0

Manufacturing Materials and Metallurgy

1996;():V005T12A001. doi:10.1115/96-GT-219.
FREE TO VIEW

In many advanced combustor concepts, such as the RQL (rich-quench-lean) combustor, the requirement of low NOx emission makes film cooling of the hot gas path surfaces undesirable. Double-walled structures with relatively low aspect ratio (height/width) rectangular passages and with well-controlled thin hot gas side metal walls are an alternative to film cooling. The additional application of other cooling techniques, such as impingement and surface enhancements, make efficient use of the limited cooling air available. However, the use of cooling channels to increase heat transfer coefficients on the coolant side may ultimately require well-bonded structures. Alternate methods of bonding have been considered for Ni-base alloy HA230-to-HA230 joints in thin-walled structures, with emphasis on bonds produced by hot isostatic pressing and by laser welding. For hot isostatic pressed (HIPed) structures, mechanical strength and ductility have been measured as a function of temperature for structures prepared after a number of different surface cleaning treatments prior to joining. The surface cleanliness has been characterized by scanning electron microscopy, and the after-HIP bond line microstructure has been evaluated as formed and after mechanical testing. Characterization of laser welds produced at Laserdyne for Acraline Products has consisted of scanning electron microscopy of the weld surfaces and metallography of weld/substrate cross-sections, looking at solidification/heat affected features and defects.

Commentary by Dr. Valentin Fuster
1996;():V005T12A002. doi:10.1115/96-GT-220.
FREE TO VIEW

Ultrasonic examination methods have been developed to qualify the locally repaired coating on gas turbine blades. The acoustic velocity of the MCrAlY coating obtained by Inert Gas Plasma Spray (IGPS) has been measured and compared with that of Vacuum Plasma Spray (VPS) coating. The thickness of the locally repaired coating on specimens and blades has been evaluated. The coating adhesion on the base material and its distribution have been investigated by A-scan and C-scan data display methods. The influence of surface roughness on the test results has also been examined.

Commentary by Dr. Valentin Fuster
1996;():V005T12A003. doi:10.1115/96-GT-380.
FREE TO VIEW

INCONEL® alloy 783 is an oxidation resistant low coefficient of thermal expansion (low CTE) superalloy developed for gas turbine applications. Turbine efficiency can be increased through the use of low CTE shrouds and case components that maintain tight blade tip clearances at different turbine operating temperatures. To achieve low CTE, alloys based on Ni-Fe-Co compositions require Cr content be maintained at low levels. Added Cr lowers the Curie temperature and thereby increases thermal expansion rate over a wider temperature range. The necessary lack of Cr minimizes resistance to both general oxidation and stress accelerated grain boundary oxygen enhanced cracking (SAGBO). Increased amounts of Al in alloys strengthened by γ’ alone also promotes SAGBO. Alloy 783 is the culmination in the development of an alloy system with very high aluminum content that, in addition to forming γ′, causes β aluminide phase precipitation in the austenitic matrix. It was discovered that this type of structure can be processed to resist both SAGBO and general oxidation, while providing low thermal expansion and useful mechanical properties up to 700°C. The high Al content also reduces density to 5% below that of superalloys such as INCONEL alloy 718. Key aspects of the alloy development are presented, including the assessment of SAGBO resistance by evaluating elevated temperature crack growth in air. The alloy, now commercially available, has been successfully fabricated and welded into gas turbine engine components.

Commentary by Dr. Valentin Fuster
1996;():V005T12A004. doi:10.1115/96-GT-390.
FREE TO VIEW

This paper describes an improvement in creep strength and thermal fatigue and hot corrosion resistance of a cobalt base superalloy developed for the applications of heavy duty gas turbine nozzles. The developed alloy has superior resistance to creep, thermal fatigue and hot corrosion over that of conventional alloys.

The optimization of alloying elements which improve the creep properties and restrain the coarsening of carbides is discussed. A reduction in the volume fraction of eutectic carbides promoted the prevention of fast thermal fatigue cracking and also increased the hot corrosion resistance of the developed alloy.

The mechanical properties were evaluated by general method for gas turbine materials. Thermal fatigue property was examined by introducing cyclic thermal stress into test pieces. During the test, crack propagation of the test pieces was observed. Hot corrosion resistance was evaluated by molten salt corrosion tests. After tests, mass loss of specimens was measured and penetration depth of sulfidized scale was observed. The developed alloy showed good properties which can allow a wide temperature margin for high temperature gas turbine nozzle designing.

