Preliminary Detailed Design
Manufacturing system to produce reed valves for accordion restorations by Scott Bellinger.
Team Vision for Preliminary Detailed Design PhaseWhat did your team plan to do during this phase?
The vision for this phase was to develop a preliminary, detailed design phase for our manufacturing process. In this we were to determine key outputs, such as the footprint of the table, first draft of a BOM, electrical schematics, and identify any key inputs that would be critical to the final design. During this phase, we also took a deeper look into our Engineering Requirements to make sure that all were being met and had a specific value associated with them that were tangible.
What did your team actually accomplish during this phase?
We continued to research materials for the reed valves and potential suppliers. There are various operations that these reed valves will undergo which required extensive testing and prototyping. This included replacing the current valves on the reed blocks and testing to see where our predicted outcome compared to our actual values. During this phase we also met with key stakeholders such as potential suppliers, and other Engineers with years of experience and knowledge who gave us feedback on our potential design. This allowed for us to see any flaws and get feedback on our system.
House of QualityDuring this phase there were quite a few changes with the customer requirements, which in turn caused changes to the engineering requirements. Due to the large changes a new house of quality (HOQ) needed to be reconstructed to ensure all requirements were being met and nothing was being missed or time wouldn't be wasted. Below shows the second revision of our house of quality. The new customer requirements are on the left wall and engineering requirements on the ceiling. The center shows that all of the engineering requirements meet up with a customer and vise-versa to ensure nothing is missed. The attic shows the positive and negative correlations between the engineering requirements, there are much fewer than last time because there is less overlap between the ERs. Finally each ER is given an overall difficulty rating in the basement, then the difficulty rating combined with the importance denoted in the center shows which ERs are most important.
Efficiency:Since the house of quality was being reworked this took much less time than the first revision. However, there were still a few hangups that did lower the efficiency. Due to this the efficiency of this was approximately 20%. This is fairly high because I was already experienced with filling this out.
Impact:The HOQ is highly impacted by the customer and engineering requirements because those are the only inputs to this document. Once this is constructed though it impacts several systems, this changes the function decomp. to ensure that the most important ERs are being satisfied by a few functions. Also, without the ERs being confirmed through this method the beginning of the design could not take place because there would be no validation that our system would satisfy all requirements.
Functional DecompositionWith the new customer and engineering requirements shown above in the HOQ, the functions required for this system to accomplish needed to be updated as well. Below is the new function tree showing the break down of the higher level functions breaking them down into twenty base level functions that our system needs to accomplish.
Efficiency: Much like the house of quality this is another revision of an existing document. For this a few extra functions needed to be added and numbered. This was a bit more efficient than the HOQ, therefore the efficiency was 25% for the function tree.
Impact: Similar to the house of quality this is directly impacted by the engineering requirements because these drive the key functions for our system. Once this is revised, now the design will be impacted due to the extra functions that need to be encompassed by our system.
Feasibility: Prototyping, Analysis, Simulation
Using the testing bench, shown above, the deflection profile of our HDPE samples was able to be found. Once the deflection profile was matched, the fusion of the material needed to be proven. For this a soldering iron was used because of its very accurate temperature of 183 Celsius which is very close to the melting point of HDPE. However, this testing did not give the desired results. HDPE is a crystalline thermoplastic, meaning that it transitions from solid to a liquid very quickly at a very specific temperature. As can be seen in the graph below an amorphous thermoplastic would be a much better fit for this application. Unfortunately, this was unknown to the group prior to the testing. Therefore, more plastic research and sourcing was required.
For this research twelve companies were contacted as well as a few RIT professors. Below is a table of the suggested plastics, as well as a few companies to buy them from and their glass transition temperature, denoted as "Tg" in the graph above. Currently, polyester (A-PET) and acetate are the highest interest, simply because the companies we contacted have these in stock that match all of our requirements. However, if testing shows these to not be viable solutions then PLA, ABS, PVC, and PMMA are all other options to test.
