This article discusses that military-sponsored research tools can improve the machines that drive civil applications. The Defense Evaluation and Research Agency (DERA) researchers tested the engine of the legendary DeHavilland Vampire single seat jet fighter in the late 1940s. This Vampire is owned by Fred Ihlenburg, president of Yakity Yaks Inc., an importer of foreign military aircraft, based in Aurora, Oregon. DERA is investigating heat transfer on turbine blades to help gas turbine manufacturers develop a cooling system that will keep blades at an optimum temperature while minimizing losses in engine performance. More efficient cooling means less air is bled from the compressor, thus improving performance while extending blade life. This work was co-funded by the Central European Commission under the Brite Euram Fourth-Framework Initiative, which is part of the European Union’s strategy to enhance European global competitiveness, and Britain’s Department of Trade and Industry’s Civil Aircraft Research and Technology Demonstration Program. The British program aims to advance the capabilities of the United Kingdom’s civil aerospace companies.
LIKE RADAR, ATOMIC fission, television, and the forging of iron, the development of gas turbines has been closely linked with defense purposes. The military tasked engineers with creating turbines that were compact, reliable, and powerful enough for jet aircraft to span the world and reach their Cold War targets. The design innovations of these engines carried across into commercial gas turbines over the years, particularly aeroderivative engines.
Today, as Western governments shrink their defense budgets in the post-Cold War world, engineers who designed the gas turbines for jet fighters and bomber engines are bending their expertise toward the civilian sector. This is the case with the Defense Evaluation and Research Agency, better known as DERA, part of the United Kingdom’s Ministry of Defence, and with several engineers working at Ohio State University’s Gas Turbine Laboratory in Columbus.
DERA combines the bulk of the Ministry of Defence’s non-nuclear research, technology, test, and evaluation establishments, involving over 12,000 staff, making DERA one of the largest European research organizations. The agency presented its latest turbine research projects at the ASME’s Turbo Expo, held in Munich during May.
The agency’s pedigree in joining civilian and military gas turbine engineers dates back to 1946, when Sir Frank Whittle’s Power Jets team first joined forces with the British government’s gas turbine facility at Pyestock. Engineers at the Pyestock facility tested the gas turbines from Britain’s first-generation jet fighters, such as the DeHavilland Vampire and the Gloster Meteor, up to the Westinghouse/Rolls Royce WR-21 marine engine being developed for the United States, United Kingdom, and other NATO navies.
The Propulsion and Performance Department at Pyestock employs approximately 170 people and focuses primarily on gas turbines and related technologies. For example, the Pyestock staff recently conducted tests that involved using a temperature sensor embedded on a blade to investigate heat transfer effects in a high-temperature gas turbine.
Probing Heat Transfer
DERA is investigating heat transfer on turbine blades to help gas turbine manufacturers develop a cooling system that will keep blades at an optimum temperature while minimizing losses in engine performance. Increasingly, high temperatures of gas turbines, up to 2,100 Kelvin, take a toll on turbine blades whose thermal resistance is limited to about 1,100 K.
If turbine designers were able to adequately predict the surface heat transfer distribution on turbine blades, they would be able to devise ways to cool the costly blades and extend their performance life. Cooling flows are taken from the turbine’s compressor stage, but this will reduce overall engine performance. More efficient cooling means less air is bled from the compressor, thus improving performance while extending blade life.
This work was co-funded by the Central European Commission under the Brite Euram Fourth-Framework Initiative, which is part of the European Union’s strategy to enhance European global competitiveness, and Britain’s Department of Trade and Industry’s Civil Aircraft Research and Technology Demonstration Program. The British program aims to advance the capabilities of the United Kingdom’s civil aerospace companies.
To measure heat transfer effects on turbine blades, DERA engineers had to insert a temperature sensor on the blade tip itself. “The most difficult part of the program was developing a method for mounting the sensor that could withstand the 30,000 g acceleration that is experienced at the tip of the rotor blade. The clearance gap between the rotor and shroud is on the order of 0.5 millimeter. Hence, the sensors had to be kept small, so they can be mounted on the rotor tip without affecting the gap,” said Kam S. Chana, a mechanical engineer and technical team leader for propulsion and performance projects at DERA.
