No abstract

Influence of curvature distribution and area-ratio (AR) distribution on the pressure fields within the curved annular diffuser are discussed. General guidelines for end-wall contouring to control the pressure gradients on the diffuser walls are evolved and further demonstrated through computational fluid dynamics (CFD) simulations. Also, detailed guidelines for controlling the adverse pressure gradients (APG) on duct walls are presented. A geometry generation methodology (GGM) which enables both design and evaluation of curved annular diffusers based on the guidelines evolved is presented. The approach presented deals with the sensitivity of the duct performance parameters to duct wall modifications. In that sense, the work per se is not a description of an automated optimization process, but rather about the physical principles that can guide such an optimization. An aggressive diffuser design space is identified with ducts of maximum slope of 50 deg and maximum divergence angle between the outer and inner walls of 10 deg for length to inlet height ratio ranging from 1.25 to 2.5. Part of the identified design space for which the flow separation can be eliminated based on the guidelines evolved is demarcated. The need for flow control, possibly passive, is established for more aggressive designs.

The use of splitter blade as a passive flow control mechanism in the design of separation free aggressive annular diffuser is explored through computational fluid dynamics simulations. The fundamental working principle of a splitter blade in case of a two-dimensional rectangular diffuser and an annular diffuser is discussed. The effects of splitter blade configuration on the end-wall adverse pressure gradients are discussed. An aggressive diffuser design space is identified with ducts of maximum slope of 50 degree and maximum divergence angle between the outer and inner walls of 10 degree for axial-length to inlet height ratio ranging from 1.25 to 2.5. One or more splitter blades are employed in the duct to eliminate flow separation for all the ducts in the aggressive design space considered. Requirement of number of splitter blades in the aggressive design space is demarcated. Performance charts for the ducts in this aggressive design space are also established.

Additive Manufacturing (AM) in turbine technology enables the manufacture of complex and detailed shapes such as optimized cooling channel designs. However, the AM components are usually produced with high surface roughness. The ability to predict the pressure loss and heat transfer during the AM components’ design phase gives the designer an extra edge to arrive at a better design. In this paper, numerical prediction for the effect of roughness on pressure loss (friction factor) and heat transfer (Nusselt number) for flow in the internal channel is discussed. Numerical simulations were performed using the commercial computational fluid dynamics (CFD) tool by ‘Siemens Star-CCM+’. Initially, the best practices for the CFD process were arrived at by comparing the CFD results with the theoretical correlations for a fully developed channel flow. Further, these best practices were invoked to validate the two test cases from the open literature. From the first test case, three test coupons with the lowest, intermediate, and the highest level of roughness were selected for the validation. From the second test case, in-line and stagger configurations were selected for validation. To reduce the simulation time, modeling the full channel domain as a single channel was explored. The single-channel results were found to be matching well with the full channel results at the two Reynold’s numbers simulated. For all the cases, the friction factor predictions are close to the theoretical and test data, whereas the Nusselt number predictions show a consistent trend with the theoretical data but over-predicts when compared to the test data.