• ETRI Journal
• /
• v.44 no.5
• /
• pp.849-858
• /
• 2022

Dynamic voltage frequency scaling (DVFS) has been widely adopted for runtime power management of various processing units. In the case of neural processing units (NPUs), power management of neural network applications is required to adjust the frequency and voltage every layer to consider the power behavior and performance of each layer. Unfortunately, DVFS is inappropriate for layer-wise run-time power management of NPUs due to the long latency of voltage scaling compared with each layer execution time. Because the frequency scaling is fast enough to keep up with each layer, we propose a layerwise dynamic frequency scaling (DFS) technique for an NPU. Our proposed DFS exploits the highest frequency under the power limit of an NPU for each layer. To determine the highest allowable frequency, we build a power model to predict the power consumption of an NPU based on a real measurement on the fabricated NPU. Our evaluation results show that our proposed DFS improves frame per second (FPS) by 33% and saves energy by 14% on average, compared with DVFS.

• KEPCO Journal on Electric Power and Energy
• /
• v.8 no.2
• /
• pp.111-118
• /
• 2022

SF₆ is used as an insulating gas because of its excellent electrical insulation properties, non-toxicity, and non-inflammability. On the other hand, the global warming potential of SF₆ is 23,900 times higher than that of CO₂. The Korea electric power cooperation (KEPCO) is responsible for 80% of the domestic SF₆ usage, and approximately 6,000 tons are currently charged in electrical and power facilities. KEPCO will gradually replace the insulating gas with SF₆-free gas from 2023. SF₆ decomposition facilities are required because more than 60 tons of SF₆ will need to be disposed of annually from existing equipment. This study developed a novel decomposition and pollution control system that can process 60 tons of SF₆ per year. This facility can decompose more than 97.7% of SF₆, with the emissions of hazardous and toxic materials below the legal limit.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.27 no.5
• /
• pp.376-383
• /
• 2022

The improved performance of computer parts, such as graphic card, CPU, and main board, has led to the need for power supplies with a high power output. The dynamic load profile rapidly changes the usage of power consumption depending on load operations, such as PC power and air conditioner. Under dynamic load profile conditions, power consumption can be classified into maximum, normal, and standby power. Several problems arise in the case of maximum power. Peak power is generated at the system power source in the maximum-power situation. Frequent generation of peak power can cause high-frequency problems and reduce the life of high-pressure parts (especially high-pressure capacitors). For example, when a plurality of PCs are used, system overload occurs due to peak power generation and causes problems, such as power failure and increase in electricity bills due to exceeded contract power. To solve this problem, a system peak power limit/compensation power circuit is proposed for a power supply under dynamic load profile conditions. The proposed circuit detects the system current to determine the power situation of the load. When the system current is higher than the set level, the circuit recognizes that the system current generates peak power and compensates for the load power through a converter using a super capacitor as the power source. Thus, the peak power of loads with a dynamic load profile is limited and compensated for, and problems, such as high-frequency issues, are solved. In addition, the life of high-pressure parts is increased.

• Journal of Welding and Joining
• /
• v.26 no.3
• /
• pp.45-50
• /
• 2008

SM45C steel rods being used generally for power transmission shafts and machine components was selected and welded by Butt-GMAW(Gas Metal Arc Welding) method. An estimation of fatigue life was studied by constructing S-N curve. Fatigue strength of base metal zone showed higher values than one of weld zone in low cycles between $10^4$ and $10^6$cycles. However, significant decrease in fatigue strength of base metal was found around $10^6$cycles, which were almost same as one of heat affected zone. This decrease was attributed that initial residual stress of the steel rods distributed by drawing process was diminished by continually applied load, and resulted in softening of base metal. The fatigue limit of the weld zone was highest in the boundary between deposited metal zone and heat affected zone, and followed by in the order of deposited metal zone, base metal zone, and heat affected zone. Based on these results, it is revealed that the stress for safety design of machine components using SM45C butt-welded steel rods must be selected within the region of the lowest fatigue limit of heat affected zone.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.15 no.2
• /
• pp.175-181
• /
• 2007

The objective of this study was to develop formability evaluation techniques in order to apply aluminum sandwich panel for automotive body parts. For this purpose, newly adopting formability evaluation (using limit dome height and plane strain test) was carried out in order to secure the fundamental data for the measurement of sheet metal forming and the establishment of optimum forming conditions of the aluminum sandwich panel. The results showed that there were good agreements between the old formability evaluation method and the new method which was more simplified than that of old one. From the results of these formability evaluation, the formability of sandwich panel was higher than that of aluminum alloy sheet alone which was the skin component for the sandwich panel. Also, it was found that sandwich panel could reduce the weight and could have the same flexural rigidity simultaneously when it was compared to the automotive steel sheet.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.61 no.3
• /
• pp.485-490
• /
• 2012

Nowadays, in metropolitan railroad, DC feeding system is being generally applied. In order to reduce damage of electro-chemical corrosion caused by stray current and leakage current, in DC feeding system, rail is used as negative-polarity return conductor for traction load current. However, it has problem of rail potential increase and there are no adequate measures to prevent it in domestic. The rise of rail potential leads to damage for human and equipments. To solve the problems, this paper presents fundamental theory and related standards about rail potential increase. And then, we analyzed field testing data and simulated a variety of operations by using PSCAD/EMTDC as an analysis program of power system. In addition, this paper suggests rail potential limit device and addresses how to the device. To verify the effect, simulation of DC feeding system before and after the application of the device is carried out in various cases.

