The Systems Design phase is where we breakdown the functions our system needs to provide and brainstorm solutions to accomplish those functions. This happened in several steps that are detailed below.
Some of the important documents that were used and created during this phase can be found in the Systems Level Design Documents directory. Feel free to explore this directory to get a better understanding of the process used during our system level design phase.
Team Vision for System-Level Design Phase
The goal for this phase is to perform a functional decomposition to determine what functions our system will need to perform. Concepts to complete those functions will be brainstormed and evaluated to select the best concept. A system architecture will be created to relate the functions and subsystems of a chosen concept.
During this phase a functional decomposition was completed to generate a function tree and transformation diagram. Using the function tree, a morphological chart was created with possible concepts to complete the functions. Concepts were selected from the morphological chart and compared using a Pugh Chart. The best concept was determined from our selection criteria and a system architecture was created to relate the functions of our chosen system.
During this phase we had a conference call with our customer, Tom Trabold, and Biomass Controls, the company that makes the refinery. We spoke to the engineers at Biomass about where to place temperature and oxygen sensors. When discussing the oxygen sensors, the engineers brought up that measuring the oxygen level would be difficult given the time frame and budget for this project. Therefore, we will only be looking to incorporate temperature sensors into the data acquisition device.
The transformation diagram shown below relates the inputs and outputs of information, energy, and materials and how our system will transforms the inputs into the outputs.
The function tree shown below shows a breakdown of how we will perform the highest level function of creating a temperature profile. As you go down the function tree, the question of how this function is accomplished should be answered and as you go up the tree, the question of why the function is needed should be answered.
These documents were edited to the current revisions based on suggestions made during the System Level Design Review
Because there does not exist very similar applications for temperature data acquisition systems, our team decided to take a comprehensive approach at benchmarking this technology. We took three DAQ systems from well known companies (Omega, National Instruments, Graphtec) with similar costs to our $500 budget, as well as the existing system (Kelvin) implemented in the biogenic refinery, and compared them using metrics from our Engineering Requirements. This gave us a general idea of how we wanted our concept to operate.
The full document can be found here.
Morphological Chart and Concept Selection
The morphological chart below shows the different solutions brainstormed to perform the functions listed in the left column. The similar colored boxes around the diagrams show a few of the potential initial concepts that will be evaluated.
From our morphological chart, initial concepts were chosen that were thought to be realistic and feasible.
Feasibility: Prototyping, Analysis, Simulation
The overall feasibility document can be found here.
Thermocouples produce a temperature-dependent voltage as a result of the thermoelectric effect, and this voltage can be interpreted to measure temperature.
An RTD, also known as a resistance temperature detector, measures temperature by correlating the resistance of the RTD element with temperature.
A thermistor is a thermally sensitive resistor that exhibits a large, predictable, and precise change in resistance correlated to variations in temperature.
After a detailed research about the suitable type of sensors needed in the refinery, it was concluded that Thermocouple sensors would be the best fit for the Biogenic Refinery system. Thermocouple is the most suitable type as it has the longest temperature range of all the sensors. K type thermocouple and J type would be appropriate as they would measure and withstand machinery operating temperatures. Although TC's do not have best accuracy as compared to RTD's or NTC's, they're still the best fit temperature sensors for this machinery.
Analysis was done on three different possible choices for the control system. The three choices investigated were a computer (e.g. a laptop), a small computer (e.g. a Raspberry Pi), and a microcontroller (e.g. MSP430). The different capabilities of the three were compared as above and it was determined that the small computer option was the best, as it had the fewest drawbacks and was reasonably priced.
For our first iteration using the Pugh chart, we determined the selection criteria to be:
- Sensor Range
- Sensor Accuracy
- Sensor Durability
- Easily Understood Display
- System Build Time
- Effectiveness of Safety Measures
- Efficiency of Safety Measures
- Data Throughput
- System Versatility
- Setup Time
We then were able to have a conference call with our customer and the company that makes the biogenic refinery and it was decided that we would no longer measure the oxygen content through the process as it would be a complicated process and beyond the scope and resources available for this project. We removed the oxygen sensors from the concepts and did another iteration of the Pugh chart comparison, adding the selection criteria of 'Expandable' in order to be sure our system can be used in the future to add different data sensing capabilities.
We determined that concept five was the best concept from the first iteration since it had the least amount of negatives and after shifting the datum it remained as having the most number of '+'. For the next iteration we refined our selection criteria to better evaluate the remaining concepts, and created concepts six, seven, and eight by picking the subsystems that had contributed to being a positive for the concept.
The final iteration compared the best three concepts determined from the second iteration. The selection criteria was refined again to remove criteria that had similar evaluations in the second iteration. It was determined that the reset time would be similar between all system concepts and was not an important criteria for the system.
We determined that Concept 7 would be the best to continue with, minus using fans to protect from internal overheating. We determined that using a small computer, such as a Raspberry Pi, would give us versatility in the programming language we could use and allow us to expand the system easily. Using a high level code, such as C, we could create a robust code to accomplish the necessary functions that would be simple and accessible to many people. Using a protective case would be the most cost-effective way to protect the sensitive electronics and keep the system compact. Using warning sounds, lights, and signs would be simple to incorporate and allow most anybody using our system to recognize when something is not working correctly. Using a touch screen to control user inputs as well as being the data display would be simple and effective. We determined that there is a low risk of the system overheating internally or being damaged by an electrical abnormality so having no safety protections would help lower the cost.
More information on the Pugh Chart can be found here.
Archived DocumentThe changes made above were based off of suggestions from the System Level Design Review. The change was for the Function "Transfer Data to Computer" from "USB" to " USB and Wireless/Network".
Current DocumentsIn our system architecture below, we relate how the different subsystems interact and the flow of information through our system. Since our system is a data acquisition device, there is no material that moves within it and the only energy would be what is required to power the small computer.
Archived DocumentsThe updates seen above were made to the original System Architecture documents based on suggestions made during the System Level Design Review.
Designs and Flowcharts
The Data Flowchart serves to define a high-level overview of the elements in the entire system. A detailed description of the flowchart can be found here.
The updated risk assessment can be found here.
Plans for next phaseOur next steps and tasks for the next phase, Preliminary Detailed Design, can be found here. The full Gantt Chart, which is a living document,can be found here.
During our System Level Design Review, we recieved great feedback from both our Customer and our guide. The summary of the review can be found here.
We were giving suggestions to on our documentation, and will be making these changes in the next coming weeks. Here is a list of those changes:
- Change micro-controller to control unit in transformation diagram
- In transformation diagram we should split up bullet
- Split up energy to electricity and thermal
- Also split up materials
- Basically change from bullets to boxes (when appropriate)
- Do check on hole dimension—risk mitigation for drilled holes
- Functional decomposition: check verb-noun format
- Expand out system architecture to include ways to mount probes on Biogenic refinery
- Final Concept - add Wireless/ Network to transfer data to computer