Final Design

The final design resulted in non-aluminum stators with a reduced amount of coils and added liquid cooling reservoirs on each side. The main components are shown here:

        (A) 2 liquid cooling reservoirs

        (B) 2 stators with 6 coil inserts each

        (C) 1 stator connection/spacer

        (D) 1 aluminum rotor with magnets

        (E) 1 hall-effect sensor

        (F) 1 hall-effect sensor mount

        (G) 1 bike frame testing apparatus

General Changes

        ➜ Gap between stators increased by 33% to reduce chances of collision

        ➜ Stators reduced to 6 coils each due to manufacturing limitations (3D-print bed space)

        ➜ Hall-effect sensor mounted directly with stator for simplicity

PETG 3D-printed Stators and Reservoirs

        ➜ 3D-printing used for rapid prototyping capabilities at a low cost

        ➜ PETG is chemically resistant and non-electrically conductive

        ➜ PETG is a relatively sturdy thermoplastic polyester with a low thermal expansion coefficient

        ➜ Elimination of metal removes diminishing returns caused by eddy currents

Liquid Cooling Reservoirs

        ➜ Mineral oil is non-electrically conductive and safe for electronics

        ➜ Lined by O-ring, aiding leakage prevention

Electromagnetic Copper Ribbon Coils

        ➜ "Pancake" coil design used for radial field strength

                ➜ Bifilar coils reduce back EMF but have significantly reduced radial magnetic fields

        ➜ Wound inside aluminum inserts for easy removal, maintenance, and replacement

                ➜ Coil inserts previously directly glued to stators, creating a difficult maintenance process

Figure: Pancake vs Bifilar Coil wire directions

Figure: Pancake vs Bifilar Coil estimated magnetic fields 

PERFORMANCE RESULTS

As of June 6, the motor is not yet complete due to manufacturing and shipping delays, there are however concrete results on the performance of the pancake coils.

    Code was developed to predict the magnetic field generation for pancake coils of different geometries using numerical integration techniques of the biot-savart law. This code was found to be working after validating it with comparisons to an actual experimentally tested pancake coil. It is important to note that during experimentation current running through the coil was measured with a multimeter and magnetic field strength around the coil at various places was measured with a tesla meter.

    First off the simulation was tested against the experiment without any data-fitting for the number of wraps. The number of wraps and geometry of the coil was calculated from straight geometric measurements and great results from this method alone were achieved.

Figure: Z-Magnetic Field Strength at Center of Coil vs Current

    Next, the theory was fitted to the data to determine the true value of the number of wraps due to difficulties in accurately measuring the small thickness of the ribbon cable. A near-perfect fit resulted with a value of 158 wraps as opposed to the 153 wraps that were predicted from straight geometric calculations.

                                    Figure: Wrap-Fitted Z-Magnetic Field Strength at Center of Coil vs Current

    After fitting the simulation to the Z-components of the magnetic field due to the ease of measuring them precisely, the simulation was tested against the actual measured radial components of the magnetic field at a constant amperage with great results.

Figure: Wrap-Fitted Radial-Magnetic Field Strength vs Distance from Center

Lastly, an attempt was made to represent the loss in efficiency of the coil, as current was increased. Using the fitted simulation, the radial magnetic field components of the coil were calculated at different points in space for different currents. The ratio of magnetic field strength to heat generated of the coils was then calculated for all of these values. The strongest magnetic field per unit heat generated always occurs at the halfway point between the first and last wrap radii. Increasing current, generally decreases this efficiency ratio.