• Laser Solutions
• /
• v.3 no.1
• /
• pp.38-45
• /
• 2000

A 4㎾ RS840 CO2 laser with a powder auto-feeding apparatus has been used to deposit multiple overlapping tracks of Ni-base superalloy on to low carbon steel. It was found that the surface roughness(turbulence) of an overlapped cladding layer decreased with the increase of the overlapping ratio in an oscillating manner. When the overlapping ratio had values of 0.62, the surface turbulence was lowest. Overlapping ratio offer significant potential for improvement of materials surface properties such as corrosion performance and wear resistance. This paper reports that the overlapping ratio shows best corrosion resistance. The tensile residual stresses generated at the higher overlapping ratio( ＞ 0.45) and the element concentration of Fe increased in the surface layer at the lower overlapping ratio( ＜ 0.45) may lead to worse corrosion resistance.

• Journal of Advanced Marine Engineering and Technology
• /
• v.40 no.9
• /
• pp.768-773
• /
• 2016

Engine valve seat and face, which are the important factors affecting engine performance, are required to have wear, heat and corrosion resistance. In order to produce surface layer with these characteristics, PTA(plasma transferred arc) surfacing procedure is generally employed, but problems, such as large HAZ and high dilution etc., frequently occurr. Laser cladding, which overcomes the drawbacks of conventional technologies, can be employed to create a superior clad layer with low dilution, small heat affected zone, and minimal distortion. However, in case cladding is to be applied to a large area, it is necessary to overlap 1 pass clad layer because of limited clad layer width. Two criteria for the overlapping ratio-beam size and clad layer width-have been considered thus far. Upon inspection of multi pass clads, produced by different overlapping criteria, it was observed that the greater the increase in overlapping ratio, the greater was the decrease in clad layer width and increase in clad layer height regardless of the criterion used. However, a multi pass clad overlapped by the beam size criterion demonstrated a higher hardness value than a clad overlapped by the clad layer width owing to decreasing dilution of the substrate. In conclusion, the beam size was defined as the criterion for the overlapping, because the clad layer width increased or decreased depending upon process parameters.

• KIPS Transactions on Computer and Communication Systems
• /
• v.10 no.7
• /
• pp.191-198
• /
• 2021

Since the size of training dataset become large and the model is getting deeper to achieve high accuracy in deep learning, the deep neural network training requires a lot of computation and it takes too much time with a single node. Therefore, distributed deep learning is proposed to reduce the training time by distributing computation across multiple nodes. In this study, we propose hybrid allreduce strategy that considers the characteristics of each layer and communication and computational overlapping technique for synchronization of distributed deep learning. Since the convolution layer has fewer parameters than the fully-connected layer as well as it is located at the upper, only short overlapping time is allowed. Thus, butterfly allreduce is used to synchronize the convolution layer. On the other hand, fully-connecter layer is synchronized using ring all-reduce. The empirical experiment results on PyTorch with our proposed scheme shows that the proposed method reduced the training time by up to 33% compared to the baseline PyTorch.

• Journal of applied mathematics & informatics
• /
• v.36 no.1_2
• /
• pp.135-154
• /
• 2018

In this paper, an almost second order overlapping Schwarz method for singularly perturbed third order convection-diffusion type problem is constructed. The method splits the original domain into two overlapping subdomains. A hybrid difference scheme is proposed in which on the boundary layer region we use the combination of classical finite difference scheme and central finite difference scheme on a uniform mesh while on the non-layer region we use the midpoint difference scheme on a uniform mesh. It is shown that the numerical approximations which converge in the maximum norm to the exact solution. We proved that, when appropriate subdomains are used, the method produces convergence of second order. Furthermore, it is shown that, two iterations are sufficient to achieve the expected accuracy. Numerical examples are presented to support the theoretical results. The main advantages of this method used with the proposed scheme are it reduce iteration counts very much and easily identifies in which iteration the Schwarz iterate terminates.

