Team Vision for System-Level Design PhaseBy the end of this phase we would like to have finished research and benchmarking, understand why the bottle braiding team’s machine didn’t work, refine engineering requirements (OSHA, quantitative), brainstormed a variety of feasible designs (electrical, mechanical, and control scheme), and selected a design to move forward to detailed design with.
During this phase, we have selected concepts for our sub-systems, we understand the drawbacks of the bottle braiding machine, and are ready to move on to detailed design.
This Functional Decomposition Tree helps uncover function details in a structured format and ensure that all engineering requirements are being addressed and satisfied by our system.
Note: Due to the nature of our project as an exploration into relatively unknown technology, it is challenging to apply standard metric benchmarking. While a few industry braiding machines exist, they are vastly more expensive and thus have near incomparable metrics, compared to our expected design.
Feasibility: Prototyping, Analysis, Simulation
Literature review did not reveal common tension values for 3D braiding, so we set up a prototype to determine if the motor we had could handle what we deemed as qualitatively appropriate amount of tension in string. The cardboard represents the horn gear in our design. Based on this testing, we determined that motors with higher torque ratings were needed.
Morphological Chart and Concept Selection
- Independent Sync Signal fixes many synchronization issues
- Daisy Chain is cheap and easy to wire
- Two way communication
- Scaling isn’t the best, but for a prototype it is cheap and easy.
- A network may be more appropriate for future iterations.
- Assumes that bandwidth is minimal
- Assumes all transmission techniques will have errors, and will need two-way communication to request a resend
- Assumes Bandwidth is minimal
- 9 Nodes could be controlled with one microcontroller, but to test expandability, a scalable approach will be used
- UART uses two bare wires, so it is cheap
- Available on everything for no extra cost
- Easy to program
- Well tested and proven throughout history
- K.I.S.S methodology is good for a prototype
- I2C GPIO expander is very cheap and can provide more IO to future uses
Zigbee\Xbee is recommended for future, larger scale
- Servos are overkill
- Steppers have plenty of control
- Pneumatics may scale better and have more power, but the complexity and effort *is too great for a prototype. Packaging is also harder
- Steppers have standardized sizing
- Driver boards can take 4 coil wires -> 2 wires (Step + Dir).
- Assume engineering effort is mostly negligible
- Assume computational power is minimal
- Nearly a GPIO per dollar comparison
- Needed enough UART for daisy chaining
- Teensy 3.6 has more future expandability if Zigbee was used
- Teensy 3.6 has more I2C channels if GPIO extenders are used
Systems ArchitectureThis Transformation Diagram helps identify and illustrate interactions between our system and the environment as well as internal interactions. In particular, it has been helpful to see the interactions between our mechanical, electrical, and control systems.
Designs and Flowcharts
Click the following link for the Risk Management active working document
Preliminary Test Plan Overview
- Run machine for 2 hours with thermocouple(s) placed
in electronics compartment
- Make sure electrical components do not overheat
- Goal: less than 158 ºF
- Weigh Machine
- Goal: less than 100 lbf
- Measure power draw during operation
- Goal: less than 1000 W
- Time study of bobbin swaps with 10 repetitions. Find
average and best repeatable time.
- Goal: less than 2 seconds per index
- Tensile Testing of Braided Rope - Maypole Style
- Compare values to online resources for expected rope strength
- Goal: ultimate tensile strength within 20% of theoretical calculated value
- Create part with Y-shaped Mandrel, resin added and
- Braid component around Y shaped mandrel such that every area of mandrel has appropriate coverage
- Goal: braided part is knit tight w/ no gaps, 100% coverage
Engineering Requirement and Constraint Updates
We added tension force estimates to our engineering requirements.
Multiple items originally on our engineering requirement list were actually constraints, so we updated both lists accordingly.
We added a constraint of fiber not loosening during braiding. This directly relates to ER20 (fiber tension)
Design Review Materials
Action Items from Review:
- Present slower! (give him time to absorb info as well as write down feedback. Maybe pause at the end of reading a slide?)
- Tension prototype: carbon fiber doesn’t allow tight radii (brittle)
Torque prototype: 3D print horngears
and account for representative friction
- If we end up needing more torque than hobby-grade, he may be willing to increase budget
- We can use the Mark4 printers to make CF/Nylon
composites and decrease weight of horngears
- Process: notify him in advance and he will make it happen, either AMPrint or 3D print classroom
- Scalability: there should be no redesign necessary to move up to bigger matrix. Scalable control scheme is critical, as well as aspects like the pegboard (it doesn’t make sense to have a giant chunk of metal raising and lowering with the progression of the part). Scaling it should be a small summer project for someone, not part of another Sr. Design
- Stall detection? Will this be possible with our controllers?
- Suggests waterjet or plasma cutter for large tracks / horngear holders
- Meet next Tues or Thurs with Dr. C and Andrew Greely (PhD, knows generally where Dr.C’s stuff is). Hopefully schedule enough time to go to the places he stores things. Rob Kranik from Brinkman also knows things
- Future integration: Q: will the printing act as the matrix of the composite, or the mandrel? A: potentially both. He can envision the fiber being pre-impregnated with resin or even coated with a custom formulation of plastic and heated after it’s braided but still in the braiding machine, such that the braid angle is set and fixed so it doesn’t get messed up when it’s removed from the machine.
Cost: Dr. C questions why we would
need such expensive motors. Is our torque estimate too
high, or are we looking in the wrong place for motors,
or are we missing companies that sell smaller spools of
fiber? If we can justify the need for expensive motors
to handle 5lb. spools, he could potentially increase
- Make a 3D printed prototype to validate inertias and friction
- What is the goal of our machine? Is it supposed to be a proof of concept, or a usable prototype? A: both! It should be able to produce parts that are of decent quality, however since it is only a 3x3 matrix, its possible parts are limited.
Plans for next phase
- For phase 3, we would like to progress on our preliminary designs and start finalizing some of our ideas. Parts will be modeled in CAD software, BOM will be updated, a preliminary ordering form will be made, and all of our designs will be improved. By the end of phase 3, our tensioning and bobbin passing subsystems will be prototyped and tested so we can move forward. By testing these systems, we will know if they are good methods for our final product.
- Design software architecture
- Create software flowcharts
- Select specific steppers and solenoid
- Price wiring and connectors
- Improve tensioning prototype
- Continue CAD development
- Quantify system loads/torques
- Tensioning prototype
- Modeling/printing prototypes
- Incorporate tension of string to torque
- Finalize choices for electrical components
- Talk to SME about motor wiring, back EMF protection, and shielded cabling
- Update flowchart with chosen subsystems
- Create CAD models for parts, as designed
- Detailed design of mechanical subsystems
- Prototype and testing of bobbin passing
- Assist with risk and scheduling
- Document our decisions and add to edge
- Keep team mindful of good practices
- Continued updates to project schedule and Gantt chart
- Coordination of design and prototyping tasks between team members
- Aid in documentation and meeting minutes