Subsystem Build & Test
Team Vision for Subsystem Level Build & Test Phase
Manufacturing UpdateParts manufactured:
- Mounting Sheet Metal Bracket (Water jet and bend)
- Hopper Mounting Plate (Water jet and mill)
- Funnel 1/8" and 1/16" (CNC Lathe and mill)
Items to be Machined:
- Feed tube
- Plastic Skirt
- Final agitator spinner
Items to be Printed:
- Feeder body V3
- feeder Arm
- Motor Cowl
- Counting Sensor Bracket (for testing)
- Sensor Housing
- Electronics Housing
Purchasing and Budget Update
Items purchased this phase
- More compatible stepper motors for shields
- Additional Arduino Megas
- Additional light sensors
- DB9 connector
- Power Supply & cables
Items that may need to be purchased (pending testing)
- Different springs for feeder
- Stepper with gearbox
- Voltage regulators for agitator
- Cable disconnects / connectors
Video of electronics test: https://youtu.be/-YHNFS9sLuY
- Running off of a single 12V DC brick, will upgrade to two 12V supplies to power the light tower and the motors separately
- 3 state (red, yellow, green) light tested and working for operator notifications
RS 422 (Command Signal)
- Shield bypassed and chosen direct analog interfacing with a voltage divider instead
- Signal range tested and is acceptable for constant and accurate readings
Currently unable to communicate with level sensor via RS-422 digital output, meeting with Professor Barrios to see if this is still possible. Tested analog output from level sensor to measure changes in distance. A change in voltage was measured due to movement of the sensor, and the Arduino was successfully able to read this change.
- All 3 photogates are tested and opperational
Photogates were first tested individually on breadboards to confirm that the voltage change from shining the LED on them would be suitable for use with the Arduino. The photogates were then tested with the Arduino, with the Arduino reading the output voltage from the sensor. This test was successful.
- Both the agitator motor and servo motor are tested and work
- Tuning for agitator motor ongoing as range defined but electrical wiring improving to tune into it more
- Code for all subsystems is written and working on integration of all subcomponents together.
We decided not to use ANSYS modeling to find the natural frequency and instead used a variable power supply to allow us to change the agitator frequency and then empirically tested to find the frequency that worked. The optimal frequency corresponded to a motor input voltage of about ~7.9V.
Video of test: https://youtu.be/DtCgfAm52XE
A behavior we noticed is that, when the hopper occasionally gets stuck, "pulsing" the motor to a high frequency for about a second and then returning to optimal fixes this problem nearly immediately every time it was tried. Therefore we are going to incorporate new functionality in the design to do this. The plan is to have two relays going to the agitator, one at 8V and the other at 12V. The 8V will run until the sensor detects its stuck, then it will switch to the 12V for a moment and then switch back.
To test the feeder subassembly we ran the stepper motor at a constant speed and analyzed the travel path of the rod through the feed block.
Video of "No Shim" Test: https://youtu.be/0DFYmLd_-SM
Video of "1/8 Shim Test : https://youtu.be/ib5neuDmT1k
Video of "1/16 Shim Test: https://youtu.be/GvQxtGsNdJA
Note:these tests we recorded after we replaced the springs with weaker springs.
Through testing we found that the motor struggled to push the rod through the feed block. We observed that this was caused by several things,
- The holes which the rod traveled through were slightly misaligned due to the wheels being more "squishy" than expected
- The springs on the feeder door were too strong and created too much friction
- The wheels were not able to engage the rod and pull it down. This was caused because only the powered wheel moves, and when the rod drops down it only contacts the unpowered wheel
- The motor was not strong enough to overcome the friction caused by the previous reasons
- Use weaker springs
- Slightly move the hole location in the feeder block
- Move the center location of the wheels together so the wheels contact each other. This way both wheels will spin to engage the rod.
- We could also use shims to prob the feeder door open slightly. This would decrease the compression of the wheels on the rod.
- Possibly purchase a stronger stepper motor to overcome any friction. This may not be needed, however we have found a suitable replacement motor if it comes to that.
Our main concern was if the Mounting Subassembly would fit in the Vader System without interfering with the operation of the Vader. We tested by placing the Mounting Subassembly in the Vader and it fit sufficiently.
Angle Feed Decision
Based on the results from testing the feed block, we have decided that the angled-feed is not feasible and are moving forward without it.
- The stepper motor used (and any reasonable replacement) in the feedblock does not have sufficient torque to feed a rod through a bend of that angle and deform the rod.
- Such a feed tube would be very difficult to manufacture and integrate regardless
- Deforming the rods to feed them is frankly not a reasonable method, they should be allowed to feed straight into the crucible.
Impact on this project:
- The device will probably not be out-of-the-box compatible with the VADER
- We plan to complete and display the bench-top model
Test Results Summary
Based on the results from testing:
- Each subsystem is currently capable of meeting spec
- For each subsystem, testing located parameters that could be improved and we are working on incorporating them into the design for phase 3
- We will be delivering a bench-top model of the device, not full VADER integration
Imagine RIT and Tech paperWe have signed up for Imagine RIT and are in the process of filling out exhibit info.
Poster will begin work next phase.
The technical paper has been started, James will spearhead the writing.
Risk and Problem Tracking
- Potential problems that we have identified for the upcoming phase are the system not working when all subsystems are integrated, the feed block jamming during operation, and not being able to communicate with the level sensor via RS-422. For the full system, it is difficult to know a potential solution until we fully integrate all parts. For the feed block jamming, solutions include using weaker springs, moving the hole location in the feeder block, and possibly purchasing a stronger stepper motor. For digital communication with the level sensor, testing and meetings with Professor Barrios will continue to understand how to communicate between the Arduino and the level sensor.
No problems have presented themselves as of yet. When they do, we will use the process outlined in Phase 1.
Planning UpdateMSD 2 Overview:
Phase 3 Plan: