Preliminary Detailed Design
Team Vision for Preliminary Detailed Design Phase
In this phase our team hoped to answer any critical questions regarding the selection of key components to our system such as the piston, solenoid, compressor, and dampening mechanism. We aimed to produce a finalized B.O.M. which reflected a fully developed design we are confident will meet all of our customer requirements.
While we are still in the process of answering some questions, we were able to find answers to many of our key questions and address our main concerns through research, feasibility testing, and discussions with subject matter experts. Additionally we were able to reduce some costs by finding alternate methods of sourcing components for our rig including discussions with the RIT Machine Shop and various professors inquiring about what unused materials were available for use.
Prototyping, Engineering Analysis, Simulation
Questions needed to be answered
- Is the compressibility of air at the systems running region stiff enough to reach our system goals?
- Does the solenoid cycle at 100 Hz?
- Is compounding timing an issue with the solenoid, piston, DAQ and Labview for the 100 Hz?
- Will the flow rate be an issue cycling in and out of the piston?
- Do we stay under our natural frequency of our system? And by how much?
1. The speed of the air needs to be under Mach 0.3 to assume incompressible flow.
If we look at 3 different sizes of inlet valve we can then get an understanding of what flow rate we will need to be under so that we can assume an incompressible flow. So our correctional area is
- A = Pi * r2 (1)
Now knowing the correctional area we can get the velocity for a range of volumetric flow rates with the fallowing equation:
- v = V_dot/A (2)
now knowing the velocity of the flow we can generate a graph of different ranges of inlet sizes. But first lets simplify the speed and put it into Mach number
- M = v/c (3)
where c is the speed of sound under Standard Temperature and Pressure (NOTE: may need to be reassessed based off pressure)
Now we can use Matlab to generate the figure bellow.
As you can see if we go with our current specced out piston we would only be able to run the air at about 1.7 CFMs before we ran into issues with comprehensibility. So knowing this we might need to choose a different piston as the other inlet sizes would be more ideal.
2. We can get a solenoid that will cycle at 100 Hz After meeting with Professor Wellin he assured me not only that 100 Hz is not a big deal for solenoid valves at but he himself had some pneumatics that we could barrow to experiment with. It should be noted that he did not say our system was no problem but 100 Hz was no problem for the solenoid valve.
3. Depending on our choices 100 Hz for a DAQ is not a big deal in any way. In my meeting with Professor Wellin when discussing DAQ devices there was no question that it could handle the speed going with the MyDAQ option, which we have elected to do. These are used in the systems lab currently and could possibly be borrowed.
4. There should be no issues with this as our Amplitudes are not large for our frequencies
To asses this first we have to assume that the flow we are looking at has a constant flow rate, and for this system we do assume that to be true.
Now we need to know what the Volume is that we are filling and for this part we will assume that it is just the piston. We will then assume that volume to be a cylinder so that the bore hole of our piston a the diameter (D) and the amplitude is the hight (H). Giving us
- V = pi*(D/2)^2*H (1)
But then we also have to account for the change in diameter for the flow from the tube to the bore size which is assessed in the fallowing way
- V_B = V*(D_B/D_T)^2 (2)
Where V_B is the volume of the bore hole, D_B is the diameter of the bore hole and D_T is the diameter of the tubing.
Now we can asses the time it will take to to fill the volume based off the flow rate that is a predetermined flow rate.
- t = V_B/V_dot (3)
Now we can asses what our system will end up looking like under our two extreme conditions. First looking at the larger amplitude we see the effects of different bore sizes.
From this graph we can see that the flow rate for our system will be close depending on the bore size we end up going with.
For the different bore sizes we can also look at the frequency for our ideal amplitude we get the fallowing graph:
This image is hard to see so bellow is a zoomed in version of the same graph:
We see that the ideal flow rate for 100 Hz is low and definitely reachable.
We realize though that pneumatics are not exactly the most precise controllers for such a small Amplitude. So we reassessed problem and used a larger amplitude to see what the results would be.
This shows though even if we try to hit this amplitude we are well within the range of our limits.
5. Still verifying
Initial assessment: read through the handout from Dr. Kempski and working on figuring it out
Questions needed to be answered
- What vibration solution will meet our damping goals?
- What feasibility analysis can be done to determine that the chosen solution can get the job done?
Answers1. Sorbothane Hemispheres
Sorbothane hemisphere mounts are visco-elastic polymers that provide a quick, cost-effective method of isolating small machinery and bench equipment. It combines shock absorption, good memory, long fatigue life and vibration isolation/vibration damping characteristics.
Sorbothane has a high damping coefficient compared to other polymers (rubber, neoprene, silicone, etc.).
2. Feasibility Analysis
Replicate spinning mass project from Engineering Applications with motor and crank system. Extra mass can be added to the base plate to simulate our overestimated 50 lb system mass. The motor provides about 4.5 Hz which will be a great starting block for determining whether or not our hemispheres can dampen our real system. Professor Wellin also has a motor and crank that we will plan to use to set up our feasibility test.
Mounting the Motor
Option A is to attach the motor to a right angle plate and drill a hole through the vertical part of the plate to allow the shaft to go through.
Option B is to sit the motor on top of the base plate, drill press two parallel holes on each side of the motor and strap the motor down with a hose clamp and bands.
User Interface and Controls
Questions needed to be answered
- How can we reduce any latency associated with the computer interface and code?
- What will the system output to the user to validate accuracy?
Answers1. After meeting with Dr. Kempski and Professor Wellin we are confident that the response time from the computer system will not be an issue. The Lab view program and associated DAQ device are designed with the intent of a near instantaneous response. However it is still best to code in parallel increase efficiency and reduce usage of excess processing power. The diagram below shows an overhead view of the critical processes running the system. The processes can be broken down into a section of providing inputs to the solenoid and flow control valves, and a section of sampling and validating data.
In order to assist in maintain a low response time, the system control portion of the program will use an internal clock to track time as an input, while the sampling and validation portion of the program will use a count of the samples to track time. The sampling and validation portion will accomplish this by applying a count to the samples and then dividing the count number by the sampling frequency to achieve the sample time. As an additional measure to reduce response time, the system will not wait for validation before sending a signal to the control portion. Instead the control portion will continue to run normally unless a corrective signal from the validation portion interrupts it.
2. The systems position sensor will output a data stream which by knowing the sampling frequency, we can use to track the peak acceleration, frequency, and time. These three values will be output to a data file which will then be used to produce a graph of acceleration vs time and of frequency vs time at the end of testing.
Sampling and Validation
Questions Needed to be Answered
- How can we be reasonably certain that we are finding the peak acceleration or displacement values with our sensors?
- How will adjustments be made if measured values are not within tolerance?
Answers1. Since the pneumatic air piston will act as an air spring, it's motion and acceleration can be modeled as a sinusoid. We can exploit this sinusoidal nature to calculate the actual peak values of the system by applying the formula Y=A*sin(B*T). In this formula A is the peak displacement, and B is a frequency coefficient which causes the necessary period of the sine wave. The figure below represents a sample set off data where the black line represents the motion or acceleration of the piston and has two sets of data points. The red set of data points represent the actual peaks and zeros (PR1,ZR1,PR2,ZR2). The green set of data points represent a theoretical set of samples (PG1,ZG1,PG2,Zg2) from a position or acceleration sensor functioning at 4 times the testing frequency (400 Hz at max testing frequency).
For the two green data points PG1 and ZG1 representing a point near a peak and another near a zero, we can adjust the general formula to:
By solving for A we can set these equations equal to each other and solve for TPG1, yielding the equation:
Where TP is the time difference between samples the two samples (in this example 2.5 ms). Since we know use the time TPG1, we can solve for the value A which will be equal to the Peak using the equation:
2.After finding the 2 most recent peaks (Maximum and Minimum), we can find the time between them by knowing how many samples were taking in between them and the frequency of the samples. We can then calculate frequency using the formula:
With the values for peak acceleration and frequency we can compare them to the theoretical values for validation using the equation:
If the values of Pf or PA are greater than 1% a signal will be sent through the feedback loop to adjust to the inputs according to the formulas:
Where SA is the amplitude input to the solenoid, and TS is the time between switching.
Sensors and Data Collection
Questions needed to be answeredHow are we going to ensure that our rig is functioning to the specifications required to pass the vibration testing? How accurate do these devices need to be and what is the ideal sampling rate?
AnswersSince our maximum vibration is 100hz, our sampling rate will need to be at least five times as high to be able to form the sine wave needed to analyze the data properly. Ideally, we would like a sampling rate of nearly 1khz or above to get a good output from our system and accurately validate the test. With the accelerations being between -1.20g’s and 1.20g’s, the accelerometer will need to be sensitive enough to collect data within this range. Since the smallest decimal point needed is 0.01g, the sensor will also need to be accurate enough to detect this small change in acceleration. Similarly, the proximity sensor will need to be able to detect movements between 0.00mm and 26.00mm. With the smallest decimal point needed being 0.05mm for an accurate test. To be able to detect this small of displacement, this sensor will need to be very accurate.
Accelerometer Since the testing specifies both frequency and acceleration, those will be the focus of the sensors and calibration. While industrial accelerometers are very precise and durable, they are also very costly. However, smaller sensors made for robotics are capable of detecting the accelerations produced with high accuracy and a fast sampling rate for what we need.
The product shown above is an accelerometer from Adafruit that meets the specifications set on it. It is capable of sampling at a maximum of 3200hz and has four ranges of acceleration. It is capable of switching between ±2g, ±4g, ±8g, and ±16g testing to be able to increase the range of the device while decreasing the resolution. When the device is set for ±2g testing, it will give us a 3.9mg/LSB resolution while giving us 31.2mg/LSB while in ±16g testing. Having this device will allow us to make changes to the program later while still being capable of using the same device. It is a minimal cost increase to the system ($5-$10) but will help to ensure that we are capable of running with a varying program.