Finally, to prove that the plastic and assembled valve will be able to withstand twenty years of use the mechanism shown below was designed. With an exaggerated estimation of the same note being played 50 times a day, every day, for 20 years, this testing mechanism would be able to simulate this in approximately 2 hours. After this the fusion area, glue on reed block, and deflection would be inspected to ensure no deterioration.
Efficiency: Due to the poor results from our plastic testing and now basically needing to restart, the efficiency for this is approximately 1.8%. If we knew that the HDPE would not have fused properly then there wouldn't be any time used on testing and only into research. Then if we knew amorphous thermoplastics were the best solution the research would have taken only a fraction of the time.
Impacts: The main impacts that the change in plastic has on the other teammates and the system, is when it comes to modeling the fusion system. Without the exact polymer being used the glass transition temperature is unknown and therefore the heating system for the heat stake can not be designed. Also this could severely impact the budget depending on the polymer and company it is being ordered from. An enormous roll of heat resistant PVC will cost much more and will be harder to design around than a cut to width roll of polyester.
Rivet Hole Feasibility
Our team created a physical model of the hole punch subsystem in the manufacturing process. This model can be seen below:
After various tests of manually applying a force to punch a hole through the plastic film, we measured this force to be about 76N. We currently are supplied a 10mm diameter bore on a linear slide piston that has a maximum pressure of 72 Psi (500,000 Pa). Calculating the maximum shear force of the system, given our parameters, comes out to 39.27 N. This value is substantially less than our estimated necessary force to break through the plastic cleanly at 76N. By rearranging the shear force equation we discovered that we would need a linear slide piston with a bore of 15.29mm.
Looking at our BOM, we do not necessarily have the room in our budget to buy a new linear slide piston. In the next phase we will recalculate the representative amount of force needed to break through the plastic film using a maximum of 72Psi and a 10mm bore by potentially reshaping the edge of the hole punch itself. We should also note that this shear force will vary slightly depending on the plastic film selected for the product.
Efficiency: In order to fully comprehend the necessary fluids and force equations used in this phase's calculations consultation was needed from various different professors. Due to the differing values from calculations to the measured force necessary to punch the hole, additional testing is will need to be done in the following phase to ensure that the maximum parameters for the current linear slide piston design supply enough force to cleanly produce a hole punch. Efficiency for this portion of the phase is about 6%, since we already developed a an entire physical model for this subsystem.
Impacts: This shear force disparity of the linear slide piston will mainly impact the budget if we are unable to produce the hole punch cleanly with the maximum parameters of our current system.
- The tension rolled will be spring loaded to keep a constant tension.
- The drive roller will be running off a stepper motor and the pinch roller will be held in place with a pneumatic piston.
Heat Stake Feasibility
A brief analysis was conducted to get an estimate the worst case time it would take for a heat stake to fuse a the layers of a valve. It was assumed that for proper fusion the layers would have to be brought to the glass transition temperature and then held there while phase change occurred. From the model constructed it was found that it should take 3.278 sec for fusion to take place. This is much less than the expected 10 sec envelope that each operation on the indexing dial must take place in.
Link directly to the PDF of the hand calculations https://drive.google.com/open?id=1zYSTAySPglQs1KnMlRkGJOND_V4LLhde
Drawings and Schematics
The below schematics show the electrical design of the project. the top schematic shows the connections to the Power Supply, PLC and additional I/Os, and the HMI. This schematic shows the main control of the system. The second schematic is sensors which are used for inputs to the PLC. These inputs will help with control of the outputs which will be completed in the next phase.
Flowcharts and Diagrams
To output 1 reed valve every 10 seconds an indexing dial will be used in the design of our project. The indexing dial allows for multiple processes to be done at the same time. The total time to output 1 valve would be equal to the time of our longest stage on the dial. Currently the longest stage is projected to be the second loading stage. If two layers are needed at this stage the estimated time of completion is 10 seconds. From this, it seems that the output of one valve every 10 seconds is feasible. This timing diagram is tied to all of the engineering and customer requirements since it maps out the entire process from start to finish of one reed valve. This chart takes into account the amount of time it takes for each component to return to its initial condition. For example, each individual process on the dial needs to be back in the home position before the indexing dial will rotate.
