P17341: Harris Near-Zero CTE Test Apparatus
/public/

Integrated System Build & Test

Table of Contents

Team Vision for Integrated System Build & Test Phase

Plans & Accomplishments

This phase was planned to finalize all component purchases, construct the entire final system, and begin testing. Purchases were finalized for all components, however, lead time of the crucial capacitive sensor component meant that the full system was not assembled in time for the end of this phase. However, in the meantime, all subsystems were fully assembled, and test plans were carried out to the extent possible on the unfinished system.

Team members' individual 3 week plans can be found here.

Updates from Subsystem Build & Test Review

Bill of Materials Update

Due to space constraints, a small snapshot of the team BOM is shown below. The full document may be accessed here.
BOM Snapshot

BOM Snapshot

The team also brainstormed a first selection of high-level components to be used to upgrade the final Harris device (with expanded budget) This lisk functions as a theoretical ideal list with no regard for price, lead time of components, etc.

CTE Test Fixture upgrade components

CTE Test Fixture upgrade components

System Update

Structure

Process Flow Diagram

The following diagrams show the progress from CAD to the machined part

Rod Clamps

The rod clamps were developed to solve the problem of securing the top plate at various heights on the composite rods. The design needed to be adjustable, strong, and not damage the rods when tightened. Compared to the original idea of a set screw, this rod clamp uses multiple points of contact to distribute the load. While this component was not part of the original design, we were able to utilize material in the machine shop and come up with this quick fix to our problem.
3D Rendering of rod clamp

3D Rendering of rod clamp

Rod clamp drawings

Rod clamp drawings

Photo of finished clamps

Photo of finished clamps

Support Rod Flange

The flanges were developed to solve the problem of securing the rods into the base plate. The original design press fitted the rods into the base plate with no additional support. The problem with this design is that the rods are not smooth on the outer surface, and this prevented a tight fit from occurring. To solve this problem, the flange is inserted into the base plate and inside the rod. Since the inside of the rod is smooth and round, a tight fit is possible and the rods remain secure.

3D Rendering of support rod flange

3D Rendering of support rod flange

Support rod flange drawing

Support rod flange drawing

Photo of support rod flange

Photo of support rod flange

Clamp

A critical part to our design, the coupon clamp serves multiple purposes. It must be able to secure a coupon sample, accommodate varying coupon thicknesses, and be levitated by the electromagnet. The design also utilizes a pin locating feature to ensure each test coupon is installed in the same spot before testing.
3D Rendering of clamp

3D Rendering of clamp

Photo of clamp

Photo of clamp

Final Design Update

Design changes were made as we began machining parts and assembling our design. Flaws were found and additional parts were built to solve our problems. Our current test structure is shown below.

Structure without and with thermal shroud

Structure without and with thermal shroud

Thermal Shroud

While individual subcomponents were tested last phase, the thermal shroud subsystem was fully assembled and integrated into the total CTE testing structure as an integrated system. The shroud consists of two halves of a cylinder of polyurethane foam insulation, each with a resistive "heater tape" attached to the inner surface. The heaters are attached to the insulation via aluminum tape, which also functions to spread the heat across the entire inner surface of the shroud and provide more uniform heat flux to the sample coupon. One half of the thermal shroud is attached to the baseplate of the CTE tester with three dowel pins, and the other half attaches directly to the first. This was done as a trade-off between attaching the shroud rigidly and being able to observe the sample as it levitates.
Half of shroud installed in CTE tester frame

Half of shroud installed in CTE tester frame

Additionally, to improve portability and ease of assembly, terminal plugs were attached to the power and thermal tape connections on the thermal tape controller. The "jaw" type electrical connections on the thermal controller are very stable, but require a screwdriver to adjust, which will not be feasible once all electrical subcomponents are housed in a case. However, it is easy to make and break connections as necessary on the plugs, and the wires can be adjusted to protrude from the case. This also allows each half of the thermal shroud to function independently of the other.

