石油炼制与化工 ›› 2022, Vol. 53 ›› Issue (5): 95-102.

• 控制与优化 • 上一篇    下一篇

润滑油基础油黏度调合模型的优化及应用

张安贵1,王汉文2,李艳1,于晓文2,梁雪美1,王恒2,燕艺楠1,郑舒丹1   

  1. 1. 国家能源集团宁夏煤业有限责任公司
    2. 长安大学汽车学院
  • 收稿日期:2021-09-24 修回日期:2021-10-31 出版日期:2022-05-12 发布日期:2022-04-24
  • 通讯作者: 王汉文 E-mail:834317176@qq.com
  • 基金资助:
    煤基高端润滑油基础油关键技术研究与示范;陕西省自然科学基金项目;中央高校基本科研业务费

OPTIMIZATION AND APPLICATION OF VISCOSITY BLENDING MODEL FOR LUBE BASE OIL

  • Received:2021-09-24 Revised:2021-10-31 Online:2022-05-12 Published:2022-04-24

摘要: 研究了天然气合成润滑油基础油(简称基础油)与其他3种基础油的混合油黏度调合数学模型的准确度和适用性。测定了天然气合成基础油GTL420分别与合成基础油PAO10、矿物基础油500N和煤基费-托合成基础油CTL10按不同比例调合的混合基础油在40 ℃和100 ℃时的运动黏度,对黏度调合模型进行优化。基于偏差率、均方根误差、残差平方和以及相关系数对模型进行评价。结果表明:Arrhenius方程在3种调合体系中的偏差率最大,不能精确描述混合基础油的真实运动黏度;Lederer-Roegiers Sr方程和Grunberg-Nissan方程表现出相似的模拟效果,但在不同调合体系中的有效性不同;Oswal-Desai方程的结构较复杂,但准确度和适用性是最优的。

关键词: 天然气, 合成基础油, 混合基础油, 黏度调合模型, 优化

Abstract: The accuracy and applicability of the mathematical model for viscosity blending of synthetic lubricant base oils (referred to as base oils),sourced from natural gas,with blends of other three base oils were investigated. The kinematic viscosities of the blends of synthetic base oil GTL420 with synthetic base oil PAO10, mineral base oil 500N and coal-based Fischer-Tropsch synthetic base oil CTL10 in different proportions at 40 ℃ and 100 ℃ were measured, and the viscosity blending model was optimized. The models were evaluated based on the deviation rate, root mean square error, sum of squared residuals, and correlation coefficient. The results show that the Arrhenius equation has the largest deviation rate among the three blending systems and cannot accurately describe the true kinematic viscosity of the blended base oil; the Lederer-Roegiers Sr equation and the Grunberg-Nissan equation exhibit similar simulation effects, but have different validity in different blending systems; the Oswal-Desai equation has a complex structure, but its accuracy and applicability are optimal.

Key words: natural gas, synthetic base oil, blended base oils, viscosity blending model, optimization