Session: 17-01-01: Research Posters

Paper Number: 142000

142000 - Streamwise Velocity Decomposition in Riblet Surfaces - Comparison Between Different Direct Numerical Simulation Domain Sizes 

Riblet geometry has the potential to reduce turbulent drag by up to 10%, which is a promising approach to reducing fuel consumption.  In the numerical simulations field, direct numerical simulations (DNS) are used as a primary tool to investigate the drag reduction of riblet surfaces. However, in many research reports the DNS setup differs. This report will investigate how one specific parameter, the domain size, has the potential to affect the results for riblet surfaces. 

These riblets, which are grooves aligned in the streamwise direction, are very small because of their ability to introduce slip conditions near the surface while minimizing the amount of additional stresses. This has been determined through DNS using a minimal-span channel, where the turbulent cycle near the surface can be observed (MacDonald et al., J. Fluid Mech., vol. 816, 2017, pp. 5–42). Given that its domain size is used as a control variable, researchers make sure that the domain size stays consistent within each DNS run. In some studies, they keep the domain size the same based on the fluid area, while others choose to use a domain size based on the area above the riblet crests. 

To evaluate the results from these two domain parameters, each will have ten different surface models varying in different sizes. Half of the models will use symmetric triangles while the other will use blade-shaped riblets. They will also use the minimal-span channel concept, because of its performance efficiency. To verify that the domain size captures all relevant data, a smooth wall case will be created and compared to a reference data set from a fully developed channel flow.

Once results are obtained from the DNS, the mean streamwise velocity difference between the smooth wall case and the riblet models will be decomposed following Endrikat’s method (Endrikat, Minerva Access, 2020, pp. 1–112). This will show the streamwise velocity distribution caused by the slip velocity at the riblet’s crest and from the total stresses (which include turbulent and dispersive stresses). This will all be plotted in a graph where the horizontal axis is the viscous-scaled square root groove area. This is because it has been determined to give a more universal result, making it easier to compare with other data sets.

Preliminary results show that the mean streamwise velocity is optimized at a viscous-scaled square root groove area of 10. From the streamwise velocity decomposition, as riblet size increases so does its slip velocity. However, the velocity caused by the total stresses decreases at a faster rate. Thus, riblets can help with drag reduction up until the viscous-scaled square root groove area reaches 16.5. These results seem to be consistent with external data.

Lastly, when comparing the results from the different domain sizes, we see that the symmetric triangle riblets are accurate from one another. Further work will be done to give a more conclusive answer, nevertheless, these preliminary results give an insight that changing the domain size, in regards to the spanwise height, tends to not have a drastic effect on the results.

Presenting Author: Alexander Ramirez Garcia Portland State University - Maseeh College of Engineering and Computer Science

Presenting Author Biography: Alexander Ramirez Garcia is an MSME Graduate Student at Portland State University. Has interned for Micro Systems Engineering Inc. and LotusWorks. In 2024 he got his MSME, and in 2023, he built a compact and portable liquid fuel engine test stand. Has an interest in fluid mechanics, and would like to work in the aerospace industry.

Authors:

Alexander Ramirez Garcia Portland State University - Maseeh College of Engineering and Computer Science
Bianca Viggiano Polytechnique Montréal - Department of Mechanical Engineering
Xiaowei Zhu Portland State University - Maseeh College of Engineering and Computer Science