Hi, I'm Blaine Tubungbanua

Angled Idler Roller Assembly: Test Jig

I designed this test jig to verify that the angled idler would convert rotation to axial translation

• 3D printed pillow blocks secure the bearings for the rollers, and a 3D printed adjustable roller assembly allows the angle of the idler to be calibrated.
• Parallel powered rollers cause the bar to spin, and the angled idler roller applies pressure from above, exerting an axial force to allow translation.
• The bottom rollers were designed to have less grip, to promote slipping along the parallel rollers while the top roller has more grip, to facilitate translation.
• In this test jig, the parallel rollers were powered by a cordless drill, and a handheld belt sander simulated the forces from the sander

This testing jig allowed us to verify the concept, and estimate how much downward force was required from the top roller to facilitate axial sliding without seizing.

Angled Idler Roler Assembly: Final Design

This was my final design for the angled idler assembly.

• The whole assembly can slide up and down along the two shoulder bolts, which locate the springs, supplying the required downwards force on the bar.
• Smooth sliding is ensured by the two brass bushings in the top square bar.
• To simplify machining requirements A shoulder bolt is used as the roller shaft, locating two bearings with 3D printed spacers.

Angled Idler Assembly: Spring Calculations

To ensure that the angled idler supplies sufficient downwards force to facilitate axial translation, I conducted spring stiffness and pre-tension calculations

• Empirical testing showed that ~15lbs of force was required to facilitate translation.
• Spring pre-tension, and buttom-out position were calculated, and a spring stiffness and length were selected to supply 15lbs-20lbs for rods over a range of motion that could accept varying bar diameters, from ½” and 7/8”, as requested by the client.

PCB Shield

To ensure robust electrical connections between the Arduino and the peripherals in a high-vibration environment, I designed a custom PCB shield.

• This shield implemented screw terminals, labels, pull-up resistors, reverse polarity protection, and a temperature sensor to detect overheating.
• This shield also improved ease of use and cable management for signal-ground pairings, allowing signal pairs to run next to each other.
• The first iteration was designed on protoboard, shown above.
• The final iteration was designed in KiCad.

Electrical Cabinet

As electrical lead for the project, I was responsible for:

• Power calculations to ensure the system could run on a 15A household circuit.
• Selecting a 120VAC induction motor to actuate the sanding belts
• Selecting a 48VDC brushless motor to actuate the rollers
• Power management for 120VAC, 48VDC, 10VDC and 5VDC systems for AC motor, brushless motor, arduino power, and arduino components.
• Programming of microcontroller in Arduino
• Selection of sensors: Limit switches, buttons, E-stop circuit

Structural Analyses

To validate the safety and reliability of the design, I conducted structural analyses.

• I conducted loading analysis on the arms, determining the forces transmitted to the pulleys from the sanding belt.
• With the loading determined, I conducted stress analysis on the arms and shafts, with a fatigue analysis for the shafts.

Instrumentation components in the electrical box. The system uses a LabJack U6 DAQ, connected to 2 multiplexor boards. The cables are then grouped at DIN terminals and connect to 4-strand cable which are soldered to the connectors interfacing with the outside of the cabinet.

Calibration of our thrust load cell on MTI tensile tester.

Successful hotfire of Mule-1, the team's test engine, verifying functionality of the test platform.

IO wired to the Arduino Mega. Push buttons control start/stop/pause functionality, with custom debounce code written. LCD screen displays number of each material counted. 47kOhm resistors are also wired into sensors, acting as low-pass filters to eliminate false triggers caused by electrical noise generated by the brushed DC motor. Clean cable management was one of the deliverables.

The system was designed as a finite state machine, shown in this diagram.

• As the system categorizes incoming cylinderes, they are stored in a linked list
• A software debounce was implemented for the push-button interface
• Interrupt Service Routines (ISRs) were programmed to detect sensor triggers, and button presses.

A reflectometer sensor was used to distinguish each distinct material. To calibrate the system, the amount of light reflected for each material was determined by measuring the voltage off the reflectance sensor. After taking a large sample size, the different materials were categorized.

• A - Aluminum
• S - Steel
• W - White Plastic
• B - Black Plastic

2D & 3D Meshing

To accelerate simulation time, 2D and 3D models were used. 2D simulations were used to quickly alter geometry and simulate results while 3D models were used to verify the 2D results and test the final configuration.

Optimization of the orifice diameter, length, and angle were investigated. To accelerate simulation time, a simplfied 2D single orifice model was used, and and open-loop parameter optimization was conducted using the parameter set tools in ANSYS Workbench.

A swirl injector was also investigated, proving to eliminate the air eddy seen in the non-swirled configuration.

3D model of reduction shaft. Labeled features are A: Retaining ring groove, B: Bearing seat, C: Bearing abutment shoulder, D: key way, E: Gear shoulder, F: Gear shoulder, G: Key way, H: Gear shoulder, I: Key way, J: Bearing abutment shoulder, K: Bearing seat, L:Lock nut threads.

Free body diagram of the reduction shaft, considering tangential and radial components of gear meshing forces.

Closeup of the meshing mechanism and the belt drive, allowing variable gear ratio.