• 한국정보디스플레이학회:학술대회논문집
• /
• 2009.10a
• /
• pp.1496-1499
• /
• 2009

A diode-pumped Nd:$YVO_4$ laser was used to obtain indium tin oxide (ITO) patterns on glass substrate with various overlapping rates. The results showed that the overlapping rate of laser beam influences on the edge structure of ITO pattern and the surface roughness of ablated groove bottom. At a laser repetition rate of 40 kHz, the optimized condition of overlapping rate was 75 %.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.25 no.5
• /
• pp.377-380
• /
• 2012

Laser direct patterning of indium tin oxide(ITO) is one of new methods of direct etching process to replace the conventional photolithography. A diode pumped Q-switched Nd:$YVO_4$ (${\lambda}$= 1,064 nm) laser was used to produce ITO electrode on various transparent oxide semiconductor films such as zinc oxide(ZnO). The laser direct etched ITO patterns on ZnO were compared with those on glass substrate and were considered in terms of the overlapping rate of laser beam. In case of the laser etching on double-layer, it was possible to obtain the higher overlapping rate of laser beam.

• ETRI Journal
• /
• v.28 no.6
• /
• pp.709-720
• /
• 2006

In this paper, we suggest a simple and efficient multiple-forwarder-based file distribution method which can work with a tree-based application layer multicast. Existing multiple-forwarder approaches require high control overhead. The proposed method exploits the assumption that receivers join a session at different times. In tree-based application layer multicast, a set of data packets is delivered from its parent after a receiver has joined but before the next receiver joins without overlapping that of other receivers. The proposed method selects forwarders from among the preceding receivers and the forwarder forwards data packets from the non-overlapping data packet set. Three variations of forwarder selection algorithms are proposed. The impact of the proposed algorithms is evaluated using numerical analysis. A performance evaluation using PlanetLab, a global area overlay testbed, shows that the proposed method enhances throughput while maintaining the data packet duplication ratio and control overhead significantly lower than the existing method, Bullet.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.19 no.12
• /
• pp.2813-2819
• /
• 2015

In this paper, we investigate the physical layer network coding (PNC) technique in a two-way relay channel (TWRC) where two source nodes send and receive data with each other via a relay node. In particular, we consider the communication scenario where packet length from the two sources is different from each other. We analyze the bit error rate (BER) of the received packet at the relay node according to degree of overlapping between two packets. The BER of the short packet remains unchanged regardless of the degree of overlapping since the entire packet is overlapped with the longer packet, while the BER of the longer packet becomes improved as the degree of overlapping decreases. Thus, we need a novel transmission scheme to enhance BER performance of the PNC technique in TWRC environments since the overall BER performance of the PNC technique at the relay node depends on the worse BER between two ovelapping packets' BERs.

• Fashion & Textile Research Journal
• /
• v.25 no.1
• /
• pp.119-126
• /
• 2023

This study aimed to determine the effects of relative humidity and fiber properties on the moisture permeability of multilayer systems by measuring water vapor transmission in the overlapping condition of various fabrics. The results confirmed that the property of the fabric in contact with the humid environment affects the moisture permeability. If the layer facing the humid environment is hydrophobic and the layer facing the dry environment is superhydrophobic, water vapor transmission increases by up to 17.8% compared to the opposite conditions. Comparing the correction values of the water vapor transmission reflecting the thickness of the specimen under the multilayer condition showed that permeability was higher when the hydrophilic or hydrophobic layer was facing the humid environment. The opposite was true from the "push-pull" effect of absorption mechanism. In the case of moisture permeability, the more hydrophilic the surface facing the humid environment, the more permeable that water vapor diffuses and passes through. It was concluded that the "pull-push" effect, in which water vapor diffuses widely through the hydrophilic facing a humid environment and then passes through the hydrophobic layer, contributes to the improvement of permeability. Permeability differed according to the multilayer overlapping condition. When the relative humidity was high, the "pull-push" effect was insignificant. This is caused by water droplets absorption after the partial migration of water due to condensation. These results suggest that the overlapping conditions and properties of fabrics should vary depending on heavy sweating or not.

• Journal of applied mathematics & informatics
• /
• v.36 no.3_4
• /
• pp.181-203
• /
• 2018

In this paper, we have constructed an overlapping Schwarz method for singularly perturbed second order convection-diffusion equations. The method splits the original domain into two overlapping subdomains. A hybrid difference scheme is proposed in which on the boundary layer region we use the central finite difference scheme on a uniform mesh while on the non-layer region we use the mid-point difference scheme on a uniform mesh. It is shown that the numerical approximations which converge in the maximum norm to the exact solution. When appropriate subdomains are used, the numerical approximations generated from the method are shown to be first order convergent. Furthermore it is shown that, two iterations are sufficient to achieve the expected accuracy. Numerical examples are presented to support the theoretical results. The main advantages of this method used with the proposed scheme is it reduces iteration counts very much and easily identifies in which iteration the Schwarz iterate terminates.