Michael Morgan and Professor Dan Maynes, Mechanical Engineering
Introduction
The objective of this research was to explore the drag reduction capabilities of a surface with trapezoidal riblets coated with a superhydrophobic coating. Drag reduction is very desirable as even small reductions can save enormous amounts of energy.
Riblets are patterned microfeatures on a surface (illustrated in Figure 1). When fluid flows parallel to them (into the page when referring to Figure 1), riblets reduce drag by dampening the turbulence perpendicular to the direction of flow. In the last 40 years, this effect has been studied both computationally and experimentally. Results of these studies show 4-10% reduction in drag. Direct numerical simulations performed by both Goldstein and Chu/Karniadakis indicated a drag reduction of 4-6%. On the experimental side, Walsh measured the drag reduction with a drag balance and Suzuki/Kasagi performed a particle tracking velocimetry experiment. Trapezoidal riblets were used in both of these experiments and the resulting drag reduction was in the range of 8-10\%. In other studies, experimental and computational, other riblet shapes showed similarly varying results.
Data from my research indicates that 1) trapezoidal riblets alone can reduce drag by up to 10% and 2) that trapezoidal riblets with a superhydrophobic coating exhibit drag reductions of up to 35%. It should also be noted that these surfaces were about twice as efficient in the fabrication process as compared to blade riblet surfaces which showed drag reductions very similar to those in my research.
Methodology
The first step was to fabricate the surfaces using photolithographic techniques. The data collection was then facilitated by the existence of an apparatus developed by a former PhD student in the lab. The entire setup and its components are illustrated in Figure 2, but I will only briefly describe the test channel. In the channel, there were placed eight smooth surfaces – four on top, four on bottom – followed by eight test surfaces placed likewise. The channel was equipped with three pressure transducers- two of them each had seven equally spaced pressure taps which were positioned to measure the pressure drop over the control surfaces and test surfaces. The third measured static pressure in the channel. A thermometer is also included in the setup. With this equipment, data was collected which was used to calculate drag reduction.
In my experiments, I explored Reynolds numbers (Re=pvL/u) in the range of 5000 to 15000. Prior to each test, measurements of the channel are taken to determine uncertainty. These, along with data collected by all instruments during the test, were recorded by a LabView program and were then loaded into MatLab which was used to process the data and compute the drag reduction (DR).
Drag reduction is defined as DR=(fcontrol-fsurface)/fcontrol, where f is the Darcy friction factor which is an estimate of the friction at the wall slowing the flow, which in these experiments is calculated by f=2Dh(-dp/dx)/pV2. The pressure gradient necessary to drive the flow (-dp/dx) is calculated by measuring the differential pressure between the first pressure tap and each subsequent pressure tap (refer to Figure 2).
Results and Discussion
A quick comparison of Figures 3 and 4 will indicate a couple of things. The first thing to note is that (besides the points from 174) the control surface data is almost identical in the two plots. This gives confidence that the data is indeed comparable. The second observation is the wide disparity in the test surface points which confirms that the superhydrophobic coating did, in fact, greatly help to reduce drag.
The data collected in the four tests of the first stage were very similar to each other; this can be seen in Figure 3. The drag reduction, which is seen in Figure 5’s plot, ranges from 6-10%, well within the range of previous experiments as mentioned above. According to this data, there seems to be an optimal Re (about 1000) for drag reduction with these surfaces. In the second stage of my research, I ran three successful tests. Again, the drag reduction results from these tests are fairly comparable; see Figure 4. As expected, the reduction in drag achieved with these surfaces is much higher than that seen in the previous stage of my research. These surfaces show a DR between 25-35%. This is very significant and is even somewhat higher than results seen with other kinds of riblet/superhydrophobic surfaces which are no greater than 30 percent. Another observation is the in this case, instead of a peak in the plots, there is a valley, a minimum DR point around a Re of about 12000. Further exploration of this trend would be required for an understanding of its cause.
Conclusion
Substantial drag reduction is possible with both trapezoidal riblet surfaces alone and superhydrophobic trapezoidal riblet surfaces. This is noteworthy because, not only can it lead to energy-saving technology, but the surfaces were fairly quick to manufacture. Though these wafers would not likely be used in real-world applications, the fabrication method could save researchers a great deal of time, allowing them to focus their efforts on testing, refining, and exploring applications for these types of surfaces. Future work to confirm the data presented here and to explore the feasibility of manufacturing the surfaces on a large scale would be beneficial.
References
- J. F. Prince, “The Influence of Superhydrophobicity on Laminar Jet Impingement and Turbulent Flow in a Channel with Walls Exhibiting Riblets,” Ph.D. dissertation, Dept. of M.E., BYU, 2013.
- F. Beibei, “Characteristics of flow over thin triangular riblet surface,” in Proceedings of the 9th WSEAS International conference on fluid mechanics (FLUIDS ’12), Cambridge, Massachusetts, 2012.
- R. Garcia-Mayoral, “Scaling of turbulent structures in riblet channels up to Reτ ≈ 550,” Physics of Fluids, vol. 24, no. 10, 2012.
- D. Bechert, “Experiments on drag-reducing surfaces and their optimization with an adjustable geometry,” J. Fluid Mech., vol. 338, pp. 59-87, 1997.
- C. Barbier, E. Jenner and B. D’Urso, “Large Drag Reduction over Superhydrophobic Riblets,” ARXIV, 2014.