Subsystem Build & Test
Team Vision for Subsystem Level Build & Test Phase (Kyle)
The Subsystem Build and Test phase is used to prove individual subsystems perform as expected, prior to full prototype assembly. The teams goal for this phase was to have the majority of the X-01 prototype assembled to be used in concurrently for testing while the A-01 final deliverable is assembled. This phase review will present the teams progress with subsystem testing, and ensure the team has captured all needed testing prior to final assembly.
The ongoing assembly and subsystem verification has been documented along the way. Pictures detailing the mechanical subsystem assembly, along with the electrical subsystem assembly and the software subsystems, will be shown.
The X-01 prototype is nearly completely assembled, with all subsystems being verified, and a video detailing high level operation of the prototype can be seen here. Along with the subsystem verification, risks have been updated. One PTR was generated and mitigated this phase regarding the force sensitive resistor. Test Results will be reviewed proving some of the subsystem level design elements.
Requirement TCR 15.1/TER 15.1 will be reviewed to accommodate latest customer input regarding the physical boundary limit of ALTAS.
StatusThe following details the current status of subsystem level testing:
Mechanical Systems (Kyle)
|Apparatus (FSR, Linear Act, Magnets, Microswitch)||100%|
|ATLAS Frame & Electronics Mount||100%|
Electrical Systems (Mark)
|Wiring (Arm, Motors, FSR, Linear Act)||100%|
|Teensy & Motor Drivers||100%|
|Power Supply Functionality||100%|
Software Systems (Sarah)
|High Level Header Files||100%|
|Draft of Motor Movement||100%|
|Linear Actuator Functions||100%|
|System Function Definition||100%|
X-01 Assembly Status (Harvick)
|Mechanical (Arm, Frame, Apparatus Assembly)||100%|
|Electrical (Wiring, Electronic Boards Soldering)||100%|
|Software (High Level Files, Draft of software functions)||100%|
Function Demos and System Pictures
Mechanical Assembly (Matt)
The X-01 prototype frame and NEMA 23 motor mounts were cut using RIT's waterjet. The frame has been mounted to the provided MFD for testing.
All inserts have been machined, and epoxied in place using the Jig. The carbon fiber arms have also been cut to length.
The arms have been assembled using the jig.
Manufacturing Updates (Matt)
ARM-01 UpdatesAn update to the Carbon Fiber Work Instruction, ARM-01, occurred during this phase to make the instruction more inclusive of information.
Current Manufacturing StatusThe current manufacturing status of the mechanical arm subsystem can be found here.
System Weight Tracking (Harvick)
|ATLAS X-01||ATLAS A-01|
Electrical Assembly (Mark)
The power supply has been wired, and is fully functional. A voltage regulator was used to lower the voltage to 6V, making the power supply capable of powering both the motors (24V) and the auxiliary components (Linear Actuator).
The electronics board for the X-01 prototype is a bread board, that will be mounted to the system using the electronics mount shown above. The electronics mount proved its capability of securing the boards, but the first revision was proven to be fragile.
To correct this, the mount will be printed using more infill, making the mount more solid, and will also be cut out of aluminum with a waterjet in the event the plastic is still fragile with 100% infill.
- Subsystem code created in C++ (view here)
- Stepper motors (running simultaneously)
- Touch apparatus (linear actuator and FSR)
- Interrupt test
- Integrated code structure created (view here)
- Source and header files for all functions
- As of right now, plan for config file is to use a header file to store all global constants
- List of all subsystems and basic structure of classes, below:
- Demos for Stepper Motors and Linear Actuator seen in videos above
- Demo of Kinematics (forward and inverse) for control system below
Assembly and Functions (Kyle)
The video above shows a majority of the subsystems operating together on the X-01 prototype. The power supply, electrical, mechanical, and software systems operated together independent of external influences. The microcontroller and motor drivers sent inputs to the motors and linear actuator to move a set amount in degrees.
The forward and inverse kinematics inputs are in progress, and will be functional during the next phase. In addition, the micro-switches will be integrated by the next phase. Their functionality has been tested, and they are functional.