• Transactions of the Korean Society of Pressure Vessels and Piping
• /
• v.15 no.1
• /
• pp.18-27
• /
• 2019

The Advanced Power Reactor 1400 (APR1400) Steam Generator (SG) uses alloy 690 as a tube material and SA-508 Grade 3 Class 1 as a tubesheet material to form tube-to-tubesheet joint through hydraulic expansion process. In this paper, the residual stresses in the SG tube-to-tubesheet contact area was investigated by applying Model-Based System Engineering (MBSE) methodology and the V-model. The use of MBSE transform system description into diagrams which clearly describe the logical interaction between functions hence minimizes the risk of ambiguity. A theoretical and Finite Element Methodology (FEM) was used to assess and compare the residual stresses in the tube-to-tubesheet contact area. Additionally, the axial strength of the tube to tubesheet joint based on the pull-out force against the contact joint force was evaluated and recommended optimum autofrettage pressure to minimize residual stresses in the transition zone given. A single U-tube hole and tubesheet with ligament thickness was taken as a single cylinder and plane strain condition was assumed. An iterative method was used in FEM simulation to find the limit autofrettage pressure at which pull-out force and contact force are of the same magnitude. The joint contact force was estimated to be 20 times more than the pull-out force and the limit autofrettage pressure was estimated to be 141.85MPa.

• Journal of Nuclear Fuel Cycle and Waste Technology(JNFCWT)
• /
• v.18 no.4
• /
• pp.537-548
• /
• 2020

Currently, radioactive waste for disposal has been restricted to low and intermediate level radioactive waste generated during operation of nuclear power plants, and these radioactive wastes were managed and disposed of the 200 L and 320 L of steel drums. However, it is expected that it will be difficult to manage a large amount of decommissioning waste of the Kori unit 1 with the existing drums and transportation containers. Accordingly, the KORAD is currently developing various and large-sized containers for packaging, transportation, and disposal of decommissioning waste. In this study, the radiation exposure doses of workers and the public were evaluated using RADTRAN computational analysis code in case of the domestic on-road transportation of new package and transportation containers under development. The results were compared with the domestic annual dose limit. In addition, the sensitivity of the expected exposure dose according to the change in the leakage rate of radionuclides in the waste packaging was evaluated. As a result of the evaluation, it was confirmed that the exposure dose under normal and accident condition was less than the domestic annual exposure dose limit. However, in the case of a number of loading and unloading operations, working systems should be prepared to reduce the exposure of workers.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.3 no.3
• /
• pp.517-531
• /
• 1999

This paper presents the development of the 2-mode variable gain high power amplifier for a transmitter of INMARSAT-M operating at L-band(1626.5-1646.5 MHz). This SSPA(Solid State Power Amplifier) is amplified 42 dBm in high power mode and 36 dBm in low power mode for INMARSAT-M. The allowable error sets +1 dBm of an upper limit and -2 dBm of a lower limit, respectively. To simplify the fabrication process, the whole system is designed by two parts composed of a driving amplifier and a high power amplifier, The HP's MGA-64135 and Motorola's MRF-6401 are used for driving amplifier, and the ERICSSON's PTE-10114 and PTF-10021 are used the high power amplifier. The SSPA was fabricated by the circuits of RF, temperature compensation and 2-mode gain control circuit in aluminum housing. The gain control method was proposed by controlling the voltage for the 2-mode. In addition, It has been experimentally verified that the gain is controlled for single tone signal as well as two tone signals. The realized SSPA has 42 dB and 36 dB for small signal gain within 20 MHz bandwidth, and the VSWR of input and output port is less than 1.5:1 The minimum value of the 1 dB compression point gets 5 dBm for 2-mode variable gain high power amplifier. A typical two tone intermodulation point has 32.5 dBc maximum which is single carrier backed off 3 dB from 1 dB compression point. The maximum output power of 43 dBm was achieved at the 1636.5 MHz. These results reveal a high power of 20 Watt, which was the design target.the design target.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.39 no.4
• /
• pp.405-413
• /
• 2015

In the US, the number of cases of subterranean water contamination from tritium leaking through a damaged buried nuclear power plant pipe continues to increase, and the degradation of the buried metal piping is emerging as a major issue. A pipe blocked from corrosion and/or degradation can lead to loss of cooling capacity in safety-related piping resulting in critical issues related to the safety and integrity of nuclear power plant operation. The ASME Boiler and Pressure Vessel Codes Committee (BPVC) has recently approved Code Case N-755 that describes the requirements for the use of polyethylene (PE) pipe for the construction of Section III, Division 1 Class 3 buried piping systems for service water applications in nuclear power plants. This paper contains tensile and slow crack growth (SCG) test results for high-density polyethylene (HDPE) pipe welds under the environmental conditions of a nuclear power plant. Based on these tests, the fracture surface of the PENT specimen was analyzed, and the fracture mechanisms of each fracture area were determined. Finally, by using 3D finite element analysis, limit loads of HDPE related to premature failure were verified.