UWE Bristol Engineering showcase 2015
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
Thomas Gabriel<br />
BEng Mechanical <strong>Engineering</strong><br />
Project Supervisor<br />
Mike Ackerman<br />
To Assess and Compare Mechanical Presses and Hydraulic Presses<br />
Introduction<br />
Aquaponics is a relatively new farming method<br />
that utilises nutrient rich fish waste water to fuel<br />
the growth of crops. The nutrient fish water is<br />
pumped up from a fish tank to a header tank full<br />
of a porous medium that acts as an anchorage<br />
point for the roots of the crops on the surface<br />
above. The nutrients within the waste water are<br />
absorbed by the nitrifying bacteria on the base of<br />
the roots, resulting in the cleansing of the waste<br />
water and the uptake of nutrients into the crop<br />
roots.<br />
Currently the construction of a functioning autosiphon<br />
involves a trial and error method so that a<br />
30 minute siphon trigger time is achieved. This<br />
process can be time and labour intensive, reducing<br />
the accessibility of Aquaponics to new,<br />
inexperienced practitioners.<br />
As a result, it would be beneficial if a<br />
mathematical model could be constructed that<br />
could take system input parameters, such as<br />
geometry of the grow bed, volumetric inlet of fish<br />
wastewater and maximum root depth and<br />
produce the required geometry for a specific<br />
siphon design.<br />
The first step to the construction of such a model<br />
would be to determine the pressure drop of a<br />
liquid-gas fluid flow within a small siphon driven<br />
drainage system.<br />
Aims of invesitgation<br />
The aims of the investigation are as follows:<br />
-Construct a Mathematical Model that accurately<br />
predicts the total pressure drop experienced by<br />
the an auto-priming siphon system.<br />
-Design and Test a prototype of the chosen autopriming<br />
siphon.<br />
-Predict the Flow Regimes present throughout all<br />
stages of the siphon cycle and at what point they<br />
move from one regime to another.<br />
Test Setup<br />
Pressure (Pa)<br />
1800<br />
1600<br />
1400<br />
1200<br />
1000<br />
800<br />
600<br />
400<br />
200<br />
0<br />
Results<br />
The Pressure drop was found to be 1400 Pa, as<br />
shown in the graphs below.<br />
The Mathematical Model predcited a drop of 4800<br />
Pa. This was due to the unmodelled error of the<br />
two stream interactions as depicted in the figure<br />
below<br />
Schematic Highlighting the Influence of Annular Layer Thickness on Effective Gas<br />
Outlet Diameter<br />
Pressure Drop<br />
0 100 200 300 400 500 600<br />
Time (s)<br />
Pressure Drop<br />
Mark 1<br />
Mark 2<br />
Mark 3<br />
Mark 4<br />
Mark 5<br />
Mark 6<br />
Full Flow Begins<br />
Full Flow Terminates<br />
Project Conclusion<br />
As it currently stands, the mathematical model using<br />
the Brill and Beggs method does not accurately<br />
predict the test geometry. The largest source of the<br />
inaccuracy has been hypothesised to occur due to the<br />
unanticipated effects of the second inlet stream on<br />
the behaviour of the EZ-T siphon system as a whole.<br />
The addition of the second stream has been<br />
hypothesised to lower the pressure gradient between<br />
the inlet of pipe one and the inlet of pipe three,<br />
resulting in a reduced total pressure drop.<br />
This hypothesis was reached after an initial analysis<br />
of the mathematical model produced erroneous<br />
results due to an incorrect gas fraction reading<br />
generated by the balloon testing method described in<br />
section. The balloon method returned a value of<br />
volumetric flow rate for the gaseous phase of<br />
0.944x10E-03 m3/s. This equated to a gas content of<br />
42.7% within the flow mixture. The error in the gas<br />
fraction measurement was found after the<br />
construction of the transparent PVC prototype.<br />
During the documenting of the flow regimes within<br />
the transparent PVC model it was observed that there<br />
were no bubbles in the horizontal or vertical outlet<br />
pipes. As a result it was hypothesised that the gas<br />
fraction measurement had been erroneous. This led<br />
to running the mathematical model again with a nonexistent<br />
gas fraction. This lead to a pressure drop that<br />
was further away from the values tested than the<br />
initial result using the erroneous gas fraction.