Introduction

Power performance and characteristics of the CRTT are analyzed using CFD commercial code, the ANSYS 14. Meshing is done using ANSYS MESHING. Mesh dependency study was carried out and the independency was achieved at mesh size of 6.1 million nodes (about 23 million tetra elements). The full rotor simulation requires high computational effort but can bring about more accurate results. Fig. 3 illustrates mesh formation for the rotor which shows very dense and fine development of elements. Yplus value at most parts of the rotor region are below 10.

Fig. 3 
Full rotor mesh formation

Fig. 4 
Computational domain

The computational domain is shown in Fig. 4. The domain has cylindrical shape and is sized 4.8D × 25.6D (Radius × Length). This is suitable for full rotor simulation and is large enough to capture the wake characteristics with high accuracy. The far-field region is gridded using multi-zone hexa mesh, while the turbine region is high density tetra mesh with well-developed boundary layer as described above. Inlet boundary is normal speed, outlet boundary is static pressure, and opening boundary is set to “Opening Pressure and Direction”. The blades and hub are all no-slip wall boundaries. The mesh connections between both rotors (front and rear) and between each rotor with the flow field are general connection interface boundaries and are set to frozen rotor for frame change/mixing model. The fluid is sea water at 25^oC, turbulence model is shear stress transport (SST) with turbulence intensity is set to low (1%) for the whole flow-field. Simulation is done in steady state by ANSYS CFX Solver and is carried out for a wide range of TSR as viewed in Table 2. Solver iteration is set to 10000, time scale is kept as low as 0.025s for convergence.

Fig. 5 
Streamlines distribution on rotor blades (TSR 1)

The visualization of flow patterns at different TSRs are in turn shown in Fig. 5 to Fig. 10. These figures present the surface streamlines of relative water velocity (velocity in rotating frame) which are distributed over the blades and hub. In all cases, the view of blade’s pressure side is oriented at the left of the figure and the one on the right views blade’s suction side.

At TSR 1 (highest tidal inflow velocity), it is clearly seen that there is flow detachment at the pressure side. On the other hand, vortex is formed at suction side (mainly at root part) and there is the existent of radial flow at all blades. Usually, fluid within turbine’s wake region should flow following the blade’s rotational movement. In this case, the streamlines on suction surface tends to move in radial direction from the root of blades toward the tips. Radial flow has no effect in transferring water’s kinetic energy into rotor’s torque due to the escape of water from the wake’s core region. The main reason for this is extremely high inflow tidal speed, the inflow speed up to around 6m/s is considered very high in terms of physical understanding. Thus, under such high speed flow, water is quickly detached from blades and this phenomenon may not beneficial to the absorption and exchange of power.

Fig. 6 
Streamlines distribution on rotor blades (TSR 2)

Fig. 7 
Streamlines distribution on rotor blades (TSR 3)

Fig. 8 
Streamlines distribution on rotor blades (TSR 4)

Fig. 9 
Streamlines distribution on rotor blades (TSR 5)

Fig. 10 
Streamlines distribution on rotor blades (TSR 6)

At TSR 2 (Fig. 6), the streamlines already became well developed. The distribution of streamlines at blade’s pressure and suction sides become more uniform without any abnormal or strange effect. At higher TSR (from 3 to 5), the patterns look similar and there is also no significant difference between front and rear rotor blades. At operating TSR (theoretically TSR 3, and is based on single rotor design), there is a tiny sign of radial flow at trailing edge of rear blade’s suction side. The cause of this is not clear and possibly due to computational error or minor meshing problem. Moreover, the finest state of streamlines seem to be found at TSR 4. With this, the CRTT's optimum operating point can be shifted. By examining the energy transfer efficiency, the effect can be figured out but in general, the interaction between rotors does have a certain changes to overall performance of the CRTT.

At the highest TSR (lowest tidal inflow velocity), the streamlines are no longer maintain the smoothness and stability. Although there is no formation of vortex or complicated flow, there is the appearance of radial flow that is similar to the pattern at TSR 1. However, the strength and intensity of the effect are weaker and mainly noticeable at rear rotor’s suction side.

Power performance of S814 CRTT is evaluated by power coefficient C_P which is calculated and expressed in Fig. 11. The figure expresses C_P as a function of TSR and shows individual performance of each rotor as well as the combined performance of the CRTT system.

Based on previous studies^(2),(3), a single rotor turbine can operate efficiently, and extra-added rotor can only contribute a small increase of efficiency to the device. Moreover, there may be effects that are not advantageous for each rotor in counter-rotating assembly, i.e. local vortex or back flow. Such effects may cause deficiency in both rotors’ operation.

Besides, because of the wake’s physical characteristics, escaped water from front rotor contacts the rear rotor in different angles of attack and influences the aerodynamic performance of the rear rotor blades. Thus, the individual C_P of both front and rear rotors are under 0.3 throughout the range which are lower than their true capability in separated installation operation. However, the overall C_P of CRTT is seen to be high all in all. From the finding at streamline characteristics, TSR 4 is the most efficient point which has the maximum C_P very close to 0.5. The performance is seen to be good since there is only a slightly decrease of C_P at the vicinity (TSR3 and TSR 5). Even at the highest TSR, power efficiency of CRTT is in an acceptable value. At TSR 1, the turbine’s performance is poor as C_P is as low as 0.15. This is reasonable to what has been seen previously at the flow patterns.

Fig. 11 
Power coefficient as a function of TSR

Fig. 11 also presents the performance of S814 tidal turbine operating as single rotor device with the same design as front blade. Compared with the CRTT, single rotor turbine has the maximum C_P at TSR 3 and this point is significant higher than the rest. When it comes to CRTT system, the front rotor's performance drops in about half of its capability seen in single operation. The installation of counter-rotated rotor at given distance in fact does influence the operation of upstream rotor and thus decreases its power efficiency. These results are reasonable to characteristics of flow patterns discussed above, where high efficiency is found at cases with well-developed streamlines. In same sizing design, the rear rotor’s performance is not so highlighted and mainly beneficial at low TSRs. Compared to recent works of other researchers as Sutikno et al.⁽⁷⁾, Grant et al.⁽⁸⁾, Kumar et al.⁽⁹⁾ and Lee et al.⁽¹⁰⁾, S814 CRTT shows similar or even more promising performance. Figure 12 shows several recent researches’ results in comparison with S814 CRTT. Lee et al.’s counter-rotating turbine is slightly more efficient than S814 CRTT, but their work is done for wind case. The other studies have one common, that is the operating range is mostly at high TSR region where either tidal current is low (low power output) or rotational speed is high (disadvantage for durability and structural concerns).

Fig. 12 
Performance comparison among studies

The configuration of S814 CRTT in this study results in bigger rear rotor but shows more promising power performance and more importantly the trend is maintained throughout a wide range of TSR and both rotors have stable equal performance with less unbeneficial interaction. With the high performance at various TSRs, we can use this device widely and efficiently at varied tidal current conditions.

In actual, other than power efficiency there are many concerns about CRTT needed to be fully discovered. Nevertheless, the S814 CRTT shows high potential for further development and improvement.