The Level 1 transformation diagram (shown above) demonstrates the flow of material, energy, and information through the system. Each function block within the system corresponds to the function decomposition.
The Level 2 transformation diagram (shown above) is similar to the Level 1, but slightly more in-depth. It shows functions from the second level of the function decomposition, and shows more of the energy travelling through the system.
The functional block diagram shows how each individual part of the design will interact with each other. The Programmable logic controllers I/Os will send signals to each of the individual steps to start and stop certain processes. This is very similar to the functional block diagram from the previous cycle, but updated to account for all the system and subsystem changes made.
The software flow chart diagram (shown above) illustrates the flow of the software from start to end of a single cycle. Each vertical rectangle shows a sub-process at each station. Immediately after the indexing dial completes its movement, if there is material in the jig at a given station, then that station's subroutine will execute starting from the black circle at the top travelling downwards.
Below are our current Test Plans.
Bill of Material (BOM)
The team has been very lucky with donations from both or customer, Scott Bellinger, as well as from several companies. Currently, the team is still working on incorporating as many of these donations as we can to decrease our cost. With the current design and assuming everything can be used as planned, the bill of material (BOM) below was formed. As can be seen in the top right corner of the image, this is currently over budget. However, the two most expensive components (the Allen Bradley I/O cards) might be acquired from LBJ or we might be able to become sponsored by Allen Bradley for this project. Other than this the BOM is fairly accurate to the most recent revision to our design. We will continue to search for better deals or any donations to lower this cost.
Efficiency: To construct the bill of materials it took the team a few hours of talking through the design, while also counting inputs and outputs of the PLC and reworking the design to lower these I/Os. Due to the multiple team members working on this at once, this efficiency dropped a bit to 12.5%. The main reason that this isn't lower is because if we had to do this all over again knowing the design, a large amount of time was still put into looking up the parts online through the approved vendors.
Impacts: The bill of material is the the main document that will impact the budget, and vise-versa. Since we are currently over budget that will have large impacts on our design as we now need to find other components for less money, or design a way to use fewer components. This in turn changes what needs to be modeled within SolidWorks and what needs to be accounted for in the electrical diagram.
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. Some of the risks identified so far are tied to certain engineering and customer requirements and this is expressed in the FMEA. Risks that do not have any ERs or CRs effect the process as a whole. This could mean effecting whether the project is completed on time or the overall quality of the reed valves.
Phase 3 and 4 Project Plan
Below is our projected Project Plan for Phase 4.
Plans for next phaseAs a team, where do you want to be in three weeks at your next review?
Between now and the end of the semester we need to have one set material locked down and approved through our supplier. We would also like to have a list of potential suppliers/ backup materials for our customer to utilize in case there was ever a need to.During this next phase we will have to look into resources on and off campus for HALT for lifetime testing and validate that our material is capable of what the customer is asking. By the end of the next phase we will have all key items identified for our prototype for this machine to begin building before Spring semester.
Justification that we will finish on time.
We found that we were a little behind on this phase in terms of things to accomplish. There are other teams who are further along then we are but we found that we needed more time really take the time to make sure that we had the proper material for the reed valves. We met our deliverables for this phase but we are not as far along as we would like to be. A lot of our design is contingent on the properties of the valves (thickness, melting point, etc) and by taking the extra time now to make sure that our design will work in the end was more beneficial than randomly choosing a material with only some of required properties. When this phase is over, we will have a sit down meeting as usual and talk about peer evals and how we can help one another for this next phase.
As an individual on the team, what are you doing to help your team achieve these goals?
To view our team goals, please click the link below.