Build & Test Prep
Table of Contents
Objective StatementManufacturing system to produce reed valves for accordion restorations by Scott Bellinger.
Team Vision for Build & Test Prep PhaseWhat did your team plan to do during this phase?
- Planned on completing the CAD modeling
- Verify electrical components function as expected
- Begin drafting/writing the ladder logic code
What did your team actually accomplish during this phase?
- Clarified team roles and expectations with our customer and new team member
- Nearly complete in the CAD models
- Test plans
- DOE for plastic testing
- Looked into HALT testing
- Proved that the life span of plastic isn’t a risk, began peel testing welded region.
- Work instructions for Punch and Cut
- Started air diagram along with pressure and size calculations for valves and lines
- PLC and all input/output cards tested with everything functioning correctly. The correct software has been installed and now we only need software for HMI programming. This will be a 30 day free trial closer to imagine. Also tested hall effect sensors and relays for 120 VAC outputs, everything works correctly so far. We have constructed some ladder logic for Feeder station 1 it can be seen below.
Lessons Learned in MSD I
- This is a complicated project and we have to keep the project in scope
- There is always stuff to be done
- Time is the biggest risk factor in the project
- Keep the process simple, complexity adds too many layers and too much time to the design
- It best to verify that the software tools are activated/working as early as possible
The dial layout was changed only slightly, adding another 6.35[mm] station to the dial instead of just leaving it blank. For this change, 4 more custom nests will need to be fabricated and the code will need to be tweaked slightly. However, this will also double the 6.35[mm] throughput.
S01- System Interface : Pre-approval revision S02- Reed Plate: Approved S03- Indexing Dial: In final revision/approval process
- Critical Parameters: Speed, System height line
- Primary Mechanism: indexing gear box, dial face, jigs
- Primary Failure Modes: Too slow, system crashes
- Mitigation of Failure: Paralyzation of processing, dictates system work plane height
S04- Material Feeding:In final revision/approval process
- Critical Parameter: Normal Force
- Primary Mechanism: Spring loaded pinch roller
- Primary Failure Mode: Slipping of drive, insufficient normal force
- Mitigation of Failure: Spring loading is adjustable via a thumb screw
- Alternative: Pneumatic slide driving pinch roller
S05-Fusing: In final revision/approval process
- Critical Parameters: Time, Temperature, Pressure
- Primary Mechanism: Soldering Iron
- Primary Failure Modes: Burn through, crinkling, no weld
- Mitigation of Failure: DEO to determine appropriate spec
- Alternative: Using a knurled or pin covered surface
S06- Punch and cut: Approved
- Critical Parameter: Shear force
- Primary Mechanism: Pneumatic slide with cut/punch tooling
- Primary Failure Modes: No cut, No punch
- Mitigation of Failure: scissor action of cutting blade, stress concentrating punch, jack block for die alignment
- Alternative: Applying Tension when cutting
S07- Test and Collect: Removed from Scope
- Will now be done by Scott and be on the dial instead of off
- This was accommodated by adding slots into the design of the dial and jigs
S08- Pick and Place: Approved
- Critical Parameter: Vacuum Pressure
- Primary Mechanism: Vacuum pick up
- Primary Failure Modes: Not picking up a valve, Dropping Early, Crashing
- Mitigation of Failure: Determined vacuum pressure required to pick up valve, built on same supports as feeder, constant pressure monitoring, buffer designed into vacuum cup
- CAD modeling remotely over break was a hassle and quite complicated.
- Splitting up the subsystems proved to be hard at times when conversing about dimensions this specific.
- Making sure we are all using the same version of software.
- Make sure you model the design in a way that it is easily editable.
The bill of material has not changed too much since the last review. This is mainly because our subsystems are just being approved now and purchasing before approval was too risky with the budget projections we currently have. Therefore, later this week the CAD package BOM will be compared to the one shown below and many more parts will be ordered. We currently have 20% of the BOM below ordered. However, please see the Detailed Design page for the full BOM of donated parts we have already acquired.
Lessons Learned: Through this process the team learned that cost adds up fast and without the major donations from Calvary and PMD, as well as Scott buying several components for us, making it his far with this design would not have been possible. Secondly, we learned working through RIT's PICS system slows down the procurement procedure a bit, therefore we needed to allow for more time in our schedule.
- PLC ladder logic design is driven by the timing diagram. All Input and Output ports are already decided, so it is a matter of telling the PLC where to read/write and when to do so
- Heat Stake, Indexing code, and Feeder code are already drafted
- Feeder 1 coded (~20% of the system), but not tested on hardware
- HMI not yet started, but begun drafting UI concepts. Cannot be started until system is built - 30 day free trial
- Scott has written code to interface with the stepper motors, now we need to determine what voltages determine what lengths and configure it to send a DONE flag
- All of the I/O port on the PLC and the I/O cards have been tested and verified as working.
