P15318: Gaseous Mass Flow Rate Controller
/public/

Subsystems Design

Table of Contents

Microcontroller Feasibility

Potential Options

Pugh Analysis

Detailed Evaluation

Conclusion

The Teensy 3.1 has the potential to greatly surpass the piccolo but it would definitely require a significant amount of extra work, assuming it is even possible. As of right now it seems the piccolo should meet our needs but due to the costs of the microcontrollers it is feasible to get both and find if the extra work required for the Teensy 3.1 is worthwhile.

Valve Design Feasibility

Potential Options

Ball and Cam

This design uses a cam with a calculated profile attached to the shaft on a rotary actuator. The cam will be in contact with a ball that is fixed to motion in one direction. The rotation of the cam will cause the linear motion of the ball. Geometry around the ball will be a conical seat, which allows for good sealing and a low change of sticking/wedging. There will be a spring that keeps the ball in constant contact with the cam but will be a variable obstruction in the flow. There is a good chance that the existing housing can be used. The flow vs. rotation angle would be a parabolic relationship, with some uncertainties due to the spring obstruction.

Ball and Cam (Clear Housing) Ball and Cam (Cutaway)

public/CAD/Ball and Cam/Ball and Cam (Clear Housing).PNG

public/CAD/Ball and Cam/Ball and Cam (Cutaway).PNG

3D PDF of Ball and Cam (Need to right click > save link as > open through Adobe Reader, not through web browser)

Ball and Cam with Separate Channel

This design is similar to the Ball and Cam design, except a separate channel for flow will be used. This separate channel would get rid of the spring's variable obstruction on the flow. The complexity of this design is higher due to the need for a ball contacting the cam as well as another ball regulating the flow. The two balls would be connected by a small shaft to ensure they move together but the shaft would be a constant obstruction in the flow. The shaft would also need to be supported in the channel, which would be an added complexity. The existing housing would not be able to be used. The flow vs. rotation angle would be a parabolic relationship, with a known restriction caused by the shaft.

Ball and Cam with Separate Channel (Clear Housing) Ball and Cam with Separate Channel (Cutaway)

public/CAD/Ball and Cam/Ball and Cam (SC Clear Housing).PNG

public/CAD/Ball and Cam/Ball and Cam (SC Cutaway).PNG

3D PDF of Ball and Cam with Separate Channel (Need to right click > save link as > open through Adobe Reader, not through web browser)

Rotating Disk

This design is based off of the existing prototype design. It uses a circular disk, with a calculated slot, attached to the shaft on a rotary actuator. The rotation of the disk will restrict the opening area of the output fitting. The complexity of this design is far less due to the fact that it has been proven in the prototype, the existing housing will most likely be able to be used and there will be less total parts. Only a few parts would require precision machining. The leakage rate of this design may be larger than the other two designs due to the precision of the geometry needed to ensure a tight seal. The flow vs. rotation angle would be a parabolic relationship with a point of inflection, with no restrictions/obstructions in the flow.

Rotating Disk (Clear Housing) Rotating Disk (Cutaway)

public/CAD/Rotating Disk/Rotating Disk (Clear Housing).PNG

public/CAD/Rotating Disk/Rotating Disk (Cross Section).PNG

3D PDF of Rotating Disk (Need to right click > save link as > open through Adobe Reader, not through web browser)

Pugh Analysis

Conclusion

The PUGH analysis, as well as our own intuition, has showed that the Rotating Disk will be the best design to pursue. It is a design that has been proven to work in the current prototype and will give us the most desirable results. Our design will be based off of the current prototype. Potential modifications that will be made include adapting the housing to accommodate pressure and temperature sensors, a larger rotating disk and a larger output port. A design consideration that will be taken into account is the ability of the housing to accommodate different size rotating disks and different size output ports. This will allow us to make and test multiple sizes/configurations to optimize our device for a relatively low cost compared to making a custom housing for each configuration. Another goal of our design is to include as few parts as possible that require precision machining. Any parts that will experience excessive wear will be easily replaceable as well.

Position Sensor Feasibility

We conducted an evaluation of our given position sensors to ensure that it was as accurate as previous test data claimed. Shown below is our test setup to evaluate the sensor. We used a large protractor to turn the sensor small angles to obtain the relationship between angular position and voltage.