To supplement an accelerometer in our system, we are also researching proximity sensors to be able to show the amplitude that the piston is driving. The two most common proximity sensors are ultrasonic and infrared. Since our test rig is going to be producing stray noise, it is best to focus on an infrared sensor. The issue we are having is trying to find a sensor with a 0.05mm sensitivity and above a 400hz sampling rate. While both of these are capable of finding in a device, they are very costly and go beyond our budget with the sources we have looked at. Industrial sensors are capable of meeting these parameters and is an option we are exploring for the next review.
Drawings, Schematics, Flow Charts, Simulations
- What does our system model look like?
- Is there anything that will be updated going forward
- Based on how the current CAD model looks, will there be any unforeseen manufacturing issues.
- The E Stop will be placed closer to the computer as that is where the operator is most likely to be.
- A form of a guide rail will be placed around the top half of our system to ensure the system is not able to go off center, or worse, bend the piston.
- We will be placing P16102's CubeSat into our model to have an more accurate representation.
- Our only small concerns of manufacturing are producing the inner rails of the CubeSat, which we may be able to make of a solid rail since each corner is not used.The other concern being securing the piston and solenoid to the damping base, which can be easily done with the proper fasteners.
Bill of Material (BOM)These are our current BOM's/ They will continue to be updated accordingly before our next presentation.
- What different parts does our build encompass?
- What parts need a longer lead time due to shipping or secondary operations.
- Is this entire project economically feasible with the $500 budget provided?
- What parts can we acquire for free or have on loan from professors?
- We do not currently have significant lead times for shipping or secondary operations on any of our parts. The largest one could possibly be is 3 weeks which is still within reason of our timeline.
- The entire project is most certainly economically feasible. After acquiring a solenoid and piston from the engineering basement, locating scrap metal in the machine shop, and finding other parts in "scrap" piles, we have saved a significant amount of money. Also changing from an air compressor to shop air saved significant money.
Fully Sourced BOM
Cost Effective BOM
- Updated Risk Assessment with new assessment of current issues being faced.
- Note: the two biggest edit is the current issue on cost and the change of importance of not hitting the right frequencies.
Our current largest concern with the vibration requirements is the inability to achieve the lowest magnitude of displacements. If the displacement is higher than required it will result in the Cube Sat experiencing a higher acceleration. This will be most prominent at a frequency of 100 hz. Below is a table and graphical representation of the resulting acceleration for a range of displacement values.
Questions needed to be answeredHow can we ensure safety for the users?
AnswersDue to the nature of our test rig, it will require human interaction with close proximity. Because of this, safety is a primary risk and a concern in the system design. After benchmarking the packaging science vibration test rig, it was noticed that the main safety device implemented is an emergency stop button. In this design, we will implement two emergency stop buttons in series to allow for differing stop mechanics. The first will be a stop button and loop control in the LabView control system to stop the program as needed. The second will be a physical emergency stop button to disconnect the power supply from the solenoid valve. This disconnect will immediately stop the piston from moving and bring the vibration test rig to a resting state.
Labview Control System
In order to stop the program, an emergency stop button and a true/false case loop will be implemented. This loop will surround the entire control system and will stop the program from running. It is a simple and free way to make sure there is a way to end the program as needed.
Emergency Stop Button
There are many choices for a physical emergency stop button that can be implemented. They range in cost, size, and level of safety. Below are some options sourced in order by cost and safety.
The first two options are simple switches that are inexpensive but are less recognizable. They don’t reflect the industry standard and may be an increased risk to the users.
The next switch type is a more robust switch for a higher price but is still not as recognizable as the traditional emergency stop button.
The next two choices are standard emergency stop buttons that are expensive but are recognizable and reflect the industry standard. The first is a plastic switch and the second is a metal version of the same switch. They are both push to stop and turn to reset style buttons.
The final choice on the chart is a high vibration emergency stop button. It reflects the standard for emergency stop buttons that need more secure wiring. This choice would be the best choice available for our test rig.
However, the machine shop faculty had an emergency stop button available for us and was generous enough to donate it to our senior design team. The button available is a heavy duty emergency stop button with a bottom panel that supports the wiring and switch. This button will be mounted to the base of the structure where the vibration requirements are significantly lessened, making it a suitable choice for our system.
Test PlanThe gantt chart below shows a high level overview of our anticipated schedule from now until the end of MSD II. It is modeled in terms ofeach month being split up into four quarters.Our goal is to provide a fully functional test rig for P16102's usage by early April so they can test their structure with at least a minimal level of functionality.
Plans for next phase
Deliverables by Owner
Our goals and tasks going forward:
1. Finite Element Analysis results
2. Source a proximity sensor that meets our requirements
1. Further analyze the natural frequency of the system.
2. Make an assessment of parts from Professor Wellin and decide what needs to be purchased.
1. Feasibility analysis of Sorbothane hemispheres using spinning mass
2. 50 durometer versus 70 durometer hemisphere analysis
1. Detailed drawings of parts in order to provide adequate knowledge of the required machining process.
2. A fully sourced bill of materials to have a precise and accurate representation of our project. This ensures we have accounted for all parts as well as the economic feasibility.
1. Further explore any challenges or potential failure points associated with the coding and controls of the system.
2. Begin purchasing of materials and components, especially those with high lead times.
3. Develop and finalize build and test schedule for MSD II.