The developed alloy can be applicable to other gas turbine hot sections. Application of the developed alloy could realize an increase in gas firing temperature or extension of lifetime for current conditions.

Commentary by Dr. Valentin Fuster
1996;():V005T12A005. doi:10.1115/96-GT-426.
FREE TO VIEW

Hot corrosion and long-time dynamic oxidation tests were conducted on several alloy (equiaxed MM-0011 and IN738. SX CMSX-4. and DS CM186LC) and coating (Simple aluminide, Pt-aluminide, and NiCoCrAlY) combinations. The coating performance in both hot corrosion and oxidation testing was, in most cases, influenced by the substrate composition. An overlay NiCoCrAlY on either CMSX-4 or CM186LC appeared unaffected visually, but there were indications during metallographic examination that this coating substrate combination was beginning to degrade after a 1000 hour hot corrosion test. The results of the oxidation test indicated that several of the alloy/coating combinations that were superior in the hot corrosion test were inferior in oxidation. After 6000 hours of dynamic oxidation testing at 1038°C (1900°F) two of the alloy/coating combinations had not yet reached negative weight changes. Metallographic results from the oxidation test show a significant reaction layer in the CMSX-4 coated with either simple or modified aluminides or NiCoCrAlY overlay coatings.

Topics: Crystals , Coatings
Commentary by Dr. Valentin Fuster
1996;():V005T12A006. doi:10.1115/96-GT-427.
FREE TO VIEW

The goal of all repair processes is to return the hardware to a serviceable condition. Diffusion braze repairs utilize metallurgical processes to achieve economical repairs of expensive gas turbine components, especially in the turbine section. Component repairs often require dimensional restoration and crack repair on the same part. To achieve this goal, a new diffusion braze repair alloy was developed that combines high strength crack repair and dimensional build up into one material. This new material has mechanical property strength approaching that of the base metal. The improved mechanical properties result from a homogenous gamma prime strengthened diffusion braze zone.

As part of an FAA approved test plan, the Howmet ESR (Effective Structural Repair) diffusion braze material was evaluated by tensile and stress rupture testing at elevated temperature. The test results showed high tensile strengths and long stress rupture life. In addition, the effect of the diffusion braze thermal cycle was evaluated on the base metal. A comparison was made between the gamma prime size and shape of engine run JT8D LPT vane clusters before and after the thermal cycle. The thermal cycle was shown to have a beneficial effect on the gamma prime size and shape relative to overaged engine run nozzles. The low cycle fatigue (LCF) life of MarM247 was also shown to improve with the ESR thermal cycle relative to a typical LPT nozzle heat treatment.

Commentary by Dr. Valentin Fuster
1996;():V005T12A007. doi:10.1115/96-GT-435.
FREE TO VIEW

Tungsten carbide thermal spray coatings are important to the aerospace industry for reducing wear on jet engine components, fanblade mid-span dampers being one example. However, the fatigue life of a component is often reduced when a coating is applied and for some cases the coating can fail due to spallation and cracking. Coating failures can result in decreased engine performance and costly maintenance. To provide insight and possible explanations for the reduced service life of coated mid-span dampers, identify the best coating and application processes for future use, and to develop methods for improving coating performance, a comprehensive experimental research program was conducted. The program involved coating performance in jet engine tests, coating crack resistance in bending, low cycle fatigue properties of the coating and substrate, and microstructures for a wide range of coating compositions and application processes. Eleven coatings were ranked according to their performance relative to the other coatings in each evaluation category. Five of the coatings were selected for engine test runs. Results from the engine test runs for more than 800h (AMT-cycle) were compared to bend and low cycle fatigue evaluations and to meassured residual stresses. Strong correlation between engine performance and the residual stresses in the coating-substrate system was found. Results from the program were used for selecting a suitable coating system for final in-service use.

