Preliminary Detailed Design
Team Vision for Preliminary Detailed Design PhaseIn this phase, our team intended to determine the initial feasibility of our design through both testing and simulation. To do this, we tested last year's project to get transient temperature data, and then used this information to run a simulation to determine the feasibility of using a convection loop to move the water in our design.
Feasibility: Prototyping, Analysis, SimulationAfter running initial tests on last year's project, we analyzed the temperature data and used the results to run a CFD simulation to determine the feasibility of using a passive convection loop to move water through our system.
Analysis: Temperature DataOur team completed a second burn trial to collect temperature data with thermocouples for the water and kiln over time. The details of the testing conditions are outlined in the timeline below:
The schematic and actual setup images below shows the placement of the thermocouples and modifications to last year's design to complete this test:
The final results from the test are summarized below:
|Starting water temperature||19.9 °C||67.8 °F|
|Increase||+10.1 °C||+18.2 °F|
|Ending water temperature||30.0 °C||86 °F|
|Amount of water heated|
|Time elapsed||1.3 hours|
|Ambient temperature||11.67 °C||53 °F|
In addition to temperature data, our team conducted a crude estimate of the volumetric flow rate by collecting data for volume and time:
With this parameter, other needed parameters for heat flux analysis were calculated and are summarized below:
Using this data, heat transfer rate and heat flux equations were applied to each temperature reading data to generate the following:
|Average Heat Transfer||5.2979||kW|
|Average Heat Flux||12.041||kW/m²|
The average heat flux was compared to the heat flux at each temperature in the graph below. The standard deviation is ±3.033kW/m² which is used to create the upper and lower bounds, which account for 75% of the data.
Due to flame dynamics being highly variable, our team will use this base heat flux measurement to create our simulation, which is discussed in the next section.
Additionally, the density change for the discrete water temperature values were analyzed in the graph below. This shows that the water density changes due to changing temperature are negligible, so a constant water density value can be assumed during analysis:
Fluent SimulationTo determine if a convection loop was a viable way to move water through our system, a CFD simulation was created to determine the temperature of the water as well as its velocity as it moves through a system. First, a CAD model was created using the drawings of last year's project, which can be seen below:
This model was then imported into ANSYS, which is a finite element software that is used to set up Fluent simulations, which is a computational fluid dynamics solver. After the initial geometry was imported into ANSYS, and a fluid domain was created to run the simulation on. This was done by filling in all the empty space in the previous CAD model, then suppressing all of the geometry used to create the fluid domain to make converging on a solution easier. A picture of the fluid domain can be seen below:
Next, the fluid domain was meshed using ANSYS, as this is a necessary step in using a finite element solver. The mesh can be seen below. Note that the mesh is much finer in the coils than it is in the reservoir, as it was expected that more elements would be necessary in the coil to properly capture the rapidly changing temperatures and velocities in the coil.
Next, a Fluent simulation was created using the mesh. Using the information we calculated from testing last years project, a boundary condition of 10 kw/m² was applied to the copper coils. Additionally, a heat loss of 100 W/m² was applied to the walls of the reservoir, to account for conductive heat losses through the side, and a convection boundary condition was applied to the top of the reservoir. The simulation being run is a transient one, and was run for 800 seconds with a time step size of 2 seconds. Temperature contours for the water, as well as velocity vectors, can be seen below:
Using this simulation, it can be seen that the water is flowing at a rate of approximately 10 cm/s, and has an average temperature in the reservoir of about 36°C. The simulation needs to be refined further, and will be after talking with professors who know more about CFD simulation. However, these results are approximately in line with what is expected, since the test we ran had a velocity of approximately 1 m/s, and an average temperature of 23°C. It makes sense that a lower flow rate would result in a higher temperature, since the slower moving water will have more time to absorb heat from the fire. Further refining of the simulation will likely result in adding more heat losses to the system, which means that the simulation will be able to run for a longer time without the temperatures getting to extreme values.
Drawings, Schematics, Flow Charts, Simulations
This is an updated schematic of the selected design. A raised platform is used to elevate the reservoir and the coils are then connected:
Potential failure modes associated with this construction include:
- Inability to substitute purchased parts with locally available parts. In an effort to reduce the cost of this project for economic needs in developing countries, our team will recommend materials that should follow a desired spec, but may be substituted. For example, a raised platform will have a desired height, but the source of this platform (e.g. wall, stacked materials) could be decided by the user.
- Convection loop not behaving as desired. This design relies on a passive convection loop to avoid external forced flows such as pumps, which would make the project more expensive.
- Unanticipated safety violations. While our team has tested last year's design and identified sources of danger involved with the operation, this current design may not behave the same way, leading to new design challenges. In particular, since the design relies on temperature changes resulting in a density gradient which drives the flow, the flow of water is going to take some time to start moving quickly. If it takes too long for the water to start flowing, then there is a chance that the near-stationary water could absorb enough heat to turn into steam, potentially scalding the operators.
Test PlansThe next step for testing is to physically test how effective the passive convection loop is. While the simulation done in a previous section shows that it is likely to work, the simulation itself is not a perfect representation of reality since it is likely missing several boundary conditions. Additionally, testing data done on an actual convection loop will allow us to further refine the model, which will in turn allow us to better make future design decisions. 3 different sets of temperature data will be taken: one at the "inlet", where water is flowing into the reservoir, one at the "outlet", where water is flowing out of the reservoir, and then one set of temperature data taken far away from either of these points to determine the bulk temperature of the reservoir. Additionally, if we can locate the proper testing equipment, the volumetric flow rate of the water through the coils will also be measured to see how effective the passive convection loop is at driving flow through the system.
Design and FlowchartsBelow is an updated flowchart of the system architecture. Teal boxes indicate a material source:
Additionally, an updated functional decomposition breaks down the main functions further for this specific design: