Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Mechanical Engineering

Committee Member

Dr. John Wagner, Committee Chair

Committee Member

Dr. Georges Fadel

Committee Member

Dr. Xiangchun Xuan

Committee Member

Dr. Todd Schweisinger


Advanced automotive engine cooling systems can positively impact the performance, fuel economy, and reliability of internal combustion engines. A smart engine cooling system typically features multiple real time computer controlled actuators: a three way linear smart valve, a variable speed coolant pump, and electric radiator fan(s). In this dissertation, several innovative comprehensive nonlinear control and optimization operation strategies for the next generation smart cooling application will be analyzed. First, the optimal control has been investigated to minimize the electric energy usage of radiator fan matrix. A detailed mathematical model of the radiator fan(s) matrix operation and the forced convection heat transfer process was developed to establish a mixed integer nonlinear programming problem. An interior points approach was introduced to solve the energy consumption minimization problem. A series of laboratory tests have been conducted with different fan configurations and rotational shaft speed combinations, with the objective to cool a thermal loaded engine. Both the mathematical approach and the laboratory test results demonstrated the effectiveness of similar control strategies. Based on the tests data and mathematical analysis, an optimization control strategy reduced the fan matrix power consumption by up to 67%. Second, a series of experimental laboratory tests were implemented to investigate the contributions of each electro-mechanical device in automotive thermal management system. The test results established a basis for several key operating conclusions. The smart valve and variable speed pump impacted the engine temperature by adjusting the heat transfer rate between the engine and the radiator through coolant redirection and/or coolant flow rate. On the other hand, the radiator fan(s) operation affects the engine's temperature by modifying the heat rejection rate of the radiator which can influence the entire cooling system. In addition, the smart valve's operation changes the engine's temperature magnitude the greatest amount followed by the radiator fan(s) and the coolant pump. Furthermore, from a power consumption aspect, the radiator fan(s) consumes the most engine power in comparison to the two other actuators. Third, a Lyapunov based nonlinear control strategy for the radiator fan matrix was studied to accommodate transient engine temperature tracking at heavy heat load. A reduced order mathematical model established a basis for the closed-loop real time feedback system. Representative numerical and experimental tests demonstrated that the advanced control strategy can regulate the engine temperature tracking error within 0.12°C and compensate the unknown heat load. The nonlinear controller provided superior performance in terms of power consumption and temperature tracking as evident by the reduced magnitude when compared to a classical proportional integral with lookup table based controller and a bang bang controller. Fourth, a nonlinear adaptive multiple-input and multiple-output (NAMIMO) controller to operate the smart valve and radiator fans has been presented. This controller regulates the engine temperature while compensating for unknown wide range heat loads and ram air effects. A nonlinear adaptive backstepping (NAB) control strategy and a state flow (SF) control law were introduced for comparisons. The test results indicated that the NAMIMO successfully regulated the engine temperature to a desired value (tracking error, |e|<0.5°C, at steady state) subject to various working conditions. In contrast, the NAB control law consumes the least radiator fan power but demonstrated a larger average temperature tracking error (40% greater than the NAMIMO controller), a longer response time (34% greater than the NAMIMO controller), and defected when the heat load was low. Lastly, the SF controller, characterized by greater oscillation and electrical power consumption (18.9% greater than the NAMIMO controller), was easy to realize and maintained the engine temperature to within |e|<5°C. An important aspect of engineering research is the knowledge gained from learning materials to fully understand the thermal management. As part of the dissertation, advanced three-dimensional (3D) visualization and virtual reality (VR) technology based engineering education methods has been studied. A series of computer aided design (CAD) models with storyboards have been created to provide a step to step guide for developing the learning modules. The topics include automotive, aerospace, and manufacturing. The center for aviation and automotive technological education using virtual e-schools (CA2VES) at Clemson University has developed a comprehensive e-learning system integrated with eBooks, mini video lectures, 3D virtual reality technologies, and online assessments as supplementary materials to engineering education.