Customer Handoff & Final Project Documentation
Team Vision for Final Demo and Handoff
At the previous design review, a major shortcoming was discovered: after upgrading the size and strength of the levitation electromagnet, the Hall effect sensor intended to control magnetic levitation in the sample coupon was unable to function in the strong magnetic field. With this in mind, the team agreed to split the twin objectives of the project into two parts: 1) demonstrate feasibility of magnetic levitation of a sample coupon, and 2) measure CTE of a coupon sample. Thus, in order to demonstrate the fulfillment of these two objectives, validation testing was performed on both functions of the device individually. Without the Hall effect sensor or some other sensing device, it was not possible to accomplish both, since without reference position data on the top of the sample, it cannot be known if a change in the reading of the bottom position sensor was caused by expansion of the sample or oscillation of the coupon+clamp as a whole.
Engineering Requirements vs Performance
Project Bill of Materials and Upgrade Componentshere.
Since the budget and timeframe of the project did not allow for the best components to minimize error in the CTE test system, the following are presented as alternatives to minimize error. The effect of these upgrades on error minimization is shown in "Error Analysis" section below.
Test Results Summary
Test Plan S1: Electromagnetic Strength
Test Plans S2, S3, and S6: Control System Stability (Large Scale Prototype)
Test Plan S4: Sensor Calibration
Test Plan S5: Danger of Shock
Coefficient of Thermal Expansion TestingWith the sample coupon and clamp held in a static position in the frame's slot fixure, CTE testing was carried out on the sample. The thermal shroud was used to provide heat flux to the coupon, and the shroud temperature control thermocouple was connected to the coupon directly by aluminum tape. This is because controlling the shroud temperature directly caused the coupon to take too long to reach equilibrium temperature. As a downside, the thermal shroud experienced a much greater maximum temperature, causing a potential safety issue.
Temperature testing was carried out on the sample coupon, with distance data collected from the capacitive sensor. At this point it was noted that the capacitive sensor signal suffered from heavy noise, which was magnified when the heating shroud was active.
It was theorized that the noise was caused by the 120V 60Hz AC power from the wall socket that powered the thermal shroud. This is because the temperature controller only uses a "two-prong" connection to the wall socket, and is not grounded. Therefore even when the relay switch is open and the shroud does not pass current (and heat up), it still experiences a voltage potential fluctuation from the one connected terminal. (The exact mechanism that results in the noise being transferred to the capacitive sensor signal is unknown, however).In order to alleviate this noise, the sample coupon was allowed to heat up and reach a steady value, then the controller was unplugged and the sample was allowed to cool back down to ambient temperature. By disconnecting the shroud from the wall entirely, the noise witnessed in the cap sensor output was reduced substantially.
An additional thermocouple was placed on the capacitive sensor mounting flange, in order to demonstrate that the flange experiences insignificant temperature rise throughout the test.
Simultaneous TestA test was run with the magnetic levitation and heating systems active at the same time. Despite the noise induced from the heating shroud, the sample was still able to achieve stable levitation. However, as discussed previously, without the use of a second sensor at the top of the structure, it is not possible to subtract out rigid body motion that may occur as the sample heats and lengthens. Therefore, it is not possible to measure CTE while levitating the sample with the current setup.
Total error (by root-sum-squares method): 2.38 x10^-6 in/in/deg F
Total error (by root-sum-squares method): 7.55 x10^-7 in/in/deg F
Note that these error values, while oftentimes informed by the actual results obtained in the lab, are still theoretical. Unaccounted for/underrated error sources may arise in actual testing, especially as the sensitivity of the device increases.
The spreadsheet with error calculations can be found here.
Risk and Problem Tracking
Risk management remained unchanged from the previous phases, as "risks" were considered more as vague, conceptual problems, not concrete issues faced by the team.
Project Problem Tracking
Final Project Documentation
- Mechanical system drawings: here.
- Project technical paper: here.
- Project poster: here.
- Instructions for system operation: here.
- Future MSD project/Harris STI proposal: here.
Hypothetical Plans for Future WorkAction items the team would like to take, provided we had an additional three-week phase following this review (also goals for future iterations of this project):
- Acquire additional funds for component upgrades. Especially a microcontroller with better bit resolution and faster processing.
- Run simulations to reduce our "noise" by understanding what is truly causing it.
- Machine and run different coupon material samples.
- Consider a new support arm design.
- Attempt to evaluate more of the error sources.
- Redesign top plate: attempt to levitate with sensor attached to top instead of bottom
- Purchase and incorporate a second sensor to measure CTE and control maglev concurrently. Run more tests in this configuration.
- Begin redesign of heating system.
- Begin to test current maglev system under vacuum conditions.
- Begin design of PCB for final circuit for further noise and size reduction.
- Create housing for electrical components.
- Begin redesign phase considering higher end components and current design weaknesses.