Below is the ladder logic code for the Feeder1 station
Try to keep the design simple as more I/Os complicate the project and also increase cost as more sensors, motors, and other forms of output devices are needed. Also verify correct software is installed and start testing devices sooner.
Below is a sample of the Pick & Place Work Instructions:
Test Plan Summary
Product Test Plans
Based of of the Engineering Requirements outlined at the beginning of the project, we developed Master Test Plan to declare whether or not our final product meets these guidelines. These refer to the safety and usability of the manufacturing system itself, as well as the functionality of the produced reed valves. This set of test plans will only be used to test the final outcomes from our project.
Subsystem Test Plans
To have continuous testing for the manufacture and build phase, the System Test Plan were written up. These clearly identify how each subsystem should work independently from one another. This way, as each part is manufactured and verified to be correct, an individual can assemble one subsystem, connect it to the HMI and PLC to ensure that it functions properly. This breaks down the build and test phase into manageable subsections so we can troubleshoot specific problems as they arise, rather than putting everything together in the entire system and testing from there.
To breakdown the subsystem testing even further the Parts Verification sheet was created to make sure that every ordered and manufactured part is exactly correct. When each part is finished from the machine shop, or comes in from being ordered, this sheet links to the CAD drawings for the required specifications. This is necessary for the upcoming phase because each subsystem is built from the parts in these CAD drawings, so it is imperative that we verify the correctness at the beginning to minimize possible errors that could occur
Finishing these test plans was crucial for this phase because it gives us outlines, dates, and what to expect for next phase. It also broke testing as a whole down to more manageable tasks that we can check as we go, rather than testing everything all at once at the end. This will maximize the time we have to build by minimizing errors that we will inevitably have to deal with up front .
Not Welded: If temperature or duration is not enough to raise temperature of all layers into the Tg range, then no bonding occurs. Layers may stick together but will not withstand a 0.25 lb force.
Warped: If heating element is too hot, the length of the layer is heated close to the Tg value, becomes rubbery, and warps. Then deflection and sealing will be impacted.
Melted: If temperature or duration is too long the layers surpass the Tg range and begin to deteriorate.
The graph below shows the results of the plastics testing conducted to define the best fusion conditions. It can be seen that as the temperature increases, the “Y” or quality goes down. As time increases, quality also goes up. Pressure does not appear to make a huge impact. This means you will try to minimize heat and increase exposure time. This output is using JMP and I used an ordinal logistic regression to fit a model to the 3x3 full factorial DOE. Ordinal logistic just means that the response is categorical and the order matters, hence our 1-5 rating scale for the quality. Through this testing it was discovered that a temperature of approximately 150 degrees Celsius for 5 seconds is most desirable.
The table below shows temperature is the most significant factor, then temperature and time. At this time it does not make sense to do proof of concept outside of the design because temp and time can be accurately adjusted in the proper environment when the machine is built.
Life Span Testing
Using the real world data shown in the graph below, the life span of the amorphous polyester being used in the valves could be calculated. Through this the plastic itself will have an "infinite life" because the glue or weld will fail long before the material.
This allowed us to drop off our accelerated life testing, and instead focus on the weld testing. For this weights were hung from one layer while the others were held fixed. This was repeated several times with both the handmade valves and existing Italian valves. The results can be seen below.
As can be seen the peak of the fitted distribution for the hand made valves is a higher value, and the standard deviation is less. Therefore, from this sample group the hand made valves are stronger and more uniform. These tests will need to be repeated again once the fusion station is set up to ensure quality. Along with this a larger sample of Italian valves will need to be assessed.
Lessons Learned: During the plastics testing done, we learned that it is never too early to talk with experts. It wasn't until I talked with Dr. DeAngelis that I understood exactly what properties the plastic required. This is the same for our fusion method and temperature requirements.
Use your test plan to summarize test results and assess effectiveness of test plans to unambiguously demonstrate satisfaction of the engineering requirements
Below is our constructed Failure Mode and Effects Analysis (FMEA) for our project. The FMEA has been updated too show new risks that have come into existence as we move forward in the design process. Risks with higher importance have corrective actions in place to minimize the severity and possible remove the risk completely. The risks identified are tied to certain engineering and this is expressed in the last column of the FMEA. These risks could affect whether the project is completed on time or the overall quality of the reed valves.
Project Plan and Budget
Further breakdown of key deliverable are listed in the project document
Things to Note
- We can see from the red that we are being in terms of deliverables due to the amount of work that was to be completed over break (mostly the CAD work)
- We worked over break but still have a lot to work up toward to complete for next phase
- We are ahead on out test plans which allows us to test things as they come in the door while other teams are still in the beginning phases of them
- Have planned individual meetings for weekly reports and to see where people need help
Plans for next phase
- Continued confirmation of testing parts
- We have dates and people assigned to tasks weekly to complete for the next 4 weeks
- We will be focusing one subsystem at a time, then ordering parts accordingly then working through the rest of them
- Putting plans in place to further mitigate a risk of not completing things on time