Film Cooling Using Pulsed Coolant

AFRL research highlights the advantages of pulsed versus continuous coolant film cooling for turbine engines.

Turbine engine designers routinely use film cooling to cool engine components in the hot-gas flow-path. Film cooling is the process of injecting coolant fluid at one or more discrete locations along a surface exposed to a harsh, high-temperature environment. The film cools and thus protects turbine engine components, enabling the engine's operation at higher turbine inlet temperatures and increasing its thermal efficiency. Current turbine engine designs employ a continuous coolant flow, typically diverting 20%- 25% of the compressor's high-pressure air to cool turbine airfoils. By reducing the volume of high-pressure air needed for turbine blade cooling, designers can proportionately increase the flow available for combustion and thus increase thrust. Therefore, coolant flow reduction is an important design goal in the development of advanced turbine engines.

Researchers recently demonstrated that pulsed jets are an effective means for controlling primary flow in low-pressure turbines. They investigated the pulsed jets, also known as vortex generator jets, with secondary flow injected perpendicular to the primary flow and found that reducing the pulsed flows to a duty cycle (DC) of 1% continued to promote reattachment of the separated primary flow. These test results indicate that pulsed flows can significantly modify the near-wall boundary layer by effecting reattachment of the separated primary flow far downstream from the location at which convection would normally separate it from the airfoil.

Figure 1. Local film effectiveness distribution for pulsed coolant (left) and continuous coolant at M = 0.75 (right)

Motivated by the pulsed flow study results, AFRL scientists initiated a research effort to determine the effects of coolant pulsing frequency (PF) and DC on heat transfer coefficient (h) and film effectiveness (η) distributions. For the test specimen, they selected a semicircular leading edge test model with an after-body and positioned the traditional, cylindrical film hole 21.5° from the model's stagnation line. The researchers then used an infrared thermography technique to determine both the heat transfer coefficient and the film effectiveness distribution with a single transient test. (Details of this new measurement technique are accessible in an earlier AFRL Technology Horizons® article.1)

The research team controlled coolant pulsation by adjusting the opening and closing times of two synchronized pulsed valves, installed in the coolant supply line, to create a combination of PF and DC. The investigation comprised respective DCs of 50%, 75%, and 100% (continuous coolant) with PFs of 5 and 10 Hz. DC is the ratio of a valve's open time to its total time in the open-close cycle, expressed as a percentage value. Therefore, for a PF of 10 Hz and a DC of 75%, for example, the pulsed valve is open for 0.075 sec and closed for 0.025 sec for each cycle of the 10 Hz pulse.

Figure 2. Effect of coolant pulsation on film effectiveness (left) and Frossling number at M = 1.00 (right)

To conduct the study, the team used hot mainstream air and injected the cooler, coolant air through a cylindrical hole in the test model. They released both flows onto the ambient-temperature surface of the model simultaneously in a transient mode. The researchers performed all tests at Reynolds number (Re) 60,000, based on the model's leading edge diameter (D). They executed the test condition matrix by setting the continuous coolant at one of four blowing ratios (M): 0.75, 1.00, 1.50, or 2.00—where M is the mass flow rate per unit area for the jet, divided by the mass flow rate per unit area of the free-stream—and subsequently generated the pulsed cases by modifying the DC and PF without changing the coolant flow rate setting.

Figure 3. Effect of coolant pulsation on film effectiveness (left) and 0Frossling number at M = 1.50 (right)

The researchers employed the infrared imaging system to capture the response of the model's surface temperature. Using both the measured surface temperatures and the mainstream and coolant temperature histories, they determined the local heat transfer coefficient and film effectiveness at every point of the imaging field (320 x 240 pixels). Next, the researchers calculated a dimensionless Frossling number (hD/k/Re0.5, where k is the thermal conductivity of the fluid studied) to represent the heat transfer coefficient. Film effectiveness is the temperature difference between the mainstream flow and the model wall, divided by the temperature difference between the mainstream and the cooling stream. The goal of film cooling is to obtain a higher film effectiveness (to better protect the turbine blade/vane) and a lower heat transfer coefficient (to prevent the hot-gas stream from transferring heat to the turbine blade/vane).

Figures 1 illustrates the local film effectiveness (η) distributions at blowing ratio M = 0.75 for both the pulsed coolant (PF = 10 Hz, DC = 75%) and the continuous coolant, respectively. The film effectiveness distribution clearly shows the coolant jet trajectory. At this blowing ratio, the momentum (momentum ratio I = 0.51) of the coolant is much lower than that of the mainstream. The mainstream (coming from the left) deflects the coolant jet (exiting upward in the spanwise direction, Z, from the film hole). The coolant then travels primarily in the streamwise direction, X, without noticeable spanwise movement for either the pulsed or the continuous flow. The film effectiveness value peaks immediately downstream of the coolant injection location and then decreases due to the coolant's dissipation as it moves downstream. Compared to the continuous coolant, the pulsed coolant provides greater film coverage (as a result of its broader coverage), which contributes to the higher spanwise-averaged film effectiveness.

Figure 2 presents the effect of coolant pulsation at blowing ratio M = 1.00 on the spanwise-averaged film effectiveness and Frossling number, respectively. The researchers converted the imaging field into the physical arc length (X) from the stagnation line of the leading edge test model and the film hole diameter (d). In this coordinate system, the film hole spans the range between X/d = 3 and X/d = 4, with the hole centerline located at X/d = 3.5. The spanwise-averaged values peak immediately downstream of the injection location for both heat transfer coefficient and film effectiveness, just as they did in the M = 0.75 scenario depicted in the Figure 1 images. As shown, pulsed cases provided higher film effectiveness than did the continuous flow condition—a positive benefit of coolant pulsation. At this blowing ratio, the continuous coolant jet possesses a momentum (momentum ratio I = 0.90) slightly lower than that of the mainstream flow. The pulsed coolant jet, however, exhibits a much smaller (although not zero) momentum when the valves are closed, because the valve timing dictates that some residual coolant will remain in the coolant chamber until the valves open again. This variable momentum characteristic of the pulsed coolant results in higher film effectiveness than continuous flow cases are able to achieve. Likewise, all pulsed cases demonstrate heat transfer coefficients comparable to or slightly lower than those generated in the continuous flow cases—another positive advantage created by the coolant pulsation. Neither the PF nor the DC, however, has much effect on the heat transfer coefficient distribution.

Figure 3 reveals the results of increasing the blowing ratio from M = 1.00 to M = 1.50. At this blowing ratio, the continuous coolant jet possesses a momentum (momentum ratio I = 2.02) much higher than that of the mainstream. This increased momentum causes (1) the coolant jet to lift off the test surface, and (2) the film effectiveness to decrease at a faster rate downstream from the location of peak values. At this higher blowing ratio, the resulting coolant liftoff also contributes to lower film effectiveness further downstream as compared to all other injection rates studied, including the continuous flow case at M = 1.00. The coolant pulsation functions as a cushion to reduce the adverse effects of the liftoff, especially during the period when the valves are closed. The coolant jet continues to possess a strong lateral momentum during the period that the valves are open, but only a portion of the coolant flow lifts off the test surface to produce higher film effectiveness for the pulsed cases. The pulsed jets exhibit lower Frossling numbers than those of the continuous jet, and the distribution depends more on DC than on the PF. It follows that the longer the DC, the greater the heat transfer becomes. The shorter DCs, with reduced coolant flow effectiveness, decrease the coolant's available time for mixing with the mainstream flow and, as a result, produce a lower heat transfer coefficient or Frossling number.

This research effort clearly demonstrates that turbine engine designers can significantly reduce film cooling flow requirements by substituting pulsed coolant flow for continuous blowing methods. Using pulsed flow, they can reduce film cooling flow ranges by 13.0%-46.5% (depending on DC, PF, and M), while improving film effectiveness and lowering heat transfer coefficients. These important findings will contribute to the design of higher-performing turbine engines.

Dr. Shichaun Ou and Mr. Tom Brown (Universal Technology Corporation), of the Air Force Research Laboratory's Propulsion Directorate, wrote this article. For more information, contact TECH CONNECT at (800) 203-6451 or place a request at http://www.afrl.af.mil/techconn_index.asp . Reference document PR-H-06-02.

Reference

  1. Brown, T., Ou, S., and Rivir, R. "Innovative Technique to Obtain Heat Transfer Coefficient and Film Effectiveness." AFRL Technology Horizons®, vol 6, no 2 (Apr 05): 18-19. http://www.afrlhorizons.com/Briefs/ Apr05/PR0407.html.