Button Press Speed (Harvick)
The linear actuator's speed at pressing a single button repeatedly is about 52 presses per minute (1.15s each press), which below the 90 presses per minute requirement. Additional optimization will be implemented to attempt to increase speed of the actuator. A change of design from using a linear actuator to using a motor driven lead screw is being considered.
|Linear Actuator Button Press Speed|
Proposed Requirements Update
Based on email conversations with customer, additional details regarding to the physical boundary restriction on ATLAS was added: The critical physical limit that ATLAS cannot be exceeded is approximately 6" below bottom edge of the MFD, with no additional restrictions placed on top, left, and right side. The original border limit was 3" on each side of the MFD.
|Proposed Requirements Update|
TCR 15.1 Change.PNG
Risk and Problem Tracking (Mike)
PTR Report (Mike)Two problems were identified during the phase regarding the force sensitive resistor. A PTR was taken out to identify and track the issue. It was found that the FSR was very sensitive to changes in the loading conditions.
The other PTR that was taken out was with the electronics mount. As mentioned earlier, the electronics mount is relatively fragile with a 20% infill, with a 3 layer wall thickness. The part will be printed again with a higher infill percentage and a larger wall thickness to increase strength. The part will also be cut out of aluminum in the mean time to ensure a working mount for the A-01 prototype.
The Problem Tracking Report will be used to track additional problems as they arise.
Cost and Budget (Mark)
Overall Budget BreakdownThe full budget breakdown shows that the team is still within the total budget, but the original X-01 prototype budget went over the initial designation. This was to be expected as final parts and prototyping was chosen.
Another large change is the arms came in more than initial expected because of some custom work done through dragon plate to help with future orders, and the help the team apply man hours to other parts of the project.
Test Results Summary
FSR Testing (Mike)The FSR was tested using a range of weight values, and a few different load resistors that change slightly change the force curve for the output voltage.
With the results of both of these changes, the Current Prototype Testing Plan and Findings can be found at this link.
The raw data for the tests can be found here.
Apparatus Magnet Test (Kyle)The apparatus has the ability to be detached via magnets if the force exerted exceeds the magnets rating.
These three magnets, each rated with a 3lb pull, make the apparatus pop at a weight on average around 3.6 lbs. The pop occurs when one magnet disconnects causing the apparatus head to shift to one side.
Two solutions are going to be tested, the number of magnets with the same 3lb pull rating will be increases, and the pull rating of the magnets will be increased using the same 3 locations.
MSD II Phase 2 Closeout
|ATLAS MSD II Phase 2|
Plans for next phase
The first 2 weeks of next phase will begin the testing of X-01 models and production of the A-01 model, which designed to work with the Avalex AVM4227Q MFD. The third week is planned for system assembly and software integration, followed by a week of spring break. The fifth week is planned for system integration test and engineering requirement validations.
|ATLAS MSD II Phase 3 (Sorted by Sequence)|
|ATLAS MSD II Phase 3 (Sorted by Member)|
Lessons Learned (Harvick)
- Consider wider exterior factors with testing. In this case, the testing MFD is too unstable to test without being supported by a team member. A method for securing the MFD (such as a support frame) must be built. At this moment, the team is looking into using 80/20 as a support frame. In the mean time, two members are needed at all times during system testing.
Team InteractionSarah held an informative crash course in the approach she is taking with the software of the system. The course lasted about an hour, and provided the basics behind of the teensy and code operate with inputs and outputs of the system.
|Team Member||Peer Review Action||Feedback|
|Harvick Tang||Wasn't able to fill out team evaluations because of not being added to the team roster because of enrollment issues.||Has now been added back to the team roster for MSDII, and will be able to do peer evaluations next phase.|
|Sarah Bentzley||Keep everyone in the loop with software scheduling, especially Harvick.||Delegated more work out as needed to help keep the schedule.|
|Matt Craven||Discuss the manufacturing time needed for cutting and machine the arm and its inserts.||Kept the team in the loop on the committed hours regarding|
|Mark Min||Be more forward with how things are going with electrical systems.||Took an initiative with assembling the electrical wiring, and overall engaged more with the team.|
|Mike Kelly||Keep track of all testing supplies.||Checked out supplies for longer periods of time.|
|Kyle McAlinn||Great job organizing the edge presentation, and communicating with the team.||Keep everyone more involved with potential problems that arise as the X-01 was built.|