Chana and his DERA colleagues met this challenge by first bonding a plastic film onto the test turbine blade, then machining it so that the temperature sensor could be flush mounted. The thin-film sensors used in the turbine tests were jointly developed by Oxford University and DERA. Each sensor measures changes in surface temperature by converting those changes into electrical signals. These signals are compared to the temperature history and thermal properties of the blade metal to calculate the heat flux experienced by the blade.
The sensors and their measurement system have a frequency of approximately 100 kilohertz. The high-frequency sensor signals are transmitted through films embedded under the pressure surface of the blades. Wires connected to the films at the base of the blades run along the face of the rotor disc and through the center of the turbine shaft.
Electronic modules are mounted in the rotating frame to condition sensor signals and power the sensors. Computers collect data from the blade tip sensor in real time via a 24-channel slip ring.
Another challenge for DERA researchers was the presence of strong, three-dimensional secondary flows and unsteadiness generated by the successive stationary and rotating blade rows, which made predicting the local heat transfer rates more difficult. The British engineers overcame this obstacle by taking time-resolved heat transfer measurements using the blade tip sensors.
The turbine blade research at Pyestock has provided DERA with significant insight into the flow around and over the tip of a shroudless high-pressure turbine blade, improving the understanding of underlying flow physics, according to Chana.
“These data are now being used to validate computational methods and to develop design rules for future advanced turbines,” Chana said. “The levels of heat flux measured can be used to assess the cooling requirement for blade tips and shrouds.”
Tiny Pockets Yield Big Information
A group of engineers at Ohio State is conducting turbine blade tip research that began in 1976, while the group worked at Calspan Advanced Technology Center in Buffalo, N.Y. “Our gas turbine work was originally funded by Wright-Patterson Air Force Base that year, and was subsequently funded by NASA, GE Aircraft Engines, Allison, and Pratt & Whitney,” recalled Mike Dunn, an ASME Fellow and director of Ohio State’s Gas Turbine Laboratory. The laboratory is part of the university’s Aerospace Engineering and Aviation Department.
Dunn and four associates left Calspan in 1995 and proved that you can take it with you. They purchased the company’s turbine research facility and moved it—lock, stock, and sensors—to Ohio State. This allowed them to use the same advanced instrumentation technology to research heat transfer and aerodynamic loading at turbine blade tips in projects funded by NASA, Wright-Patterson AFB, GE Aircraft Engines, Honeywell, Pratt & Whitney, and Siemens Westinghouse.
Rather than place sensors in a layer of plastic, the Ohio State engineers, using electric discharge machining, burn tiny pockets directly in the steel of the turbine blades. They insert sensors in the holes and bond them in place with adhesives. The pockets are approximately 1 mm wide, 1.5 mm long, and 1 mm deep.
The thermal sensors are platinum thin-film devices attached to an insulating substrate that are calibrated for temperature versus resistance. A known constant current is passed through the thin films. The resistance of the sensor, and thus its measured voltage, changes during the experiment, reflecting a temperature change in the film, which can be converted to the imposed heat rate via a one-dimensional heat conduction analysis.
“Our research locates the areas of greatest heat transfer on the tips, lips, and floor of recessed tip turbine blades, which are often used in aircraft turbines,” said Dunn.
Ohio State’s turbine tip research dealing with recessed tips has been conducted on an Allison turbine, and is described in paper 2000-GT- 0197 of the Proceedings of Turbo Expo 2000. Earlier studies concerning flat blades were performed for a Garrett turbine and a GE Aircraft Engine turbine. “We are now studying recess turbine tip heat transfer among other important turbomachinery issues with GE Aircraft Engines under the aegis of their University Strategic Alliance program,” noted Dunn.
