Introduction

To achieve high thermal efficiency, and low cost of a gas turbine engine, turbine inlet gas temperature should be increased. However, the penalty is high thermal load, which exceeds the durability of the turbine components. Therefore, turbine blade is equipped with improved cooling techniques such as film cooling and internal cooling. Internal cooling is achieved by low enthalpy air circulating in multi-pass flow channels inside the blade structure. To increase the effectiveness of the internal cooling, the internal surfaces usually are roughened by angled ribs to trip the boundary layer and increase the turbulence, which enhance the heat transfer. As the turbine blade rotates, Coriolis and buoyancy forces appear and cause different heat transfer behavior from the leading and trailing surfaces. Coriolis force produces secondary flow in planes perpendicular to the main flow direction, which encourages the migration of core region flow toward the trailing surface in the first pass and leading surface in the second pass.

Over the bast few decays numerous studies have been made experimentally on the flow field and heat transfer in the internal coolant passage of gas turbine rotor blade. Metzger et al.1 studied forced convection in two-pass smooth rectangular channels by varying the divider location and the gap at the 180° turn. Fan et al.2 extended the Metzger et al.1 work by varying the channel width and the conclusion was, increasing channel aspect ratio results in smaller azimuthal heat transfer variations and increased overall channel heat transfer. Han and Park3 performed experimental studies on heat transfer characteristics in a non-rotating rib roughened rectangular channel. Han et al.4 studied the effect of the rib angle orientation on heat transfer distributions and pressure drop in a non-rotating square channel with two opposite in-line ribbed walls. They found that the 60° (or 45°) V-shaped rib performs better than the 60° (or 45°) parallel rib and, subsequently, better than the 60° (or 45°) crossed rib and the 90° rib. The V-shaped rib produced the highest heat transfer augmentation, while the crossed rib had the lowest heat transfer enhancement. Ekkad and Han5 performed a detailed study on heat transfer characteristics in a non-rotating square channels using liquid crystals techniques. The results show that the parallel, 60° V, and 60° V inverted ribbed channels produce similar levels of heat transfer enhancement in the first pass, while the 60° inverted V ribbed channel produced higher enhancement in the second pass. Wagner et al.6' 7 conducted the detailed experimental study to determine the effects of rotation (buoyancy and Coriolis forces) on the local heat transfer of a multi-pass square channel with smooth walls. They concluded that in the first pass of the coolant passage rotation created a thinner boundary layer on the trailing surface and a thicker boundary layer on the leading surface but in the second pass the performance was different and opposite to the first pass. The leading surface Nusselt number ratios in the second pass were higher than the trailing surface Nusselt number ratios because of the reversal of the Coriolis force direction. Johnson et al.8 performed a systematic investigation of the effects of buoyancy and Coriolis forces on heat transfer coefficients distribution of four-pass square channels with trips angled to the flow (45° ribs). Han et al.9 investigated uneven wall temperature effect on local heat transfer in a rotating two-pass square channel with smooth walls. Zhang et al.10 analyzed the heating condition effects in a duct with angled rib turbulators with rotation. They suggested that an uneven wall temperature had a significant impact on the local heat transfer coefficients. Parsons et al.11 presented wall-heating effect on local heat transfer in a rotating two-pass square channel with orthogonal ribs. Johnson et al.12 and Parsons et al.13 studied the effects of channel orientation and wall heating condition on local heat transfer coefficient in a rotating two-pass square channel with ribbed walls. They found that the effects of the Coriolis force and cross-stream flow were reduced as the channel orientation changed from normal b =90° to an angled orientation b=135°. Dutta and Han14 also investigated the local heat transfer coefficients in rotating smooth and ribbed two-pass square channels with three channel orientations. Dutta et al.15 presented experimental heat transfer results for turbulent flows through a rotating two-pass rib-roughened triangular channel, with two channel orientations with respect to the direction of rotation. Taslim et al.16' 17 studied the heat transfer characteristics in rib-roughened square and rectangular orthogonal rotating channels. They used a liquid crystal technique to study the effect of rotation on heat transfer distributions on the walls. First part, leading and the trailing walls of the test channel were roughened with staggered transverse ribs, while second part was, the opposite walls were rib-roughened at 45° with respect to the main flow, in a criss-cross arrangement. They found that rotational effects were more pronounced in rib-roughened channels with a higher channel aspect ratio and a lower rib blockage ratio. They investigated heat transfer effects only in an orthogonally rotating single pass channel. However, they did not consider a two-pass channel and the effect of channel orientation. Prabhu and Vedula18 investigated the pressure drop distribution in a rotating rectangular channel with transverse ribs on one wall. They conducted experiments for a rotation number up to 0.21 and rib pitch-to-height ratios of 3, 5, 7.5, and 10, and a rib height-to-hydraulic diameter ratio of 0.15. They found that a rib array with a pitch-to-height ratio of 5 caused the largest pressure drop. Azad et al.19 experimentally investgated the heat transfer distribution in two-pass rectangular channels (AR=2:1) connected with a sharp 180° turn. The results showed that the roughened surfaces exhibit better heat transfer distribution than the smooth surfaces. Also the results showed that parallel 45° angled ribs produced higher heat transfer distribution than crossed 45° angle ribs case. However, from the above-mentioned research, few papers can be found in the open literature studied the rectangular or triangular cross section channel especially with rotation condition. Hence, the first motivation of this paper was to study two pass rectangular channels (AR=2:1) that is connected with a sharp 180° turn. The second motivation was to find different ribs configuration that trip the boundary layer and promote more cooling effect inside the two-pass rectangular channels. However, it wis found from a previous study by Han et al.4 that the 45° V-shaped ribs show higher heat transfer performance in a one-pass non-rotating square duct compared to other ribs configurations (45° angled ribs or transverse 90° ribs). Thus, we have chosen 45° V-shaped ribs to be placed on the principle surfaces in the two pass rotating rectangular channels since they have shown a potential for higher heat transfer performance. Moreover, the effect of the channel orientation with respect to plane of rotation was investigated for two positions b =90°, 135°. Such experimental data is not available in the open literature, which shows the combined effect of the 45° V-shaped ribs induced secondary flows, and rotational induced secondary flows on the heat transfer enhancement in the two-pass rectangular cross sectional channels.

0 0

Post a comment