Task 11Roughness effects on film cooling performance

A series of experiments were completed to evaluate the effect of surface roughness on the film cooling performance from the first row of holes on the suction side of the simulated vane. As shown schematically in Figure 12, the simulated vane has six rows of coolant holes in the showerhead region, two rows on the pressure side, and three rows on the suction side. The adiabatic film cooling performance for this simulated vane has been documented in numerous previous publications, eg. Polanka et al. (1999) and Ethridge et al. (2001). The focus of this study was the effects of roughness on the adiabatic effectiveness for film cooling from the first row on the suction side of the vane (x/d = 0 on Figure 12).

Experimental conditions

The array of conical roughness elements used in these experiments were described in the previous semi-annual report (Mar-Aug, 2001). The roughness Reynolds for the array of roughness elements is Rek = 60, which is a "fully rough" condition. Three roughness configurations were studied: roughness upstream of the film cooling row of holes, roughness downstream, and roughness upstream and downstream. The configuration for roughness upstream of the row of holes had a roughness section that started 44 mm upstream of the row of cooling holes, and extended to a point immediately upstream of the holes, i.e. a total streamwise extent of about 11d (where d is the coolant hole diameter). The downstream extent of roughness elements was 85 mm or 21d.

All film cooling holes had a diameter of d = 4.11 mm. Holes for the first row on the suction side of the vane had an injection angle of 50° with respect to the surface, and were oriented in the streamwise direction. Adiabatic effectiveness tests were generally conducted with a density ratio of DR = 1.6, though some tests were done at a lower density ratio of DR = 1.2. Low and high mainstream turbulence levels were used, Tu = 0.5% and Tu = 20%, respectively. Tests were conducted with the showerhead holes taped over to produce a smooth leading edge condition. When the showerhead was blowing, a blowing ratio of M = 1.6 was used (based on the average blowing velocity from the showerhead and the approach velocity to the vane). All tests were conducted with an approach velocity of U0 = 5.8 m/s.

Adiabatic Effectiveness Results

As indicated in Table 1, the repeatability of the test results was established by a number of repeated experiments with various operating conditions. The repeatability of the laterally averaged adiabatic effectiveness was generally better than 8~tf = ±0.015. The following results are presented with a focus on three blowing ratios: M = 0.7, 1.0, and 1.4. In each case the maximum adiabatic effectiveness was found at the lowest blowing ratio. At the higher blowing ratios the coolant jets are progressively more separated.

Figure 13 shows the effect of having roughness upstream of the cooling holes with low mainstream turbulence and no showerhead blowing. The upstream roughness was found to cause a dramatic decrease in adiabatic effectiveness. This decrease can be attributed to the effect of the surface roughness on the approach conditions of the boundary layer. The roughness probably causes a much thicker boundary layer that is less effective in turning the coolant jet towards the surface. Consequently the coolant jet separates a greater distance from the surface resulting in decreased adiabatic effectiveness.

With a roughness section only downstream of the coolant holes, shown in Figure 14, there is still a decrease in adiabatic effectiveness, but the decrease is much less than for the upstream roughness. The decrease in adiabatic effectiveness with a downstream roughness section may be attributed to an increase dispersion of coolant by increased turbulence levels caused by the surface roughness.

The effect of roughness sections upstream and downstream of the cooling holes is presented in Figure 15. At M = 0.7 there was still a significant decrease in adiabatic effectiveness, but not as much as for the upstream alone roughness section. At higher blowing ratios of M = 1.0 and 1.4 there was only a slight decrease in adiabatic effectiveness. Apparently the downstream roughness section has a compensating effect so that the combined upstream and downstream roughness causes a smaller decrease in adiabatic effectiveness that upstream roughness alone. Recall that upstream roughness was speculated to cause increased separation of the coolant jets. The downstream roughness could mitigate the effect of jet separation by increasing turbulence and hence transport of coolant back to the surface.

With showerhead blowing, and high mainstream turbulence, the operating conditions best simulate actual turbine airfoil conditions. Experiments were conducted under these conditions with smooth surfaces and with a roughness section upstream of the cooling holes. Adiabatic effectiveness results for these conditions are presented in Figure 16. These results show that the upstream roughness causes a significant decrease in adiabatic effectiveness, approximately 20% to 25%, for all blowing ratios. Of particular interest was the An = 0.05 decrease for M = 1.4 that suggests that the roughness is dispersing the coolant from the showerhead injection as well as disrupting the coolant injection from the suction side holes.

0 0

Post a comment