Conventional Radial Flux Coaxial Magnetic Gear (CRCMG) on the left and a Flux Angle Mapping (FAM) magnetic gear on the right. Both have the same gear ratio of 20:1. Notice the significantly thinner back irons on the FAM design.

Project Overview

Project statement:

  • Implement the “flux angle mapping” technique in a prototype and validate FEA results

  • FAM allows for a overall lighter magnetic gear due to the fact that compared to a conventional coaxial MG, the magnets on the inner and outer rotor are much closer in size allowing the back iron to be much thinner. This can be observed in the two designs to the left.

Main objectives:

  • Create a detailed electromagnetic FEA model in Ansys Maxwell to perform design optimization and performance simulations

  • Develop the mechanical design of the magnetic gear

    • This was particularly challenging because the of the shape and size of the ferromagnetic piece rotor (the flux angle “mappers”)

  • 3D print parts for the housing, support structures, and other pieces

  • Iterate on the mechanical design

  • Characterize the performance under via dynamometer tests and compare to FEA to validate the EM model

Challenges (ongoing so no results section):

  • The outer bridge dilema:

    • The ferromagnetic pieces rotor (FP rotor) consists of 91 flux angle “mappers”, each is a different size and shape. To get the lamination stack manufactured, a thin “bridge” was added to outside diameter to hold all the FPs together

    • The bridge successfully made the laminations manufacturable, however it significantly decreases the maximum torque of the gear box since it shorts out the magnetic flux from the magnets, so it must be removed

    • This creates the challenge of “suspending” the FPs between the rotor magnets. The gap between the rotors and the magnets is only 1mm on each side.

Potential solutions:

  • Injection molding

    • Would mold the whole FP rotor in some kind of plastic and then machine away the bridge

    • Issue is the very high initial cost of injection molding

  • 3D printing

    • As pictured, create “fingers” that would fill the negative space between the FPs and hold them in place. Again the bridge would be machined away after all the pieces are held.

    • Issue was the rigidity of the 3D printed parts was low and accurately getting a press fit between the fingers and the FPs was difficult

  • Resin/epoxy casting

    • Create a mold for the FP rotor and then cast the whole thing in epoxy resin to hold everything in place, and as with the other methods, machine the bridge away

    • First attempt failed on the last pass of machining due to the tool getting slightly caught on a protrusion on the inner diameter

    • Second attempt in progress with a different machining order and reinforced parts

A much more detailed explanation and analysis of the FAM concept can be found in my research paper. The paper can be accessed via IEEE Xplore or directly. The paper that has the prototype results and FEA model validation is currently a work in progress.

The 3D printed parts concept (top left), a cross section of the mechanical design using the epoxy resin method (top right), a video showing the idea behind the epoxy resin mold (bottom left) and the first failed attempt at machining the FP rotor (bottom right).

The inner permanent magnet (PM) rotor is shown in the top left, the outer PM stator is shown in the top right, the ferromagnetic piece (FP) rotor is on the bottom left and a flux map showing how the flux moves across the FPs is shown in the bottom right. All three of the rotors are made of hundreds of thin sheets of electrical silcon steel that are stacked and bonded together (called laminations). This done to reduce Eddy current losses.

Flux Angle Mapping (FAM) Magnetic Gear