• 한국분무공학회지
• /
• 제27권2호
• /
• pp.57-65
• /
• 2022

The purpose of this study is to use machine learning to build a model capable of predicting the flash boiling spray characteristics. In this study, the flash boiling spray was visualized using Shadowgraph visualization technology, and then the spray image was processed with MATLAB to obtain quantitative data of spray characteristics. The experimental conditions were used as input, and the spray characteristics were used as output to train the machine learning model. For the machine learning model, the XGB (extreme gradient boosting) algorithm was used. Finally, the performance of machine learning model was evaluated using R² and RMSE (root mean square error). In order to have enough data to train the machine learning model, this study used 12 injectors with different design parameters, and set various fuel temperatures and ambient pressures, resulting in about 12,000 data. By comparing the performance of the model with different amounts of training data, it was found that the number of training data must reach at least 7,000 before the model can show optimal performance. The model showed different prediction performances for different spray characteristics. Compared with the upstream spray angle and the downstream spray angle, the model had the best prediction performance for the spray tip penetration. In addition, the prediction performance of the model showed a relatively poor trend in the initial stage of injection and the final stage of injection. The model performance is expired to be further enhanced by optimizing the hyper-parameters input into the model.

• 한국가시화정보학회지
• /
• 제20권2호
• /
• pp.38-44
• /
• 2022

Due to increasingly strict emission regulations for carbon-based fuels in the shipping industry, there is a significant motivation to investigate the alternative fuel. Ammonia is one of the attractive alternative fuels as a carbon-free fuel. Since ammonia has different properties such as high vapor pressure and low boiling point compared to conventional fuels, further research into ammonia spray behavior is important. In this work, the spray characteristics of ammonia and other fluids (ethanol, n-decane) were compared by using numerical simulation. The results show that the spray characteristics of ammonia differs from those of the others due to the occurrence of flash boiling. The narrow-dispersed spray with accelerated velocity at the center have been observed for ammonia. It is also found that droplets of ammonia achieve smaller diameter with more uniform distribution, leading to better atomization behavior compared to the others.

• 한국안전학회지
• /
• 제13권4호
• /
• pp.192-198
• /
• 1998

MRSM(modified response surface methodology)-2 model is presented for the prediction of boiling points in combustible solution of quaternary systems. This model requires only normal boiling points of pure substances and group-group parameters which are based on the group-group concepts without the use of experimental data under consideration. By means of this methodology, it is possible to predict the boiling points of the combustible mixture of quaternary systems by plotting of isothermal lines using computer graphics. The proposed methodology has been tested and compared successfully with reported boiling points in journals for the combustible solution of quaternary systems. It is hoped eventually that this methodology will permit prediction of the flash point and flammability limit for the combustible mixture of multicomponent systems.

• Journal of Advanced Marine Engineering and Technology
• /
• 제28권3호
• /
• pp.466-474
• /
• 2004

The characteristics of liquefied butane spray are expected to be different from conventional diesel fuel spray, because a flash boiling spray is expected when the surround pressure is below the saturated vapor pressure of the butane(0.23MPa 98K). The axial velocities. radial velocities. and size distributions in butane sprays were measured with PDPA(Phase Doppler Particle Analyzer) system. Sprays were macroscopically observed by using the high speed camera in case that the surround pressure is 0.37MPa and 0.15MPa. respectively. Compared with the conventional spray. the reversed results were investigated when the surround pressure is below the saturated vapor pressure of the butane.

• 한국분무공학회지
• /
• 제5권1호
• /
• pp.30-40
• /
• 2000

The characteristics of liquefied butane spray are expected to be different from conventional diesel fuel spray, because a kind of flash boiling spray is expected when the back pressure is below the saturated vapor pressure of the butane(0.23MPa at 298K). The ambient pressure was held at a pressure above(0.37MPa) and below(0.15MPa) the fuel vapor pressure. The axial velocities, radial velocities, and size distributions in butane sprays were measured with PDPA(Phase Doppler Particle Analyzer) system. The PDPA measurement showed a smaller SMD at the 0.15MPa chamber pressure, compared to the 0.37MPa case. Log-hyperbolic density function for the droplets size distribution can be fitted to the experimental results of a liquefied butane spray.

