Modeling of natural convection induced oil temperature variation in an enclosure and single bubble dynamics in shear flow

• Main Author: Yang, Songyuan
• Other Authors: Li Hua
• Language: English
• Published: Nanyang Technological University 2015
• Subjects: DRNTU::Engineering::Mechanical engineering::Fluid mechanics; DRNTU::Science::Physics::Heat and thermodynamics
• Online Access: http://hdl.handle.net/10356/65298
• Institution: Nanyang Technological University

The long-distance transportation of crude oil using tankers requires a lot of fuel to heat the oil and maintain it at a constant temperature. As an important step to reduce the fuel consumption in the oil heating and heat preservation process, the present work focuses on the modeling for analysis of the oil temperature variation inside cargo oil tanks. Specifically, the work includes the two portions: (1) the oil temperature variation in an enclosure due to natural convection, and (2) the single bubble dynamics in shear flow, as detailed below. (1) In the first portion, the focus is on the oil temperature variation due to turbulent natural convection in a tank-like enclosure, which is the major heat transfer mechanism in oil tanks. The enclosure is partially filled with oil, with a nitrogen layer on top of the oil for safety reason. The related characteristics of the oil temperature variation are numerically studied using a commercial computational fluid dynamics (CFD) code, ANSYS Fluent (v14.0), where the effects of various parameters are determined as follows: In terms of nitrogen layer, it is an effective thermal insulation layer, which keeps the average oil temperature at a relatively large value and is necessary to be included in CFD simulations examining oil temperature variation in partially filled enclosures. Regarding wall thickness, it only has little influence on the oil temperature. It thus makes sense to employ a uniform wall thickness of 20 mm in the simulations.For number of heating surfaces, its influence on the oil temperature variation is trivial. It is reasonable to apply a single large heating surface in the simulations. As regards oil density, its increase leads to a dramatic decrease in the average oil temperature. A formula is developed to correlate the average oil temperature with the oil density, making it possible to extend the simulated temperature of a certain crude oil to other types of oil which have similar material properties except density. For oil viscosity, its increase results in a slight oil temperature increase. A formula is also proposed to describe the relationship between the average oil temperature and the oil viscosity. With this formula, it is convenient to predict the temperature of new types of crude oil which mainly differ from the present one in viscosity. 6. As well-known, CFD simulation is always time-consuming. In order to reduce the CPU time, only a small portion of the CFD simulation results are coupled with the lumped capacitance model to predict the oil temperature variation over a longer period of time, resulting in two hybrid models: constant model and linear model. The constant model is robust, accurate, and efficient, and is finally chosen to predict the time histories of the oil temperature in an industrial project. (2) In the second portion, single bubble dynamics is investigated in shear flow, since the presence of bubbles in crude oil possibly influences its rheological property and subsequently the temperature variation. Using an in-house code based on the immersed boundary method, the governing equations are numerically solved to study the behavior of single bubble, where the evolutions of bubble orientation and deformation in shear flow are elucidated. In addition, the effects of several important parameters are examined, including the capillary number, shear rate, and temperature gradient, as detailed below. A larger capillary number leads to a larger bubble deformation, a larger deformation speed, and a longer time for the deformation to reach steady state. A larger shear rate also leads to a larger bubble deformation, a larger deformation speed, and a longer time for the deformation to reach steady state. Regarding temperature gradient, it mainly influences bubble location in the absence of shear flow, and affects both bubble location and shape in the presence of shear flow. Without shear flow, the bubble moves in the temperature gradient direction, in which a larger temperature gradient results in a larger velocity. With shear flow, the bubble migrates in both the temperature gradient and the shear flow directions, and becomes non-centrosymmetric, in which a larger temperature gradient results in a longer migration distance, a smaller bubble deformation, a smaller deformation speed, as well as a shorter time interval for the deformation to reach steady state.