An electromagnetic vibration energy harvester based on ring magnets with Halbach configuration: numerical modeling

Abstract

The power supply of electronic equipments is one of the major problems, during the construction of new structures. To provide this power supply, the main answer today is by building an electric network. However, this will increase the construction cost, and the technical maintenance, not to mention the cases where these networks will be impossible to setup. In response to these difficulties, autonomous power supplies like the electromagnetic vibration energy harvester started to be created. The electromagnetic vibration energy harvester is a set of three magnets, with the aim of obtaining an Halbach configuration, placed around a coil. The magnets will move when receiving vibration. Using this deplacement, the coil generates an electric current. The set of three magnets seems to be the most efficient system, but others parameters have to be taken into account, and research of other models keep on going. That’s why numerical modeling is used. This method allows us to observe different models, with the use of other parameters, to improve efficiency or study specific cases. The numerical modeling is based on two software : FEMM and EMS. The First, FEMM, creates 2D models, easy to handle, and can be driven by MatLab. The second, EMS, is an extension of SolidWorks or, in our case, Autodesk Inventor, that allows us to apply electromagnetic study to the 3D model of these softwares. At the start of this project, FEMM was already used to compare different models. But, the software allows only 2D and static studies. A change to EMS would allow 3D and dynamic models. That is why, the main goal of this project is learning how EMS works, the differences between the two software, and if an observation made on a software can be admitted on the other software. In a first time, a simple model, with only one magnet that replaced three other magnets with the same magnetic flux, was studied. The use of FEMM allowed to show different stages of this model, according to the distance between the magnet and its metal support. But the use of these simplified models on EMS did not provide me with similar results. The power of the magnetic flux was very low, and nothing was usable. And, the use of half model won't change anything. But, if we look at the model, composed of three magnets, each having a different polarization, the results are more similar to that of FEMM results. This observation allows us to deduce that EMS has to be used with complete models. All the data of FEMM can’t be transposed in an EMS model, because of this difference.