Improving turbine’s efficiency by the clocking technique

Dieter Bohn, Jing Ren, Institute of Steam and Gas Turbine, RWTH Aachen University Germany

The so-called "clocking technique" has recently received a great deal of attention in the multistage turbomachinery world. The technique aims to improve performance by modifying the relative circumferential positions of the blades in consecutive stages (whether fixed or rotating). Clocking works by modifying the interaction of the wakes and unsteady flow-fields generated by the upstream row on downstream components. In order to gain maximum benefit from the clocking technique, it is important to gain insight into the sometimes subtle changes in flow features that drive large increases in performance.

As is well known, real flow in turbines is incredibly complex, being viscous, unsteady and three-dimensional. Using STAR-CD's fast transient solver and sliding mesh approach, we were able to investigate the time-dependent flow field including the rotor-stator interaction inside turbines.


Fig.1: Stage configuration at midspan	Fig.2: Computational grid

The investigated turbine cascade comprised two stages of an axial turbine. Fig. 1 shows the airfoil cascades at midspan, in which stators and rotors are located on their trigger positions. The second stator was clocked in steps of 2.5o during one stator pitch to receive three stator-stator clocking positions. They are indicated in Fig. 1 as clocking positions S1, S2, and S3. As the vane/blade number ratio was 1/1, the computational domain comprised the passage between two adjacent stator vanes and two adjacent rotor blades. A multiblock structured mesh was generated for each clocking configuration. Fig. 2 shows the computational grid as well as the enlarged leading and trailing edges at the clocking position S1.

User programming (via the subroutine POSDAT) was used to estimate the total-to-total efficiency of the unsteady process. While the stator-rotor position in the steady solution is spatially repeatable, the unsteady solution is time periodic. Hence, two periods of the total-to-total relative efficiency distributions based on the two approaches are presented in Fig. 3 to show the tendency more clearly. An enlarged graph with a high resolution is located at the lower right corner of Fig. 3 which shows the results in more detail. It is clear that the efficiency curve at the clocking position S1 has the highest peak while the curve at the clocking position S2 has the lowest peak.


Fig.3: Averaged relative efficiency distributions	Fig.4: Relative efficiency distributions

The averaged total-to-total relative efficiency is calculated as the averaged value of the relative efficiency in one blade passing period. As illustrated in Fig. 4, clocking position S3 has the highest averaged relative efficiency while the clocking position S2 has the lowest. The maximum averaged relative efficiency variation is 0.03% in one blade passing period.

Entropy distributions were calculated and visualized in pro-STAR, from which the wake trajectories in the flow passage can be implied. The entropy distributions at the clocking positions S2 and S3 are shown in Fig. 5 and Fig. 6, at the maximum and minimum turbine efficiency respectively. In each figure, the resulted entropy distributions is illustrated at six times during one blade passing period.


Fig.5: Entropy distribution at the clocking position S2	Fig.6: Entropy distribution at the clocking position S3

It is clear that the wakes' trajectories are unsteady, but periodically repeatable in the turbine cascade. Wakes start at the blade tail and extend outward along the direction of the trailing edge. They bow, shear and stretch on the midspan, while being convected through the midpassage. Wakes sweep one pitch of the next stator or rotor with different directions and magnitudes while the rotors move. Comparing the entropy graphs at the same time in the Fig. 5 and Fig. 6, it is obvious that the clocking causes the different entropy distributions. It implies different clocking positions will reach different efficiency.

Based on these CFD studies with STAR-CD, we were able to show how to improve the turbine efficiency by the clocking technique. Meanwhile, the advanced computational technologies developed by CD-adapco also allow for investigation of the mechanism of the clocking effects on the performance and the durability of turbine. ACKNOWLEDGEMENTS The investigations were performed as a part of the joint research program "500 MW auf einer Welle (AG Turbo)". The work was supported by the Bundesministerium für Wirtschaft (BMWI) under file number 0327061E. The authors gratefully acknowledge AG Turbo and ALSTOM Power, especially Dr. Michael Sell, for their support and permission to publish this paper.