Power and half of tape connected to controller

Power and half of tape connected to controller

Integrated System

Plans were also put in place for the general shape of the final system. This includes a box to house all electrical components and provide a unified place for all power and data connections.
Fully assembled electrical box

Fully assembled electrical box

Exploded view of electrical box

Exploded view of electrical box

A workstation footprint was drafted to provide a rough visual of the total benchtop space needed to house all components. Note that the laboratory power supply and PC are not included in the CTE design that will be provided to Harris, but any USB-compatible PC and 12V multi-channel power supply will suffice.

Total workstation space for CTE test equipment

Total workstation space for CTE test equipment

Test Results Summary

Structural Strength Testing

System load test information

System load test information

System load test results

System load test results

Clamp strength test information

Clamp strength test information

Clamp strength test results

Clamp strength test results

Thermal Shroud Output Testing

Motivation The long term behavior of the thermal shroud is not yet understood. In order to identify and quantify any errors in the system, the temperature at the center of the shroud must be observed.

Procedure The shroud was tested for 30 minute intervals, with three thermocouples, T1, T2 and T3 placed at the center of the shroud. The controller was set to 90F and the unit was allowed to heat from an ambient temperature of 70F. The shroud was allowed to rest between tests until all thermocouples read ambient temperature.

Arrangement of thermocouples on shroud

Arrangement of thermocouples on shroud

Test results for shroud testing

Test results for shroud testing

Results Several tests were performed, with each following the same pattern as shown above. An initial overshoot is observed when rising from ambient temperature to the controller's setpoint. Each peak after this appears to have the same magnitude.

Observations Although the temperature seemed uniform with respect to height, it is not constant with time, following a cyclical pattern. It is also important to note that the temperature inside the shroud is lower than the thermal tape controller setpoint.

Conclusion The sample could experience a temperature fluctuation of about 4F over the course of testing. The shroud appears to experience steady state behavior after the first initial spike, so a minimum of time is required for sample testing to begin. If another setpoint is chosen, this test should be repeated to understand the temperature dynamics for that setpoint.

Structure Heating Testing

Motivation Expansion of the structure could cause error in sample measurements, due to the change in position of the measuring sensors. The temperatures of these components should be monitored to quantify this error.

Procedure Two thermocouples, T2 and T3 were placed on the composite support rod, top plate and bottom plate during operation of the thermal shroud at a setpoint of 90F. One thermocouple, T1, was placed inside the shroud to monitor the heating cycle and how it affected the component temperatures. The tests were run for 15 minutes, and the system was allowed to return to room temperature between tests.

Thermocouple arrangement on composite rod

Thermocouple arrangement on composite rod

Thermocouple arrangement on top plate

Thermocouple arrangement on top plate

Thermocouple arrangement on sensor flange

Thermocouple arrangement on sensor flange

Thermocouple arrangement on clamp

Thermocouple arrangement on clamp

Test results for composite rod

Test results for composite rod

Test results for top plate

Test results for top plate

Test results for flange

Test results for flange

Test results for clamp

Test results for clamp

Results Despite cyclical heating of the shroud, only the sensor flange experienced any heating.

Observations It should be noted that the set point is relatively low at 90F. It could be possible at much higher temperatures for some of the components to experience some degree of heating.

Conclusion The expansion of the major structural components (with the exception of the flange, see below analysis) should be neglected from the error analysis, since it does not appear that any of them experience a measurable change in temperature. However, if the setpoint is increased significantly from 90F, then this test should be repeated to confirm the results at the new setpoint.

