Numerical simulation on improvement of a Savonius vertical axis water turbine performance to advancing blade side with a circular cylinder diameter variations

The use of fossil fuel will generate particulate gas in the atmosphere and forming the greenhouse effect. One of the ways to reducing greenhouse effect is used renewable energy as hydropower without generating particulate gas impacted in human life. The present study uses hydropower as the renewable energy by using a Savonius turbine. The main objective investigates the performance of Savonius water turbine numerically due to the installation of a circular cylinder beside of the advancing blade with circular cylinder diameter variations. The method used is numerically toward Savonius turbine disturbed a circular cylinder. The numerical simulation using two-dimensional (2D) analysis of Computational Fluid Dynamics (CFD) simulation by using ANSYS 17.0-Fluent and sliding Mesh technique used is to solve the incompressible Unsteady Reynolds Averaged Navier-Stokes (URANS) equations. The turbulence model uses Realizable k-epsilon (RKE) and transport equation uses the finite volume discretization method with the second-order upwind scheme and the SIMPLE algorithm. Firstly, the numerical model has been validated by the published experimental data toward the torque coefficient by using air fluid at Reynolds of 4.32.105. Then the fluid is changed the water fluid at the same Reynolds. The circular cylinder diameter relative to the turbine diameter (ds/D) is varied of 0.1, 0.3, 0.5, 0.7 and 0.9 at X/D of 0.5 and Y/D of 0.7 kept constant with TSR of 0.5, 0.7, 0.9, 1.1 and 1.3. The numerical simulation uses the transient and two dimensional (2D) simulations. The results show that the increase of disturbance diameter (ds/D) will improve the performance of the conventional Savonius turbine and the maximum power coefficient increase about 18.04% at ds/D of 0.7 with TSR of 0.7.


Introduction
With REmap, more than half of all renewable energy use in Indonesia in 2030 would be in the form of bioenergy used for process heat in industry or as liquid biofuels in transportation. Solar applications (including Photo Voltaic (PV) and thermal) account for 15% of renewable energy use in all sectors in Indonesia as envisaged by REmap, followed by hydropower (14%) and geothermal power (9%) [1]. By using fossil fuels in combustion can produce carbon dioxide, methane and nitrogen oxides which contribute to greenhouse gas emissions improving global climate change. This present study uses renewable energy to reduce the fossil fuels and greenhouse gas emissions by improving the performance of the Savonius turbine numerically. Generally two types of a turbine; such as vertical axis turbines (VATs) and horizontal axis turbines (HATs). It based on shaft alignment between their axis of rotation and fluid direction. VATs used to generate small scale power because the turbine performance is not depended on fluid flow direction [2]. Type of VATs generates torque by combining between drag effects and side forces. The energy of hydro from the river, sea current, waves is the best renewable energy sources and very predictable compared to wind energy or the other. Application of vertical axis turbine is generally Savonius turbine, helical turbine and Darrieus turbine [3]. The present study focuses on the hydropower by improving the performance of the Savonius vertical axis turbine and the type of water turbine has the lowest performance compared to others types of the water turbine. That is why various studies have been carried out to improve the performance of the Savonius turbine. Among others, Sheldahl et al. [4] have varied the number of the bucket and the gap spacing between the buckets. Patel et al. [5] conducted a study on the Savonius turbine numerically by varying the overlap ratio of 0, 0.1 and 0.2. They have observed it for varied wind speed the maximum torque has obtained at the overlap ratio of 0.2.
Freitas [6] state the numerical study was carried out and developed to improve the performance of the turbine. The study of numerical simulation has been tested to avoid uncertainty numerically. It produces some numerical simulation policy and some policy was essential to improving the results of numerical as discretization second-order upwind and comparison numerical toward experimental data.
The result of the numerical simulation is carried out to compare 2-D and 3-D simulation, where has it shown a good approaching experiment result. The results of 2-D simulation have power coefficient which is more approaching experiment than 3-D simulation [7].
The Spalart-Allmaras (SA) [8] turbulence viscosity model is a one-equation turbulence model where near wall gradients of the transported variable are much smaller than the turbulent kinetic energy equation based k- models and the standard k- model [9] is more suitable when flow is fully turbulent and has given better results than SA model for turbine analysis [10]. Previous studies for Savonius turbine has shown that two dimensional simulations give acceptable results [2] [11].
This present study is numerically expected to reduce greenhouse gas using Savonius water turbine. Simulation analyzes the unsteady flow around the Savonius turbine by adding circular cylinder placed to advancing blade side with circular cylinder diameter (ds/D) variations.