Our results confirmed that the output of this position sensor is linear and the step size that we obtained from testing matched that of the one from the previous test data. We found that the average step size was about 0.047V/deg when supplying the potentiometer with 5V. The previous data claimed that the step size was 0.925%/deg, which comes out to be 0.04625V/deg when supplying the potentiometer with 5V.

public/Detailed Design Documents/Miscellaneous/Position Sensor Testing Setup.jpg

Position Sensor Testing Data

Rotary Actuator Feasibility

We conducted a test on the rotary actuators given to us by our customer to determine their angular operating range. The actuators were hooked up to a DC power supply and various amount of voltage and current were applied. The response was observed and examples of the range of angular motion is shown below. Our testing concluded that the average maximum angular rotation when testing at a maximum of 8V was about 65° and when testing at a maximum of 13V was about 70°.

From this testing, we concluded that were are going to base our design off of the rotary actuator having a maximum angular rotation of 55°. One reason for this is because we want our device to work with any input voltage above 8V. Therefore, we wouldn't want to design it with a maximum angular rotation of 70° because we would never get that much travel with 8V. To make sure that any given input voltage above 8V will work with our device, voltages above 8V will be handled using PWM to ensure that we get consistent rotation results.

Testing at 8V

Clean Shaft Actuator Position 1 Clean Shaft Actuator Position 2 Rusted Shaft Actuator Position 1 Rusted Shaft Actuator Position 2

public/Photo Gallery/Actuator Max Rotation Test Pictures/Clean Shaft Actuator 8V Max Rotation Test-1.jpg

public/Photo Gallery/Actuator Max Rotation Test Pictures/Clean Shaft Actuator 8V Max Rotation Test-2.jpg

public/Photo Gallery/Actuator Max Rotation Test Pictures/Rusted Shaft Actuator 8V Max Rotation Test-1.jpg

public/Photo Gallery/Actuator Max Rotation Test Pictures/Rusted Shaft Actuator 8V Max Rotation Test-2.jpg

Testing at 13V

Clean Shaft Actuator Position 1 Clean Shaft Actuator Position 2 Rusted Shaft Actuator Position 1 Rusted Shaft Actuator Position 2

public/Photo Gallery/Actuator Max Rotation Test Pictures/Clean Shaft Actuator 13V Max Rotation Test-1.jpg

public/Photo Gallery/Actuator Max Rotation Test Pictures/Clean Shaft Actuator 13V Max Rotation Test-2.jpg

public/Photo Gallery/Actuator Max Rotation Test Pictures/Rusted Shaft Actuator 13V Max Rotation Test-1.jpg

public/Photo Gallery/Actuator Max Rotation Test Pictures/Rusted Shaft Actuator 13V Max Rotation Test-2.jpg

Drawings, Schematics, Flow Charts, Simulations

Microcontroller

Distribution Plate

The end state of the distribution plate is that it will be mounted underneath an engine's throttle body. Its function is to efficiently mix fuel with the air passing through the throttle body. To accomplish this, fuel will enter the distribution plate through a port and exit through output nozzles. The size, direction and spacing of the nozzles will be critical to ensure even distribution and vortex characteristics. SolidWork's built in CFD simulation toolbox was used to evaluate our distribution plate designs.

Design 1

This design was our initial proof of concept, which shows that a vortex can be produced with the correct nozzle orientation. However, this design was not manufacturable and not able to be 3D printed, as we had initially planned.

Design 2

This design was made with manufacturability in mind, and made with standard tool sizes. It was also composed of two peices; a bottom half which contained the nozzles, input port and O-ring groove and a top half that enclosed the flow path. Each half contained corresponding holes to properly locate the two peices. The size of the nozzles and input port of this design were determined by making sure that their area restrictions are larger than the restriction at the output port of the GMFRC device.

Design 3

This is the current design. Design 2 used equal sized nozzles, and this meant that going from the first nozzle to the last, there was a drop off in flow. This made sense, as the further down the line you go, there is less total flow in the channel. To address this, the size of the nozzles were varied. As you go from the first nozzle to the last, the nozzle sizes get larger to have a more even distribution of flow.

Design 1 Design 2 Design 3

public/CAD/Distribution Plate/Vortex Simulation.png

public/CAD/Distribution Plate/Vortex (B Channel, Equal Area Nozzles).PNG

public/CAD/Distribution Plate/Vortex (B Channel, Nozzle 5 Percent Increase).PNG


Home | Planning & Execution | Problem Definition | Systems Design | Subsystems Design | Detailed Design | Build & Test | Photo Gallery