In Phase 4, Team Spring Breaker confirmed that its proof of concept (POC) worked and from that, the team moved forward on its final detailed design. This phase content of our Edge page includes detailed information and outcomes of our journey through our detailed design phase (Phase 4). It also includes links to important documentation completed during this design phase.
Team Vision for Detailed Design Phase
In our last phase review, each team member developed plans for our next phase. Here’s a list of those plans:
- Ensure that the final design meets all engineering/customer requirements.
- Update cartridge design.
- Construct new POC out of aluminum tubing.
- Conduct testing on POC and analyze data.
- Keep track of items in BOM and budget.
- Research sensors (linear potentiometer and pressure transducer) and their outputs
- Research how to interact with sensors to record data and plot it
- Update MATLAB to have all functions required by customer
- Finalize CAD and develop detailed drawings
- Produce ANSYS on stressed structures
- Consult with client regarding design direction
- Call vendors about hydraulic cylinders to determine our best path.
- Schedule plans for MSDII and determine critical path and slack in the schedule.
Here, we summarize our accomplishments and other important topics as we complete Phase 4, our detailed design.
What the team plans to accomplish in this Detailed Design Review
- Our team will present our final design to our clients Meggitt.
- We will formally request permission to move forward on this final design.
- We will show what elements of the design have remained from Phase 3, what elements have changed, and what calculations we have completed for the final design so that we can begin ordering materials.
- We will verify that our proposed final design meets the CRs and ERs of the project. We will also ensure that it upholds the House of Quality.
- We will also review our planned schedule for this phase, showing that we are now able to complete critical path analysis.
Tasks have been accomplished so far
This section summarizes our major accomplishments broken out by functional areas.
- Completed spring disc calculations and analysis and determined type and number of springs for final design.
- Completed preliminary CAD drawings for final design.
- Completed design on cartridge stop off.
- Determined possible sensor for final design.
- Converted spring disc calculations into MatLab.
- Completed test plans for next phase.
- Completed end of semester status report.
- Confirmed that design meets ERs and CRs
- Completed final Proof of Concept testing and proved concept.
- Completed critical path analysis for Phase 3 and started one for Phase 4.
- Determined what materials we can order now.
Tasks remain, and who is the owner of each
List of Tasks
Decisions that have been made so far These key decisions have been made regarding the design:
- We will use a combination of die springs and spring discs
- We will use only one piston.
- We will not incorporate the graduated reservoir design unless time allows.
Team Progress Reports for Phase 4
Throughout Phase 3 and 4, our team kept detailed weekly minutes of our meeting and accomplishments. We have included a PDF of all Phase 4 weekly meeting minutes. WeeklyMinutesPhase4
Prototyping, Engineering Analysis, Simulation
During the previous design phase, we created a prototype to test our spring concept and discover and potential problems in our design concept. The prototype contained two concentric springs in a PVC pipe with 3D printed backstops. A piston was placed inside another PVC pipe. The plan was to pressurize the piston housing with a bike pump, forcing the piston into contact with the springs. We would measure the pressure and the spring displacement to recreate the pressure vs. displacement graph Meggitt requires.
Before testing, we learned from Prof. Wellin that PVC pipe is not rated for compressed air and could potentially shatter. We therefore abandoned this test method in the interest of safety. Our initial thought was to replace the PVC piston housing with an aluminum pipe. We discussed this with Dr. Gomes to double check that the aluminum would be safe to pressurize. During our meeting, he gave us the idea of testing the springs with various masses, letting gravity apply the force. For our second test, we placed the spring housing vertically and hung a “basket” from the bottom to contain the weights. The basket was attached to a rod which went through the springs. At the top of the rod was the piston head, which would compress the springs as the masses pulled on it.
We measured each mass added to the system and calculated the force due to gravity that was acting upon the springs. We also marked the rod for each added mass to show the displacement, and measured it with a caliper. This gave us a force vs. displacement graph. Force is proportional to pressure, so this graph is analogous to what Meggitt wants.
The data did not give us the expected curve. We calculated a turning point at around 33 lbs of extra weight, which is where we get the flat slope in the middle. The first and third slope are also not distinct enough to be the slopes we wanted. Our initial guess was that this could be corrected manually. When attempting to code the setup in MATLAB, it was found that there is a transition slope between the two main slopes unless the second slope is given a y-intercept to connect it to the first slope. If this was the case, we would get the following graph. There are two distinct slopes, though the ratio between them is smaller than desired.
After conversing with Kyle about our results, he suggested that we may have simply missed the first curve during our test. We revisited our calculations and discovered he was right: the turning point was actually around 7 lbs, and our basket was over 8 lbs. By adding the basket to the assembly, we skipped the first curve entirely. We tested the prototype for a third time to get more data. This time we factored in the lessons we learned from the previous test: add masses at linear increments, and be more precise when measuring displacements. Instead of adding a heavy basket, we hung the masses from the end of the rod.
We added masses in 1 lb increments until we were past the turning point, and then added a few 2 lb masses for extra data points. Our graph clearly shows the two slopes.
Adding in the data points from the second test gives us the desired and expected graph, showing that our proof of concept is correct.
CAD / Designs
Spring Disc CalculationsDesigns were developed on expected spring disc stack heights and the dimensions of the spring discs used.
Step 1: Using Meggitt's previous test data for six different brake programs, ranging from Very Small Jet to Large Commercial Jet, hydraulic displacement data and pressure were converted into desired linear displacement and force respectively. From this desired spring rates for each section of the curve were found. A MATLAB script shown here SpringCalculations.m is used to convert the data. Using nominal pressure values of 300O PSI, and a bore diameter of 3.25 in. data is ported to step 2.
Step 2: An excel spreadsheet NewerSchnorrSpringDiscCalculator.xlsx uses the MATLAB data as an input and determines the necessary parallel and series stacks of a certain kind of spring disc to achieve the desired spring rate and displacement with +/- 5%. Using the spring force equation for a single spring disc as shown, basic Hooke's laws are applied to determine the necessary spring disc values. Discs stacked in parallel multiply load, while discs stacked in series multiply possibly displacement.
Step 3: The entire Schnorr Spring Disc Catalog is searched by the excel spreadsheet to determine the optimal type of disc and stack configuration capable of achieving the desired displacement, force and spring rate for a given brake program. Design parameters for selecting the spring disc are linearity, tolerance from desired loading, tolerance from desired spring rate, and number of spring discs per stack. What results is the table of spring discs below.
The piston will push the cylinder holding the larger spring discs as one unit, as the active spring discs will the ones with the smaller k value. Once the cylinder assembly reaches the stop off, then the larger spring discs will become active. The base will then be subjected to the largest portion of the applied force. Stress analysis can be performed to verify that the base will withstand the force.
Final Design: Spring Pack Configuration
The second design is functionally and conceptually equivalent to the original design, but with the addition of a preload that is applied to the springs to ensure that there are smooth transitions between k values in the pressure vs. displacement curves that will be output, and to increase the working life of the spring discs. The base in this design is more robust than in the original, and incorporates a bolting pattern that allows the base to be fastened securely to the steel plate and the cart in order to better withstand the applied force.
As shown, Pack A (smaller cylinder) and Pack B (larger cylinder) provide the small and large curves necessary to replicate a brake's behavior. These packs can be customized to any spring rate desired using the analysis method as shown earlier.
Pack A uses a center rod and nylon shims to keep the spring discs in place. Pack B uses a cylindrical nylon shim to center its spring discs. Both the centering rod of Pack A, the centering shims of Pack A and the centering nylon cylinder of Pack B must be machined to match the ID and OD respectively of the spring discs used. As such for each program a new set of these shims must be produced. Additionally, the Pack B head must be machined to match the ID of the Pack B cylindrical shim for each new program. As such, these four (Pack A rod and base, Pack B head and nylon shim) components are the only ones that will need to be swapped when adding programs. As per the team's detailed design review on May 4th, 2018 we will be moving forward with the spring packs final design for the AABS system.
Manufacturing Prints for Final DesignThe prints as shown will be used in MSD II to fabricate the spring pack system for use in conjunction with the piston to provide displacement data.
Complete SystemThe final design incorporates a hydraulic cylinder, the spring pack loading system, as well as bleed air system, micrometer valve, check valve, fluid restrictor and red oil reservoir. In additional a magnetic position sensor and fluid pressure transducer will be used to measured the systems output linear displacement and on-board pressure respectively, as shown in purple. A hand-pump provides fluid pressurization and a reservoir supplies the fluid. These two items are not included on the cart.
Bill of Material (BOM)
For this phase, two BOMs were made in order to compare the total number of components, sub-assemblies, and cost of the cartridge and spring pack designs. The BOMs include a level of assembly, manufacturing number (if applicable), description, whether it is bought or made, and which sub-assembly each component fits into. The spring pack design was found to be more cost-effective, easier to manufacture, but included more purchased items.
Preliminary Test Plans
- Zero pressure transducer
- Zero linear potentiometer
- Verify that graduated cylinder is empty
- Fill system with hydraulic fluid.
- Let some fluid out of the tap so that air is fully bled.
- Choose program corresponding to jet size
- Place program [cartridge/spring packs] in fixture
- Adjust distance of stop-offs as required
- Slowly pressurize fluid until maximum program pressure is reached
- Record pressure and displacement
- Loosen micrometer valve to release a small increment of hydraulic fluid into the reservoir
- Record volumetric/linear displacement
- Volumetric will be subtracted from the initial volume
- Repeat steps 5 and 6 until pressure has returned to zero
- Plot pressure-displacement points
- Compare resulting plot to brake tests if available
- Setup #1-4 must occur within 30 minutes to comply with ER3.
- Setup #4 must occur within 5 minutes to comply with ER5.
- Simulation #2 must occur within 8 hours to comply with ER4.
We added three new risks for this phase. The first two deal with time management for the end of the semester. We need to get the design finished and the components ordered by the end of MSD I so that we can start building and testing immediately in MSD II. The big problem is the spring disk calculations, but now that those are solved, we are able to move forward with the rest of the tasks we need to complete. The third risk is that the spring disk calculator gives a stack of spring disks for the specific slope we input, but the output is not exactly what we wanted. There is a tolerance of about 5% between the desired and expected slope. We will work to close that gap, and also inform Meggitt.
We updated all risks involving the prototype to a likelihood of 0. Two of them involved the prototype, and we found a testing method that excludes that. The third was that we would not get an accurate pressure vs displacement curve and we would have to rethink our design, but we collected our data and found that we got exactly what we expected.
We also set the risk of being unable to find good springs to 0, as we switched to spring disks to avoid this problem. Switching to spring disks allows us to lower the likelihood of the simulator being too large as well, as the spring disks have smaller displacements.
Finally, we were able to lower the likelihood of hydraulic leakage from 6 to 1, as we are going to buy a hydraulic cylinder instead of manufacturing one ourselves. This was our most important risk, with an importance of 54, but it is now done to an importance of 9.
Design Review Materials
Plans for next phase
- Elias's P5 3 Week Plan
- Evan's P5 3 Week Plan
- Lily's P5 3 Week Plan
- Mike's P5 3 Week Plan
- Tony's P5 3 Week Plan
- Garrett's P5 3 Week Plan
- Sabrina's P5 3 Week Plan
- Lori's P5 3 Week Plan