共 61 条
Navigating the effect of tungsten oxide nano-powder on ethylene glycol surface tension by artificial neural network and response surface methodology
被引:22
作者:
Ali, Vakkar
[1
]
Ibrahim, Muhammad
[2
]
Berrouk, Abdallah S.
[2
,3
]
Algehyne, Ebrahem A.
[4
]
Saeed, Tareq
[5
]
Chu, Yu-Ming
[6
,7
]
机构:
[1] Majmaah Univ, Coll Engn, Dept Mech & Ind Engn, Al Majmaah 11952, Saudi Arabia
[2] Khalifa Univ Sci & Technol, Mech Engn Dept, Sas Al Nakhl Campus,POB 2533, Abu Dhabi, U Arab Emirates
[3] Khalifa Univ Sci & Technol, Ctr Catalysis & Separat, POB 127788, Abu Dhabi, U Arab Emirates
[4] Univ Tabuk, Fac Sci, Dept Math, POB 741, Tabuk 71491, Saudi Arabia
[5] King Abdulaziz Univ, Fac Sci, Dept Math, Nonlinear Anal & Appl Math NAAM Res Grp, POB 80203, Jeddah 21589, Saudi Arabia
[6] Huzhou Univ, Dept Math, Huzhou 313000, Peoples R China
[7] Changsha Univ Sci & Technol, Hunan Prov Key Lab Math Modeling & Anal Engn, Changsha 410114, Peoples R China
基金:
中国国家自然科学基金;
关键词:
Surface tension;
WO3/EG nanofluid;
Experimental study;
RSM;
ANN;
THERMAL-CONDUCTIVITY;
RHEOLOGICAL BEHAVIOR;
HEAT-TRANSFER;
NANOPARTICLES;
NANOFLUID;
VISCOSITY;
HYBRID;
PERFORMANCE;
MODEL;
ANNS;
D O I:
10.1016/j.powtec.2021.03.043
中图分类号:
TQ [化学工业];
学科分类号:
0817 ;
摘要:
The quiddity of ethylene glycol (EG) surface tension is attributed to the cohesive forces between EG molecules. The presence of tungsten oxide (WO3) nanoparticles certainly affects the cohesive forces and the surface tension will be affected. In this study, SITA dynotester was utilized for measuring surface tension at 20-50 degrees C and 0.005-5 wt%. Loading nanoparticles intensified the adhesive force between EG and WO3 molecules and simultaneously attenuated cohesive forces between EG molecules. According to the results, the positive effect of the adhesive force was overcome by the negative effect of reducing the adhesion force and therefore the surface tension was weakened. At 20 degrees C the EG surface tension was attenuated up to 15.57% by adding WO3 nanoparticles (5 wt%). Performing response surface methodology and artificial neural network for navigating the surface tension was successful. The acceptable R-squared for RSM (0.982) and ANN (0.99) led to maximum error of 1.43% and 0.95%. (C) 2021 Elsevier B.V. All rights reserved.
引用
收藏
页码:483 / 490
页数:8
相关论文