Effect of Dimple on Aerodynamic Behaviour of Airfoil

-In order to boost the efficiency of an airfoil, surface of the airfoil is altered. A two dimensional airfoil was analysed with and without dimples on the upper surface using CFD software. NACA0012 non cambered airfoil with and without dimples were used for analysis with k-ɛ turbulent model. Both were compared keeping in mind the coefficients of lift and drag. Dimples were located at four different positions and compared mutually with smooth airfoil. The velocity of flow was keeping constant for different angles of attack. In CFD analysis results were fluctuated with size of grid so as to get rid of the fluctuating, a grid independency test was done before final analysis. During grid independence test numbers of nodes were increased until constant results come.

) Separation bubble and lift coefficient fluctuation with time was observed during study. Laminar separation bubble become unstable and developed primary and secondary vortex. Secondary vortex was much stronger than primary vortex. Analysis was done from 0o to 10o angles of attack. As soon as the angle of attack increased, the fluctuation also increased. Laminar separation bubble started moving forwarded for increased angle of attacks and started to reattach to surface of airfoil, hence lift coefficient increased suddenly.
(C.K. Chear & Dol, 2015) Dimples delay the flow separation for bluff bodies. Author simulated car model with different ratios of dimple using k-ɛ turbulent model. Ratio of dimples was taken as depth to diameter. (Mustak & Harun, 2017) At zero degree angle of attack, dimples on airfoil do not shows changes in drag compared to smooth airfoil. But at high angles of attack it behaves like bluff body. It leads to delay in separation and wake formation. Also it increases the angle of stall. In this work NACA4415 airfoil was used and drawing was first made in solid works. Hexagonal outward dimpled profile was compared with smooth profile of airfoil. Physical model was prepared with wood and analysed in wind tunnel. Hexagonal surface delays starting of flow separation by 4 degree angle of attack. In case of smooth surface it starts at 12o and for outward dimpled it happens at 16 o angle of attack. Velocity of the air was taken as 43m/s. III. TURBULENT MODEL Turbulent models are used because of limitations of Navier stroke equations. There are many turbulent models used in CFD analysis. Generally k-ɛ and k-ω models are used in fluid flow. Both models are used for streamlined and bluff bodies. Kinetic energy of turbulent fluid flow is solved by k-ɛ turbulent model. This model is less complicated compared to other. Time of computation is also less. This model can be used in low memory computers. A. k-ɛ turbulent models equation (Muralidhar & Sundrarajan, 2008) The Grid distribution scheme has many limitations. Global error cannot be controlled. But we can control the local errors. For controlling this author increased the no of nodes until he got constant result.

IV. GRID INDEPENDENCE TEST
As we increase the number of nodes, result varies with respect to it. But there comes a stage where the results become fixed. This fixed result shows that this is our required number of nodes on which we have to do work as shown in fig.1. In this work 102180 numbers of nodes have been used.

V. COMPUTATION METHOD
NACA0012 airfoil smooth profile and dimpled airfoil were used to study the aerodynamic behaviour of the airfoils. The shapes of the airfoil models is shown in fig.2 and the farfiled and meshing is shown in fig.3, is used for computation in CFD software. Diameter of dimple was taken as 0.02 % of chord.
(airfoiltools.com, 2016) Practical data were taken from this reference. These data were validated to check the accuracy of the work. (Confluence, 2015) Coordinates of airfoil NACA0012 was downloaded from this source.
Smooth airfoil's computed results were compared with practical data. This ensured us that we have followed right way for calculation. After that dimples on the airfoil at different location were created. Dimples location affects the results. In this work five dimpled airfoil were used, one is smooth and remaining each have dimples at 10%, 25%, 50% and 75% of chord length. Results of smooth airfoil were compared to outcomes of these five dimpled airfoil. Flow of air was taken 7.3 m/s and density was 1.225kg/m3.   A of 9-13. At 10o angle of attack fluid starts separating and generates wakes. This leads to pressure drag. As we reaches 16o angle of attack separation reaches maximum value, after that lift starts decreasing.   14-18 (A), this represents that pressure on both surface is approximately similar. At higher angle the difference of pressure near the leading edge is wider, this represents that lift starts from leading edge. It was also proved in contours of pressure diagram.  fig 19 (B) of the smooth airfoil. Flow separation delayed in the airfoil which is dimpled at 75% of chord, can be seen in fig  23(B). So it proves that dimpled airfoil performs better than smooth airfoil. was seen that dimple at 10% of chord showed worst result than smooth airfoil. Coefficient of lift was increased and drag was decreased in dimpled airfoil at 75% of the chord .Dimples at 25% and 50% of the chord length also did not performed well.    Fig 24 shows that coefficient of lift has been increased by 7% for airfoil having dimple at 75% of chord length, compared to smooth airfoil. In the same manner it was noticed for coefficient of drag as shown in fig 25. Coefficient of drag has been reduced by 3% for the same airfoil. The location of the dimple on the airfoil plays an important role. In this work we noticed that dimple at 75% of the chord length is the best location for the dimples.
ACKNOWLEDGMENT I am using this opportunity to express my gratitude to God, my Parents and everyone who supported me throughout the course of this Research Paper. I am thankful for their aspiring guidance, invaluably constructive criticism and friendly advice during the work. I am sincerely grateful to them for sharing their truthful and illuminating views on a number of issues related to the project.
I express my warm thanks to Dr. M.P.Singh, Jagannath University and Dr. Tej Singh Chouhan for their support and guidance in working in this work.
We would also like to show our gratitude to the Dr. Vivek Sharma, Jagannath University for sharing their pearls of wisdom with us during the course of this research.