Topics: Engines , Dampers , Blades
Commentary by Dr. Valentin Fuster
1996;():V005T12A008. doi:10.1115/96-GT-436.
FREE TO VIEW

Metallurgical evaluation of platinum aluminide coatings applied to industrial gas turbine components, for oxidation and high temperature hot corrosion protection, were conducted. Coatings were processed by electroplating a thin layer of platinum followed by aluminizing using either the pack cementation or the chemical vapor deposition (CVD) processes. Laboratory and field data on the performance of these coatings are presented. Results from these tests showed that both aluminizing processes produced coatings that provided adequate environmental protection. However, the CVD coating experienced less coating growth during engine service and was therefore determined to be thermally more stable than the pack cementation coating in this application.

Commentary by Dr. Valentin Fuster
1996;():V005T12A009. doi:10.1115/96-GT-451.
FREE TO VIEW

Performance of turbine components can significantly be affected by surface/subsurface characteristics. Techniques applied today and being developed further entail introducing into a components surface a residual compressive stress of predictable magnitude and depth followed by superfinishing to improve the surface finish.

The effect will be to lower the mean stress which will increase component life and fatigue strength or enable higher loads through present designs; develop a damage tolerant layer capable of withstanding corrosion pitting or strike damage while in service and produce a final roughness capable of improving flow characteristics on turbine blades/buckets.

The processes to achieve the above include Controlled Shot Peening and Superfinishing. In combination, an optimised surface condition will result.

Commentary by Dr. Valentin Fuster
1996;():V005T12A010. doi:10.1115/96-GT-467.
FREE TO VIEW

Diffusion brazing is a joining process utilized both in the manufacture and repair of turbine blades and vanes. CMSX-4 is an investment cast, single crystal, Ni-based superalloy used for turbine blading and vanes, and has enhanced mechanical properties at elevated temperatures when compared to equiaxed, directionally solidified and first generation single crystal superalloys. The objective of this work was to develop a diffusion brazing procedure to achieve reliable joints in the manufacture of a hollow turbine blade (for a prototype engine in South Africa), and to verify the coatability of the diffusion brazed joints. Two commercially available brazing filler metals of composition Ni-15Cr-3.5B and Ni-7Cr-3Fe-4.5Si-3.2B-0.06C and a proprietary (wide gap) braze were utilized. With the aim of eliminating brittle centre-line boride phases, the effects of temperature and time on the joint microstructure were studied. Once the metallurgy of the joint was understood, tensile and stress rupture tests were undertaken, the latter being one of the severest tests to evaluate joint strength. The results demonstrated that the diffusion brazed joints could satisfy the specified stress rupture criterion of a minimum of 40 hrs life at 925 °C and 200 MPa. After mechanical property evaluations, an investigation into the effects of a low temperature high activity (LTHA) pack aluminide coating and a high temperature low activity (HTLA) pack aluminide coating on the braze joints was undertaken. The results showed that diffusion brazed joints could be readily coated.

Commentary by Dr. Valentin Fuster
1996;():V005T12A011. doi:10.1115/96-GT-488.
FREE TO VIEW

The demand for increasing gas inlet temperatures in modem gas turbines up to 1500°C and above is the main reason for the need for more reliable thermal barrier coatings. New ceramics should provide higher phase stability and better resistance against chemical attack by pollutants in the combustion gas.

Electron-beam physical vapor deposition (EB-PVD) processed, ZrO2-based TBCs were generated on bond-coated superalloy directionally solidified (DS) samples. Common yttria-stabilized zirconias of two different compositions, as well as novel stabilizers like CeO2 and La2O3 were investigated. A columnar structure was established during high-rate deposition in all cases. Diameter, degree of ordering of the columns and phase composition depended on stabilizer oxide and content. The role of differences of vapor pressures is addressed with regard to chemical homogeneity of the coatings.

The performance of the TBCs having various stabilizers was investigated in a cyclic oxidation furnace test and in a burner rig at Mach 0.3. The results were correlated to the type and content of stabilizer with special emphasis on phase analyses.

Evaporation of new ceramic compositions necessitates special precautions because the vapor pressures of the components may differ too much. A new dual-source evaporation coater allows the production of these innovative TBCs with close control of chemistry. The potential of the equipment will be discussed.

