Flywheel Energy Storage System

The goal of this project is to develop math models, that with certain inputs will calculate needed dimensions and values for both the electrical and mechanical aspects of a stator for a high speed flywheel.

Background
One of the problems a potential lunar mission faces is energy generation and storage. While both solar and nuclear power generation are options, they do have drawbacks. Generating electricity using solar power would not be possible during the approximately 14-day lunar night, and nuclear power generation would have to decrease during the lunar day due to heat. As a result of these drawbacks, an efficient method of storing energy would be needed to ensure the success of any planned lunar mission. The University of Idaho has put forward the idea of using a FES system (FESS) to accomplish this. FES is advantageous due to its high energy density and low maintenance compared to other methods of energy storage.

The previous years, a low speed FESS was manufactured, but now the research team has started to focus on developing a high speed FES. The goal is to store 100x more energy than the low speed version. Whether this is still feasible or not, is still being determined due to a plethora of aspects, such as material property constraints within the flywheel, how quickly it can be accelerated, and numerous other factors. The purpose behind this senior project is to develop math models that will calculate various stator perimeters for both the electrical and mechanical aspects. Once the research team is able to determine the constrains for the flywheel, these will but plugged into our models to produce the required stator.

Stator
The UIFESS stator contains 24 coils divided into 4 poles. A voltage is supplied to these coils in order to generate a strong enough magnetic field to induce a force on the flywheel.

Rotor
The UIFESS rotor is the flywheel itself, and contains 4 salient poles to allow effective rotation.

Stabilization Bearing
The top bearing purely acts to provide corrective forces to the flywheel in order to maintain a 1 mm air gap.This 8 pole dedicated AMB was added to the original UIFESS design in order to create restoring moments on the rotor.

Self-Bearing Machine (SLFBM)
The SLFBM portion of the FESS reduces the hardware complexity of the system as well as the mass and size by removing the need for a dedicated AMB. Instead, the SLFBM acts to both rotate and stabilize the flywheel. However, combining these functions into one component increases the complexity of controlling the FESS. The SLFBM portion of the FESS takes the form of a Field-Regulated Reluctance Machine (FRRM).

Control Systems
At the time of writing, three distinct control systems were needed to ensure stability of the FESS: position, velocity, and current control. When the project was passed to the Flyrollers Senior Design Team, only the position control system was fully developed.

Power Electronics
Multiple Pololu H Bridges are used to apply a voltage to the FRRM coils.

Vacuum Chamber
To eliminate windage losses in the FESS, the entire apparatus is contained within a vacuum chamber. All wiring between the FESS and outside environment thus needed to be run through vacuum couplings in order to preserve the seal and maintain this vacuum.

Problem Statement
The UIFESS was not capable of rotation when the Flyrollers team took over the project. A rotational control algorithm needed to be developed and implemented in an embedded system in order to solve this problem. The flywheel also needed to be reconstructed and several purchases made to complete the test setup.

Specifications
High Priority: Target Velocity: 2500 RPM Time to Reach Target: 1 day Maximum Steady-State Offset: 2 rad/s Minimum Allowable Air Gap: 0.5 mm Acceleration Control: Target Acceleration Rate set by User

Medium Priority: Energy Efficiency: 70% During Acceleration Maximum Coil Temperature: 100 C Maximum Velocity Overshoot: 10 rad/s Maximum Settling Time: 1 s Maximum Rise Time: 10 s

Design Development
As stated, the Flyrollers were responsible for implementing current, position (displacement), and velocity control schemes for the flywheel. The position control schemes utilized PID controllers implemented digitally, while the current control was implemented using a simple hysteresis band.

Current Control
In order to generate the desired current through the coils of the FRRM portion of the FESS, a control scheme needed to be developed to monitor the most recent current level and send it to one of the microcontrollers. The microcontroller would then compare the measured value to a desired reference as determined by an encoder selected by the Flyrollers team. It was decided to use a simple hysteresis band control scheme to accomplish this, where the measured current value would be compared to the desired current value and the coil voltage polarity reversed or left alone accordingly.

Position Control
The lateral displacement of the flywheel must constantly be monitored, and coil current values adjusted in order to maintain the desired gap between rotor and stator. In order to ensure a smooth output response with minimal overshoot and high stability, a PID control scheme was used.

Encoder
An encoder needed to be purchased which would allow tracking of the flywheel position within a rotational frame of reference. The encoder was a key component in implementing the current control algorithm, as it established the position of each coil with respect to the rotor which determined what current value the coil needed to be operating at.

Power Supply
For testing purposes, a power supply needed to be specified and purchased which would provide the voltage to the coils in both the FRRM and SB portion of the FESS. The power needs of the UIFESS were calculated to be 51.76 A at 24 V, giving a total typical consumption of 1.242 kW.

Testing
Simulations performed in MATLAB indicated the current control algorithm would perform to an acceptable level, though with a moderate degree of overshoot and some ripple in the steady state response due to the constant voltage switching. Another key factor which needed to be tested for was the how quickly the current in a coil could be changed, in order to ensure that the current could be changed quickly enough to achieve the desired state before the rotor had moved on. Simulations indicated that even at the ideal speed of 2500 RPM the current could be changed quickly enough to avoid issues.

Team Bios
The 'Fly Rollers' team contributed to this project from Fall 2015 - Spring 2016 and was primarily responsible for wiring the FES system and getting it spinning in a controlled manner. The team consisted of the following members: