Integrated System Build & Test
Table of Contents
This phase our team has been working to ensure that all subsystems are properly tested and verified in order to be integrated into one system. The documents for the phase are located here in the Integrated System Build and Test directory.
Team Vision for Integrated System Build & Test Phase
Our goals from last phase:
1. Continue to work with accelerometer board and validate on packaging science test rig with the position sensor
2. Work with the team to understand and use the new solenoid valve and controller
1. Work with new solenoid to get an understanding of how it works
2. Order and work with the new Piston
1. Continue thorough testing of system with accelerometer, solenoid and new piston
2. Assist with solenoid
1. Complete construction of P Pod
2. Order the rest of the necessary parts
1. Implement DAQ connections into labview code
2. Test Solenoid with sine wave output
Test Results Summary
Accelerometer and Damping Testing
Melissa and Tim went to the Packaging Science lab to validate the accuracy of the data from our accelerometer and to test the damping hemispheres. We were able to use the vibration table with preset programs to see how well the data gathered correlates to the applied accelerations.
Unfortunately, the data came back inconclusive as we experienced a few issues. The first was that the data simply did not match up to what was being applied. We believe that the primary cause of this is a power supply issue on the arduino board where the sensor isnt getting full power. To fix this issue, we are going to test this process again with a different arduino board. If this is the cause, then we can go through to purchasing a new arduino board to be able to sample the acceleration for each test.
On top of that, the system was not able to sample at the rate that we needed. We are looking into ways to increase the output rate from the sensor such that it hits its maximum sampling rate as it is rated for more than what we need. This could be a communication issue or a delay in the code causing it to be slower than desired. To fix this problem, we will be reevaluating the code to work on bringing it to its maximum sampling rate.
Shop Air Piston Flow Verification & Validation
We wanted to purchase a flow meter to help us determine accurate levels of mass flow going into and out of out piston. After doing research on digital air flow meters, we concluded that we would not be able to purchase one due to cost. An example of a suitable flow meter (shown below) was $522 and would not fit into our current budget.
To work around this constraint, we plan to ensure that the resulting frequencies and amplitudes we expect from our system are accurate through accelerometer and position sensor readings.
Testing the pneumatics this phase was focused on getting an understanding of our new components and how they react inside our system. This included the new piston, the controller and the Solenoid valve.
We first started with just testing our piston to see if it worked with the components (old solenoid valve) we already understood. This provided us with a hands on understanding with how the piston worked in case there was any unforeseen elements which there weren’t. We connected the piston to a bold with the right threading but look to attach a threaded rod to the connector to connect to our test mass we have been using.
The controller is powered by a +/- 12 Vdc from a power generator and then the function generator is connected to the signal input and grounded. The solenoid then hooks into the m+ and m- ports. The big thing that needs to be noted is that the board recommends a maximum amperage of 1 amp. This is crucial to note to protect the board. There is a green LED on the board that will light up to signal a +/- 12 Vdc connection.
The solenoid is hooked up to the hooked up to the controller via the brown and blue wires that were hooked up to the m+ and m- wires respectively. The  and  ports were hooked up to the piston connections and the  port is hooked up to the feed of the flow.
The new valve was put through a few different tests. The first of which was a dry run of the piston connected to nothing but the controller and an oscilloscope. This test proved that we could in fact cycle at 100 Hz and was validated off the current feedback from the oscilloscope. The next test was hooking the solenoid up to the old piston and seeing what kind of responses we could see. This resulted in movements similar to what we saw before validating the use of our solenoid. The most recent test we did was hooking the new piston up to the system with no weight and weight on the piston. The test with the weight resulted in a reaction of the system which is the first step to what we need to accomplish and the second test of running the piston without any weight resulted in a reaction at 100 Hz.
Design and AssemblyBuilding has gone well, just been taking much longer than expected.
- Many delays were due to consulting with the machine shop and Brinkman lab to solve tolerance stack up issue. Therefore the consulting time, followed by the time required to update CAD.
- Rails are split in half to two 8" segments, for a total of 8 pieces. This is an updated from the initial design of having it split into 3 pieces. The trade off is some slight tolerance loss on ends, but a larger beneficial trade off of machining time.
- 2/3 of the required rails are machined. The final third can be completed within 1-2 days of machining. Melissa helped and learned to use the mill more proficiently so she can even make future parts if needed.
- Tolerance stack up is a concern. Have been machining for nominal, but the mills are only so "accurate". All parts are within tolerance as machined.
Note: Screws are not correct length. Will be updated to correct length pending sheet metal purchase.
Going forward, I look to finally complete this. I have learned to better estimate machining time. Also, that sometimes there comes a point when you have to stop taking advice and just act upon your own, thus to prevent delays.
John Bonzo in the Brinkman Lab was nice enough to begin waterjetting our p-pod sides. Unfortunately, there were difficulties machining and the parts were not usable. However, before we arrived to pick up our initial request he had re ran the parts out of 0.060" thick 304 stainless steel that was available to replace our stock. Running these new dimensions through a COMSOL analysis, it was found that the eigenfrequency was only 95hz. The current pieces can be used for lower end testing still.
The original dimensions included 0.100" thick walls of the P-Pod which had passed the COMSOL simulation with a good margin.
To remedy this problem, we are going to purchase a new piece of sheet metal with a thickness that passes the COMSOL analysis. Below are our two options showing that both 304SS and 1018 steel are viable options. For now we are going to order and use 1018 steel as is is approximately 1/8 the cost of stainless sheets. The stainless steel would be an easy and quick aesthetic upgrade to keep the future users in mind, so budget permitting, this may be later pursued.We are also still working on sourcing the sheet metal, thus it is not yet reflected on the BOM, local suppliers may have more cost effective options.
A sequence loop was used to add a time delay between the start of the program and the beginning of testing. It has a default value of 30 seconds with a minimum value of 5 seconds.
The required components were added to the code in order to create a waveform signal to out put to a DAQ device. The maximum and minimum values for output voltage, the type of waveform, and number of samples were all set as constants to prevent input mistakes and ensure ease of use. The waveform generator pulls the frequency values from the if loop structure previously implemented. If an error occurs in the DAQ process, it outputs and error and the program will stop and report the error.
Listed is our updated BOM. As shown, we are still within a comfortable zone for our budget.
Risk and Problem Tracking
- We have been able to take away 1 Risk from our list based off the fact that we no longer see them as a risk
Plans for next phase
For the next phase our team is aiming to have a functional test rig to present. This would allow sufficient time to make minor adjustments and fixes to the system, improve ease of use, and test P16102.
This will be accomplished by sourcing new sheet metal for the P-pod and submitting it for water jetting over break. The P-pod will be constructed and assembled early in the next phase, leaving time for the remaining top assembly to be constructed and connected with the solenoid-piston assembly.
The base will be drilled and tapped and the base assembly will be constructed and assembled. Testing will occur using connections between the Labview controls code and solenoid-piston assembly.
The accelerometer will be tested again using a more suitable arduino so that we may achieve more accurate validation data. After receiving this data we will implement code into our controls program to receive the data from the arduino, analyze it, and output the critical information.
As we approach the end of the semester and Imagine RIT we plan to allocate more resources towards our final paper, design poster, and presentation. For next phase we are aiming to have a paper outline and high level poster design completed. These will place focus on critical design decisions and the rational behind them, key testing and validation, and the final product. Furthermore our paper will offer a more in depth analysis into other design options considered, feasibility testing, and challenges that have been meet and overcome during the process. Lastly our paper will discuss potential future improvements to the rigs and opportunities for future design teams to utilize the lessons we have learned during the design and build process.
Our goals for next phase:
1. Work with new arduino board and validate the accelerometer
2. Work with the team to begin assembling subsystems
1. Test our system at different frequencies with an equivalent mass
2. Start to work along side with the myDaq and code to test the system
1. Complete base plate machining to integrate with piston and damping hemispheres.
2. Continue to assist the team with the tasks required to assemble our system.
1. Actually complete P Pod
2. Continue to machine all parts for our system, utilizing team members as they are available.
1. Begin testing solenoid-piston assembly with DAQ connection.
2. Incorporate accelerometer data collection system into code.