Thermal expansion of flange in ANSYS

Thermal expansion of flange in ANSYS

Capacitive Sensor Test Plan

Due to the long lead time and late arrival of the capacitive sensor, verification of its calibration could not be completed in time for this review. However, the plan that will be followed to verify the sensor is shown below.
Capacitive sensor calibration verification

Capacitive sensor calibration verification

Manufacturer provided cap sensor calibration curve

Manufacturer provided cap sensor calibration curve

Risk and Problem Tracking

Overall Problem Tracking

Critical problems

Critical problems

Major project problems

Major project problems

Major project problems (continued)

Major project problems (continued)

Ordinary problems

Ordinary problems

Hall Sensor Problem Tracking

Problem Statement

Hall Effect sensor reaches saturation limits prior to magnet levitating the clamp.
Operating limits of Hall Effect sensor

Operating limits of Hall Effect sensor

Problem Description

The Hall Effect sensor measures proximity of ferromagnetic object, they output a voltage dependent on how close an object is. The sensor does capture ferromagnetic object from each direction. What is happening is the electromagnet is so powerful it is saturating the sensor which does not allow the sensor to read the clamp. This problem did not present itself in earlier testing due to a smaller mass being levitated as well as a less powerful electromagnet.

The figure below shows the PWM signal going to the electromagnet, as well as the hall sensor reading. It shows that the sensor reads zero long before the electromagnet is at full power.

PWM magnitude compared to sensor reading

PWM magnitude compared to sensor reading

Causes

This problem essentially means that the control system has no feedback. Since this is a non-linear, inherently unstable system, the loss of feedback means there is no way of controlling the system. This problem also means there is no way of determining distance from the levitation clamp.
Control system with sensor feedback removed

Control system with sensor feedback removed

Solutions

A new Hall Effect sensor with a higher saturation point

The magnetic field of the electromagnet can be "isolated" via hardware through the usage of two Hall Effect sensors. Positioning the sensors at the top and bottom faces of the electromagnet allows for this isolation. The Hall Effect sensor on top only has the capability to detect the magnetic field of the electromagnet, as the levitating object is too great of a distance from to impact this sensor. The Hall Effect sensor on the bottom face of the electromagnet, however, reads the combined effort of both magnetic fields. Subtracting the output voltages of the two sensors gives the voltage proportional to the levitating objects distance from the bottom sensor, and that voltage can be transmitted through the systems negative feedback loop. Through this methodology, a stable closed loop system can be obtained, and thus the object can be levitated with oscillation invisible to the human eye.

An inferred proximity sensor

Inferred/optical proximity sensors

Inferred/optical proximity sensors

Optical proximity sensors

Optical proximity sensors

A capacitive sensor
Capacitive sensors

Capacitive sensors

An ultrasonic sensor
Ultrasonic sensor

Ultrasonic sensor

Project evolution, plus projected outcome

Project evolution, plus projected outcome

Shroud Heating Problem Tracking

Problem Statement

The inside of the shroud does not maintain a constant temperature.

Problem Description

As seen from the thermal shroud testing, the inside temperature fluctuates significantly during heating cycles. Since the coefficient of thermal expansion is defined as strain per degree F, not having an accurate measure of the sample temperature will be a significant source of error.

Causes

Possible causes include:

Solutions

Quantify error associated with the temperature fluctuation, make no modifications to the shroud

Attempt a different layup of the thermal tape

Purchase new controller and/or thermal tape for better performance

Error Analysis Update

Error analysis values were updated as necessary with the available data. The most significant update is to the nonuniform heating of the shroud. A value of .02 microinches per inch CTE was assumed, given a carbon fiber composite test coupon 8 inches in length. The roughly 4 degree F nonuniform heating results in the uncertainty in length shown. Note that this value is material dependent, and the greater the CTE, the greater the error.
Error analysis values

Error analysis values

Graph of error ranking

Graph of error ranking

Plans for next phase

Project plan

Project plan

Major Accomplishments

Major Obstacles

Projected State Week 14

Our projected state for the next review is to have the full system integrated, and a working maglev system that is capable of levitating the full clamp. Additionally, the error analysis will be updated based upon the experience gained from integration and testing of the full system.