Another major area of study by DERA and the Ohio State turbine lab team is to measure unsteadiness in turbine loads so that turbine manufacturers can improve the design of their high work turbine stages. “Unsteadiness in turbine loads is caused by vane/blade interaction, or VBI; that is, the rotor blades cutting through the wakes of the stator vanes, and the shock waves caused by transonic turbines,” explained Dunn. This can have a profound influence on component fatigue life.
DERA ’s engineers make use of powerful computers and advanced instruments to develop computational fluid dynamics techniques to analyze turbine unsteadiness for Alstom Power in Whetstone, U.K. The company will use DERA s on-site flow measurement and analysis service as part of its program to develop high-efficiency, high-load, high-pressure turbine stages.
Alstom Power believes this research will yield turbine designs capable of providing a minimum 4 percent increase in turbine efficiency. The turbine manufacturer asked DERA to measure the unsteadiness of three different build standards of its high-load, high-pressure stage turbine, and to perform CFD analysis of the turbine design geometries.
The first phase of DERA s work, begun in July last year, involved testing the current datum, high-load HP turbine in Alstom Power’s warm-air turbine rig, based at the company’s research center in Whetstone. The first measurements consisted of traverse tests made by positioning sting- mounted probes downstream of the turbine. Stings are slender metal rods used to project the sensors into place.
The DERA engineers used two types of probes in the July test: a four-sensor pyramid probe and a single-sensor probe. The four-sensor probe was designed and manufactured at the Osney Laboratories in the Department of Engineering Science at Oxford University. This instrument measured the pressure and speed of turbine unsteadiness, as well as its yaw and pitch. The frequency response of the four-sensor probe system was on the order of 100 kHz, sufficient to track pressure changes caused by rotor passing events, which occur at 9.5 kHz.
The single-sensor probe was fitted with a high-response, XCQ 062 Kulite pressure transducer. This device was designed and manufactured by Kulite Semiconductor Products Inc. of Leonia, N.J. The company has collaborated with Oxford on the development of jet engine sensors for a quarter-century.
The XCQ 062 is a piezoresistive pressure transducer consisting of two layers of single crystal silicon wafer. One layer is a monolithic structure composed of an atomically fused, dielectrically isolated Wheatstone bridge integrated circuit on a silicon substrate that acts as a forcesumming diaphragm. As the layers flex in relation to pressure, they create a voltage change that is proportional to the pressure exerted on the face of the wafer.
The DERA researchers used the single-sensor probe to measure the time-varying, total pressure fluctuations at the rotor exit. Again, DERA worked with Oxford’s Osney Laboratories to keep the probe tip’s diameter to a maximum of 2 mm to reduce the flow disturbance that the probe would cause.
The probe was rotated through 180 degrees around the stem centerline to the ends of the turbine wall, enabling the DERA technicians to take measurements to within 0.5 mm of the turbine casing’s length and end. After collecting measurement data, DERA processed it using its in-house CFD solver TRANScode.
The second phase of the Alstom Power high-load turbine project began in June, and will focus on developing a new, state-of-the-art turbine design. The final phase of the project will involve incorporating lessons learned from the DERA tests to produce an improved turbine design that Alstom Power will use.
The Ohio State engineers began researching turbine unsteadiness back in 1986 in cooperation with Allison Gas Turbines Co. and with the support of Wright-Patterson AFB. They used a turbine designed by Allison to study vane/blade interaction. Dunn’s group continued the research for NASA, and today conducts the work under the sponsorship of three private sector turbine makers, namely GE Aircraft Engines, Pratt & Whitney, and AffiedSig- nal. Again, Dunn and his colleagues machine pockets on the turbine blades to flush-mount miniature pressure transducers that are bonded in place by adhesive. They also attach strain gauges to the blade surface, using conventional adhesives to measure the structural response of the blades in response to the unsteadiness of the loads.
The Ohio State researchers flush-mount and bond Kulite XCQ 062 pressure transducers, the same models used by DERA at Pyestock, on the vanes the same way. “We also burn pockets into the blades to flush-mount piezoelectric crystals in order to excite them in a vacuum before the airflow is commenced so that mechanical damping can be separated from aero damping in a single experiment,” Dunn explained.