Rotational Effects on Heat Transfer at Advanced Engine Conditions

Abstract

Cooling of Gas turbine buckets to ensure adequate life margin with coolants at advanced engine conditions of pressure and temperature requires that the internal heat transfer as influenced by rotation be known with sufficient accuracy. The existing data base, comprised of information readily available in the open literature has limited advanced design applicability considering the range of geometric and fluid-thermal dimensionless parameters of interest. Further, studies conducted with the aid of computational fluid dynamics (CFD) have demonstrated that, as the range of such parameters is extended, the characteristics of the predicted heat transfer capability change significantly, to the extent that bucket coolant passage designs that are fully acceptable at state-of-the-art conditions are marginal or unacceptable at advanced conditions. The purpose of this paper is threefold, to: a) demonstrate the need for extending the current internal heat transfer data base, accounting for the effects of rotation, b) provide a physical understanding of the expected heat transfer characteristics at advanced conditions, and c) describe a test rig and test program focused on obtaining the data of interest.

Effect of Aspect Ratio on the Buoyancy Driven Reverse Flow Near the Leading Wall of Rotating Cooling Passages

“…The buoyancy number (Bo) has been used as a parameter to study the dependence of flow behavior and heat transfer in rotating channels in earlier investigations (Staub et al, 1995 andTelcriwal, 1994b) for flows at a given Reynolds number. It is also the ratio of Grashof number (Or) to Reynolds number (Re) squared and can be derived to be the product of rotational, buoyancy and geometrical non-dimensional parameters as in the work of Tekriwal (1994b):…”

“…The drop in heat transfer (Nusselt number) at the leading wall has been experimentally observed by several researchers (Wagner et al, 1991a and b;Yang et al, 1992;Han et al, 1992 and1994) and also numerically predicted by several investigators (Prakash and Zerkle, 1992;Telcriwal, 1994a, c and1995;Dutta et al, I 994a and b). The drop in heat transfer can' be very significant in the stagnation region causing a concern for design engineers, especially at high buoyancy number flows (see Staub et aL, 1995). The reverse flow near the leading wall can penetrate all the way close to the inlet if the buoyancy number is high at low to moderate Reynolds number flow.…”

“…The reverse flow near the leading wall can penetrate all the way close to the inlet if the buoyancy number is high at low to moderate Reynolds number flow. The rise in fluid velocity near the trailing wall can cause a significant acceleration and friction pressure drop and affect the flow adversely in the parallel flow passage configuration designs (Staub et al, 1995).…”

“…The coolant selected is a pressurized (approximately 10 atmospheres) high density fluid, R -134A so that relatively low rotational speeds (170-750 rpm) are needed to attain tbe required rotation numbers (0.08 to 0.35 in current computations). Low speed requirement reduces the cost and complexity of an actual test rig planned to acquire experimental data (see Staub et al, 1995). For all the three ducts, computations have been carried out for Reynolds number equal to 80,000.…”

“…Depending upon the flow parameters the effect of the centrifugal buoyancy can lead to radially outward flow velocity near the trailing wall so high that a flow reversal near the leading wall (that is, radially inward flow) is caused in order for the continuity equation (conservation of mass) to be satisfied (see Telcriwal, 1994b). This effect at high rotational speeds can cause the reverse flow to penetrate all the way near inlet of the duct, thus causing hydrodynamic instabilities in the flow (see Staub et al, 1995 andTekriwal, 1994b). A typical radial velocity profile in the case of reverse flow situation is shown in Fig.…”