Engineering Update

Update as of 5-7-2018:

For our Automatic Aquarium Water Changer there is crucial analysis to be done. These types of analysis include:

Water flow through the tubing and pumps.
Economic
Environmental

The analysis that we intend to perform for item (1) is important because the flow rate delivered by the pump will change depending on how high the pump has to lift the water. Since everyone’s aquariums are at different heights and the pump actuation is timing based, knowing how the flow rate changes with height is an important factor to measure and model. Once the flow rates are modeled we will be able to effectively predict the time required to drain and fill the aquarium effectively.

Analysis for item (2) is important due to the fact that consumers are not going to want to pay a large sum of money to perform a task that is arduous but not difficult.

Analysis for item (3) is important since there will be waste water (from the aquarium) that will need to be disposed of. This disposal could have potential effects of soil quality (from potential increase in ammonia or nitrite levels). These effects can propagate to potentially disturb fauna and have an effect on the environment.

Below is a Volume Flow Rate vs. DC Voltage plot for the Anself pump to predict what will happen to the volume flow rate when the voltage changes. Both plots below can be used to quickly predict the time it takes to pump a volume of water at different voltages.

Below is a graph of filling time vs. pump voltage for a variety of volumes of water. This graph will help to determine the optimal range of flow rates that the pump should operate at in order to accurately predict the time it takes to do a water change for volumes from 2 to 20 gallons. The speed at which the system performs is not a major requirement, therefore, it would save power to run the pumps at 3V. From the graph below, it should only take about 15 minutes to do one cycle of a 20% (4 gallon) water change. The system could theoretically run on 2 AA batteries in series. This could provide back-up power for the system, and allow it to operate when far away from a wall outlet.

Below is the data from the experimental and theoretical analyses. The slope calculated from experimentation as related to the pump voltage is 0.0708 (gal*Volts)/min. This is lower than the rated slope of 0.0881 (gal*Volts)/min. However, this information is necessary to predict the timing of the pumps so that they can be programmed accordingly. Using this relationship, the automated water change can be timed accurately.

It is important to note that the minimum voltage the pump can operate on is 4 volts, otherwise the pump simply does not have enough power to move water. Also, we found through experimentation, that the supply voltage is attenuated when connected to the pump directly. For example, when a 12V supply voltage is across the pump, the actual reading at the pump is 9.71V. This is important in understanding the power the pumps will use, as well as the voltage the group should choose to run the pumps at for the final prototype. The group will likely choose a 12V supply voltage (9.71V pump voltage) for prototype.

Below is the updated time to transport gallons of water vs. pump voltage plot. This plot is updated using the new relationship between the volume flow rate and the pump voltage shown above. Comparing the theoretical plot above to this plot, it is shown that it will take a little more time to transport any volume of water, thus, our Arduino code will have to be updated accordingly. For example, from the graph above, to transport 4 gallons of water, it would take approximately 5 minutes when supplying the pump with 9 volts. The updated plot below shows this same operation would take approximately 6.3 minutes. Knowing this is crucial in programming the pumps to perform an accurate water change without the need for human intervention.

Below is a plot of Power vs. DC Voltage of the Anself pump. This is derived from the specified maximum power of 4.2 W. The relationship of the voltage and the power is:

P = (V*V)/R (1)

Where P is the power of the pump, V is the pump DC voltage, and R is the electrical resistance of the pump. Rearranging equation 5 to solve for the resistance at P = 4.2W, and V = 12V DC, the electrical resistance of the pump is found to be:

R = (V*V)/P = (12V*12V)/4.2W = 34.3Ω

The graph below represents the energy consumed by one Anself pump as a function of time at various supply voltages. It was generated using the equation:

W = Pt (2)

Where W is the energy consumed in joules, and t is the time of pump operation. The energy was converted to kilowatt-hours using the conversion below:

1 kJ=0.000277778 kWh

According to an article on npr.org the average cost per kilowatt-hour in the United States is:

1 kWh consumed ≅ $0.12

These results show that even when the pump is running at full power (4.2W) for 60 minutes it will cost much less than $0.12. This shows that power consumption is not a important aspect of the project in regards to cost for the consumer, and environmental concerns in regards to power consumption. Nonetheless, this can be used to estimate power consumption of the Anself pumps.

The table provided below shows an an average resistance that was experimentally found by measuring the current at each pump supply voltage. The average resistance found is approximately 29 Ω, as shown in the table. This is a little less then the expected electrical resistance (34.3Ω). However, the group plans to use a 12V supply voltage which corresponds to a 9.71V pump voltage. Therefore, the resistance using that configuration will be very close to the expected resistance. The resistance using 9.71V is shown as 32.4Ω. This means that the graphs above closely shows the power consumption of the Anself pump that is to be expected in real-life.

The plot was generated through experimentation by measuring the relationship that the pumping height has with the volume flow rate. It shows that as the height increases the volume flow rate decreases. The equation for this is specified as:

Q = -0.0059H + 0.8475 (3)

Where Q is the volume flow rate in gal/min, and H is the height that the pump has to transport volumes of water to in inches. Our findings indicate that the average height from the ground to the top edge of a typical aquarium is between 45 and 50 inches. We have used a setup height of 45 inches. Furthermore, the fresh-water pump will have to be programmed differently than the waste-water pump (the pump inside the aquarium) because they have to pump water to different heights. The waste-water pump only has to pump the water the height of the aquarium (13.5 inches), then gravity plus the pump power will bring it down to the waste water bucket. The waste water pump will pump the water out of the aquarium faster than the fresh-water pump can pump water into the aquarium.

Updated Information:

The time that the waste water and fresh water pumps is set to for the prototype was found from the Q vs. H plot above. For the waste water pump (the first leg of the water change) the height of 13.5 inches was input into equation 3. Using a target volume of 2 gallons, the result of this is that the waste water pump has to be activated for 2.6 minutes. Additionally, the height that the fresh water pump has to transport water is 45 inches, and was input into equation 3. Again using 2 gallons as the target volume, the fresh water pump has to be activated for 3.44 minutes.

The results from equation 3 were fairly close to the actual timing. The actual timing the pumps need to be activated for is 2.73 and 4 minutes for the waste and fresh water pumps, respectively. The reason for the differences are the dissimilar voltages of each pump in which the waste water pump has approximately 7V across it, and the fresh water pump has 8V across it. This could be alleviated through the usage of a relay. This would make it possible to supply each pump with the same voltage without the pumps leaching amperage off each other. This is likely another reason for the different and significantly smaller voltages than previously intended for the pumps.

Google Sites

Report abuse