schedule T4-C2 and those from 62.1% to 8.3% were 0.070%/hr. and 811 hours for drying schedule T3-B1. 2. Drying rates for 28mm- and 31mm-thick boards showed similar tendency, but were significantly different from 25mm- thick board in drying schedule T4-C2 and those for 22mm-, 25mm- 28mm- and 31mm-thick boards showed similar tendency but were significantly different from 19mm- thick boards in drying schedule T3-B1. 3. The moisture gradients for drying schedule T4-C2 were steeper than those for drying schedule T3-B1 during drying period. and especially in early drying stage slow slope of moisture gradients of drying schedule T3-B1 was effective in preventing serious problem of surface checks. 4. Drying stresses were lower in drying schedule T3-B1 than in drying schedule T4-C2 during drying period. 5. Drying schedule T4-C2 was appropriate for 25mm-thick board but not for 28mm- and 31mm-thick board because of strong drying condition. Drying schedule T3-B1 was appropriate for 28mm- and 31 mm-thick board but not for 19mm-, 22mm-, and 25mm-thick board because of weak drying condition.

• 한국정보통신학회논문지
• /
• 제26권6호
• /
• pp.834-841
• /
• 2022

최근 스마트팩토리와 인공지능 기술의 수요 증가로 인해 다양한 분야에서 인공지능 기술을 적용하는 연구가 진행되고 있다. 결함 검사 분야에서도 인공지능 알고리즘을 도입하기 위한 노력을 기울이고 있다. 특히, 금속 외관의 결함을 검출하는 연구는 다른 소재(목재, 플라스틱, 섬유 등)의 결함을 검출하는 연구에 비해 많은 연구가 이루어지고 있다. 본 논문에서는 머신러닝 기법(서포터 벡터 머신(SVM: Support Vector Machine), 소프트맥스 회귀(Softmax Regression), 결정 트리(Decesion Tree))과 차원 축소 알고리즘(주성분 분석(PCA: Principal Component Analysis), 오토인코더(AutoEncoder))의 9가지 조합과 2가지 합성곱신경망(CNN: Convolutional Neural Network) 기법(자체 알고리즘, ResNet)의 금속 외관의 결함 분류 성능 및 속도를 비교하고 분석하는 연구를 수행하고자 한다. 두 종류의 학습 데이터셋((i) 공용 데이터셋(Public Dataset), (ii) 실측 데이터셋(Actual Dataset))에 대한 실험을 통해 각 데이터셋에 대한 성능 및 속도를 비교 분석하고, 가장 효율적인 알고리즘을 찾아낸다.