Abstract:
An important parameter for turbomachinery designers is “clearance control”,
because the clearances between interfaces must be set to optimum values to
maximize power output, operational life and efficiency. Leakage of hot gas result-
ing from excessive clearance, can lead to flow instabilities, components overheat-
ing, lower cycle efficiency and a dramatic increase in specific fuel consumption
(SFC).
Seal segments are used to reduce blade tip leakage, maintain coolant air flow
and the stability of rotor-dynamic systems, helping to maximize blade perfor-
mance. Seal segments in the High-Pressure Turbine (HPT) stages are one of the
hottest components as they face the hot gases coming from the combustion
chamber with temperatures which can reach 1700 0 C and which makes them sub-
ject to oxidation, erosion, and creep. Thus, seal segments need to be protected.
They are currently cooled using jet impingement techniques, passing cooling air
(supplied by the high-pressure stage of the compressor) through channels to di-
rectly impinge on the hot surfaces.
The focus of this research was to improve the jet impingement cooling of the
seal segments in HPTs by investigating methods that provide more effective heat
removal. The role played by configurations of ribs (surface roughness using be-
spoke turbulators), custom-made seal-segments, and surface features such as
contouring, both in isolation and combination, were investigated using numerical
methods. A set of 174 simulations were carried including the use of uniform and
non-uniform roughness elements with different shapes and heights.
Firstly, three different uniform roughness elements were tested, a square
cross-sectional continuous rib, a hemi-spherical pin-fin and a cubical pin-fin for
three jet impingement angles of α=90°, 60° and 45°. Each roughness element
was also tested for six different heights (e) between 0.25 mm and 1.5 mm in increments of 0.25 mm. Results are presented in the form of average Nusselt
number within and beyond the stagnation region.
Secondly, the effect of using a roughness element with a square cross section
in the shape of a circle, on the average Nu was investigated for four different
radial locations (R), three jet angles (α) and six rib heights (e). Finally, the roughness element used was continuous, of square cross-sec-
tion, in the shape of tear drops and reversed tear drops. This meant the rib did
not act as a total barrier to flow in either the uphill or downhill direction.