Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Mechanical Engineering

Committee Member

Dr. Gang Li, Committee Chair

Committee Member

Dr. Mohammed Daqaq

Committee Member

Dr. Lonny Thompson

Committee Member

Dr. Xiangchun Xuan


Computational modeling and performance analysis are carried out for a ferrofluid based electromagnetic energy harvester which converts ambient vibratory energy into electromotive force through sloshing motion of a ferrofluid. The system consists of a tank partially filled with ferrofluid, magnets placed on the opposite sides of the tank and a copper coil wound around the tank. In the presence of an external magnetic field, magnetic dipoles in the ferrofluid rotate and produce a net magnetic moment aligned in the direction of the field. When the device is subjected to an external excitation, the ferrofluid in the tank undergoes a sloshing motion which induces a time-varying magnetization in the fluid, causing a time-varying magnetic flux and electromotive force in the copper coil according to Faraday's law of induction. Compared to traditional solid-state vibratory energy harvesters, this liquid-state harvester provides better conformability, sensitivity, tunability and response bandwidth. This study provides useful insights for designing high performance ferrofluid based energy harvesters and is divided into three sections. First, A continuum level finite element model is developed and implemented for the multi-physics computational analysis of the energy harvester. The model solves the coupled magnetic scalar potential equation and Navier-Stokes equations for the dynamic behavior of the magnetic field and fluid motion. The model is validated against experimental results for eight configurations of the system. The validated model is then employed to study the underlying mechanisms that determine the electromotive force of the energy harvester. Furthermore, computational analysis is performed to test the effects of several modeling aspects, such as three-dimensional effect, surface tension and type of the ferrofluid-magnetic field coupling, on the accuracy of the model prediction. Second, a series of numerical simulations are performed to investigate the influence of several design parameters on the electromotive force of the energy harvester. From the eight configurations used for model validation, two configurations that give the highest electromotive forces are chosen for further performance analysis. The design parameters considered in this investigation include the device's geometric parameters, external excitation amplitude and material properties of the ferrofluid, which affect either the magnetic flux in the device or the sloshing behavior of the ferrofluid. Third, non-equilibrium molecular dynamics (NEMD) simulations are employed to obtain an understanding of the dynamic magnetization behavior of the ferromagnetic nano-particles and microscopic structures of the ferrofluid. The results from the continuum level numerical simulations reveal that the magnetic susceptibility/magnetization of ferrofluid greatly influences the performance of the energy harvester. Since the ferrofluid in the energy harvester undergoes sloshing motion under external mechanical excitations, it is also expected that fluid motion would significantly influence the aggregation behavior of the nano-particles, thereby playing an important role in determining the magnetization of the ferrofluid and the performance of the energy harvester. In this study, ferrofluid systems containing both small and large particles under the influence of both magnetic field and shear flow are considered. The computational model involves long-range dipolar interaction as well as short-range repulsive interaction of the nano-particles. The factors investigated include solvent friction coefficients, particle size, magnetic field strength and direction, and shear rate.