Equations and mathematics
The Savonius turbine rotates 360 degrees starting from the first position and returns on the same position relative to the center axis for achieving one rotation. The time step size (TSS) represents the increment angle for each the rotation of step and the number of the time step (NTS) represents the total of turbine rotation.
The equation of mathematics of the number of the time step (NTS) and the time step size (TSS) can be written as follows: 15915 ω x Number of time step (2) Where N is the number of rotations,  is the increment angle or time step rotation degree,  is the rotation speed of turbine (rad/s) and 0.15915 is a constant (conversion from rad/s to rot/s unit). The equation of TSR (tip speed ratio), the coefficient of torque, the coefficient of power can be written as follows: Where  is the angular velocity, U is the free stream velocity, As is the rotor swept area, As = D.H and D is the Savonius turbine diameter, P is the power of the turbine, Cp is the coefficient of performance that a non-dimensional parameter used to evaluate the turbine power and Cm is the coefficient of moment from simulation results.

Numerical method
Firstly, the numerical model has been validated by the published experimental data Sheldahl et al. [4] toward the torque coefficient by using air fluid at Reynolds of 4.32.  This study uses the unsteady incompressible Reynolds-Averaged Navier-Stokes (RANS) equation based on the cell-centered finite volume method and then has implemented the rotation by using the sliding mesh technique to rotate the space of the turbine area. This main study, the structured grid has been employed for all the grid system of the rotor, and the computational domain reaches to 10D in the inlet direction, 10D in the outlet direction and 6D in the vertical direction towards Savonius turbine. The calculation was made based on the two-dimensional unsteady flow assumption for its relative simplicity. There are three different meshes defined as a fixed domain, wake domain and rotating domain as shown in Figure 2. The grid of computation was generated using the MESH tool in ANSYS 17.0 with twodimensional, quadrilateral elements are more desirable and have high accuracy in numerical solutions shown in figure 2 [15]. The surface was made first prism layer with setting the y+ value for the first elements from the wall between 30 and 100, depending on the rotation velocity of the rotor and the position of the elements on the blade shown in figure 2(c).  The validation of this numerical simulation was carried out by comparing with the results experimental published by Sheldahl et al. [4]. Comparison of the torque coefficient (Cm) was as the function of the tip speed ratio (TSR) between the results of this numerical simulation to the experimental results of Sheldahl et al. [4] as shown in Figure 4. A range of tip speed ratio of 0.5 to 1.3, Figure 4 shows that the present study is in very good agreement with the experimental results reported by Sheldahl et al. [4]. As a conclusion, so the mesh characteristic, which is operated in the simulation above is considered to be valid and will be used to simulate the real problem proposed in this present study.

Results and discussion
In this present study, the air-fluid would be changed to the water fluid after the grid independence is reached. The numerical study was carried out at free stream velocity kept constant in about 0.22 m/s with the rotor diameter of Savonius turbine (D) of 0.4 m. Numerical simulation has achieved grid independence, so the next numerical was carried out by installing a circular cylinder to the advancing blade side. The diameter of the circular cylinder relative to the diameter of the turbine rotor (ds/D) was varied 0.1, 0.3, 0.5, 0.7 and 0.9. The vertical position relative to turbine rotor (Y/D) was kept constant at 0.7 and the horizontal position relative to turbine rotor (X/D) was also kept constant at 0.5 and the simulation was carried out for tip speed ratio (TSR) from 0.5 to 1.3. Graph of torque and power coefficient was shown in figure 5 and 6. It is clear that the numerical simulation by adding a circular cylinder as passive control was installed to advancing blade side influenced the performance of Savonius turbine. The numerical simulation varied circular cylinder diameter relative to Savonius turbine (ds/D) of 0.1, 0.3, 0.5, 0.7 and 0.9. The results showed that the performance of the turbine increased by the change ds/D from 0.1 to 0.9 compared to the Savonius turbine. The maximum power increased at ds/D of 0.7 about 18.04% at TSR of 0.7. Meanwhile at ds/D of 0.9, shown that installation of circular cylinder decreased the performance of the turbine because the formation of streamline on the upper side of circular cylinder caused the flow direction blocking free streamline, however, the performance of the turbine reduces. The peak power coefficients produced by varying circular cylinder relative to Savonius turbine (ds/D) are compared with the highest performing on conventional Savonius shown in Table 1. Table 1 shows the benefit of the effect of circular diameter variations with a peak power coefficient of 0.2709 obtained at a tip speed ratio of 0.7, which is 18.04% higher than a conventional Savonius turbine.

Conclusion
A circular cylinder as passive control contributed great effect to the Savonius turbine performance. From above discussion result, it shows that the increasing of ds/D will increase the torque and power coefficient, which maximum performance occurred at ds/D of 0.7 and power coefficient increase in about 18,04% at a tip speed ratio (TSR) of 0.7. Using renewable energy and increasing the turbine performance will reduce dependence on fossil fuels so that they will reduce the greenhouse gases as a cause of climate change.