If the team had an additional 4 weeks to prepare the system, we would like to have an incredibly stable maglev system that minimally impacts measurement error, as well as a shroud with very little temperature variance during testing. Additionally, it would be ideal to have CTE measurements for various materials from our test apparatus, so that they could be compared to third party results. This would help us to get a good understanding of how accurate our system is.

Mid-Phase Update

Note: This section is to provide an overview for the team's check-in with our faculty guide in between the Integrated System Build & Test and Customer Demo design reviews. The content is meant to address shortcomings identified at the Integrated System Build & Test design review that had to be addressed more promptly than at the next design review 4 weeks out.

Updated Team Vision

At the Integrated System Build & Test review, the team presented a major roadblock that had presented itself right as the full CTE test fixture was assembled: the Hall effect sensor, a key part of the feedback control system, was saturated by the larger electromagnet and no longer functioned. The decision was made to proceed with using the capacitive sensor to control the maglev by measuring from the bottom of the sample coupon. Unfortunately, this means that the fixture can no longer read CTE while the maglev is active, since there is no top sensor to provide a delta with respect to the bottom sensor. Since all other components were successfully implemented, with the exception of the cap sensor, the plan for this 2-week half phase was to fully miplement the capacitive sensor and use it to control the levitation of the sample coupon.

As an actual accomplishment, the capacitive sensor was fully integrated in the mounting hardware with the help of a 3D printed sleeve to provide a tight fit in the mounting flange. Additional circuitry was built to step down the cap sensor's 10V output signal to the 5V max that the Arduino controller can read, and the control scheme was updated to allow levitation based on the cap sensor's input signal. Aluminum test coupons were machined to

A significant problem arose when work began on tuning the controller to utilize the capacitive sensor signal: the coupon clamp is not lifted uniformly by the electromagnet. A slightly uneven center of gravity causes one edge of the clamp to pick up higher than the other, and the resulting angled condition creates instability in the levitation system.

Another issue arose as a consequence of the physics behind the capacitive sensing mechanism. The capacitive sensor needs an electrically conductive surface to read a signal, so a thin aluminum foil "target" has to be placed on the bottom end of the coupon in order to measure position on insulating materials, e.g. carbon fiber. This target suffers from internal eddy currents induced by the alternating PWM-driven output of the electromagnet. This results in significant noise in the capacitive sensor signal that is not present when the capacitive sensor is run without the magnet on.

Team members' individual 3-week plans and accomplishments can be found here.

Problem Tracking Update

Full system integration problems

Full system integration problems

Sensor Options Update

New Hall Effect sensor with higher saturation point
Pros Cons
  • It could allow feedback
  • Relatively inexpensive
  • Budget constraints
  • Difficult to tell if saturation point will be great enough
  • Still produces electromagnet bias

Infrared/optical proximity sensor

Pros Cons
  • Would allow proximity feedback
  • Would allow true position readings for CTE
  • Would not have electromagnet bias
  • Budget constraints
  • Difficult to find close range-high resolution infrared proximity sensors
  • Most likely would need a top plate design change
Infrared/optical proximity sensors

Infrared/optical proximity sensors

Capacitive sensor

Pros Cons
  • Would allow proximity feedback
  • Would reduce CTE error significantly
  • Budget constraints
  • Would require a top plate design change
Capacitive sensors

Capacitive sensors

Ultrasonic sensor

Pros Cons
  • Would allow proximity feedback
  • Budget constraints
  • Electromagnet may affect signal readings
  • Would require a top plate design change
  • Cannot be used in a vacuum
Ultrasonic sensor

Ultrasonic sensor


Home | Planning & Execution | Imagine RIT

Problem Definition | Systems Design | Preliminary Detailed Design | Detailed Design

Build & Test Prep | Subsystem Build & Test | Integrated System Build & Test | Integrated System Build & Test with Customer Demo | Customer Handoff & Final Project Documentation