• 한국추진공학회:학술대회논문집
• /
• 한국추진공학회 2017년도 제48회 춘계학술대회논문집
• /
• pp.709-712
• /
• 2017

극저온 액체 질소가 초임계 상태에서 분사될 때 역광기법(backlight method)을 이용하여 사진을 얻었다. 이 때, 유체의 온도를 함께 측정하였고, 이를 바탕으로 가짜 감압비등이라는 새로운 분열기구를 제안하고자 하였으나, 역광기법은 정성적인 분무의 외관은 보여주지만 분사되는 유체의 큰 밀도로 인하여 밀도 변화를 보여줄 수 없었다.

• 한국자동차공학회논문집
• /
• 제11권5호
• /
• pp.22-28
• /
• 2003

Flash boiling mechanism in the injector interferes with fine fuel metering in a liquid phase LPG injection engine. This study presents a mathematical model to precisely predict an injection quantity. A calibration procedure of injection quantity, which is very prompt and precise in measuring, is developed using a gas analyzer. According to this procedure, injection quantity can be obtained under various fuel compositions, temperatures and injection pressures. The release pressure of liquid phase LPG is estimated based on these experimental data. Although the release pressure is much lower than the saturation pressure, it is linearly proportional to the saturation pressure.

• 한국화재소방학회논문지
• /
• 제19권4호
• /
• pp.26-31
• /
• 2005

화학공정에서의 안전하고 최적화된 조작과 내재되어 있는 화재 및 폭발 위험성 평가를 위해서 연소 특성치를 알아야 한다. 폭발한계는 가연성물질의 화재 및 폭발위험성을 결정하는데 주요한 특성치 가운데 하나이다 본 연구에서, 알코올류의 폭발한계에 대해 용액론을 근거로 표준끓는점과 인화점을 이용하여 예측하였다. 제시된 예측식에 의한 예측값은 문헌값과 적은 오차범위에서 일치하였다. 제시된 방법론을 사용하여 다른 가연성 물질의 폭발한계 예측이 가능해졌다.

• 한국응용과학기술학회지
• /
• 제8권2호
• /
• pp.145-151
• /
• 1991

Hazadous properties were evaluated for Alkyl nitrates such as hexyl nitrate, decyl nitrate, dodecyl nitrate and 2-methyl pentyl nitrate, 2-hexyl ethyl nitrate. The thermochemical properties such as heat of vaporization, boiling point, flash point and kinetic parameter for aliphatic nitrate were measured to determine the hazadous properties of these compounds. The boiling points and heat of vaporization increase as the increase of alkyl chain length in alkyl nitrates. Flash point is a linear function of boiling point as same as alkanes. The rate equation in isothermal decomposition are 1/2 order and compensation effect is found between logarithm of frequency factor the activation energy, then the decomposition preceeds with simlar reaction mechanism for each nitrate.

• 한국안전학회지
• /
• 제22권5호
• /
• pp.84-89
• /
• 2007

폭발하한계는 가연성물질의 화재 및 폭발 위험성을 결정하는데 사용되는 중요한 연소특성치의 하나이다. 본 연구에서 에스테르 화합물에 대한 폭발하한계는 액체 열역학이론을 근거로 표준끓는점과 인화점을 이용하여 예측하였다. 그 결과, 문헌값과 예측값의 A.A.P.E.(average absolute percent error)는 8.80vo1%이고, A.A.D.(average absolute deviation)는 0.18vo1% 그리고 상관계수는 0.965로써 문헌값과 예측값은 일치하였다. 제시된 방법론 사용에 의해 다른 가연성물질의 폭발하한계 예측이 가능하다.