Commentary by Dr. Valentin Fuster
1996;():V005T12A012. doi:10.1115/96-GT-493.
FREE TO VIEW

From a cost point of view SNECMA has found DS columnar grain manufacturing technology to be highly attractive compared to single crystal. CM 186 LC alloy exhibits enhanced mechanical and environmental properties and temperature capability compared to MAR M 200 Hf alloy; these properties are close to first generation single crystal alloys up to 982°C (1800°F). The alloy is shown to be amenable to various coating and brazing high temperature processes. The longer term creep-rupture/phase stability data base on the alloy has now been extended out to 8300 hours at 1038°C (1900°F). Castings for engine test have been produced using CM 186 LC alloy.

Topics: Alloys , Airfoils
Commentary by Dr. Valentin Fuster
1996;():V005T12A013. doi:10.1115/96-GT-499.
FREE TO VIEW

The rainbow test was performed to evaluate the effects of cooling hole design modifications and the MCrAlY coating deterioration with and without top aluminide. The cooling design modifications considered in this program include a) camberline cooling hole design, b) peripheral cooling hole design with a turbulated leading edge hole and c) peripheral cooling hole design with all turbulated holes. Following approximately 16,700 hours of operation, one blade from each category was removed for metallurgical evaluation. The MCrAlY coating on the camberline cooled blade without the top aluminide and coat had degraded the most at the blade tip region and along a portion of the leading edge. In some regions, the coating on the leading edge of the blade was oxidized through thickness. However, no oxidation of coating was noted on the leading edge of the blade with the MCrAlY with top aluminide coating. The results show that the top aluminide coating and the cooling hole design modifications extend service life of the MCrAlY coating. The coating deterioration results on these blades are correlated with the β NiAl depleted zone in the coating and the oxide thickness at the cooling hole surfaces. Variations of coating degradation and cooling hole surface oxidation among the blades are discussed in terms of estimated service temperatures measured from γ′ particle size. The cooling hole design modifications significantly lower metal temperature and extend service life of the MCrAlY coating.

Topics: Coatings , Corrosion
Commentary by Dr. Valentin Fuster
1996;():V005T12A014. doi:10.1115/96-GT-500.
FREE TO VIEW

This study examined the ability of two High Velocity Oxyfuel (HVOF) systems, the Hobart Tafa JP 5000 and the Sulzer Metco DJ 2600 Hybrid, to apply MCrAlY coatings. The results were compared to a Low Pressure Plasma Spray (LPPS) applied material of the same chemistry. Each system is described and the optimum parameters are given. Also, an analysis of the powders is included.

The results were metallurgically evaluated for porosity level, oxygen content and yttrium distribution. Static oxidation tests were also performed on the NiCoCrAlY coating applied by each system. Conclusions were based on the relative performance of the JP 5000 and DJ 2600 Hybrid coatings as compared to the LPPS baseline material.

Topics: Coatings
Commentary by Dr. Valentin Fuster
1996;():V005T12A015. doi:10.1115/96-GT-510.
FREE TO VIEW

The row 1 blade from a heavy industrial gas turbine has been examined after ∼21,000 fired hours to assess the level and nature of the damage originating from the cooling holes. This consisted primarily of creep cracking/voiding coupled with deep internal oxidation/nitridation.

The difficulties the manufacturer had in addressing the problem are discussed within the context of component assessment methodologies and the ‘new’ technologies which are being incorporatd into new unit types, but for which little service experience has yet been generated.

Commentary by Dr. Valentin Fuster
1996;():V005T12A016. doi:10.1115/96-GT-519.
FREE TO VIEW

Single crystal superalloys had greatly improved cyclic oxidation resistance when their sulfur content was reduced from impurity levels, typically 5–10 ppmw in past years, down to 1 ppmw or below currently. Excellent alumina scale adhesion has been documented for PWA 1480, PWA 1484, Rene’N5, Rene’N6, and CMSX 4, all without reactive element (Y) additions. Hydrogen annealing was used for effective desulfurization to below 0.1 ppmw, as well as for achieving controlled intermediate levels. This paper summarizes the direct relationship between cyclic oxidation behavior and sulfur content. An adhesion criterion has been proposed based on the concentration of sulfur needed to initiate spallation due to a monolayer of interfacial segregation. This suggests that a level down to ∼0.2 ppmw would be needed to maximize adhesion for a 1 mm thick sample. It is in reasonable agreement with the experimental results.

Commentary by Dr. Valentin Fuster
1996;():V005T12A017. doi:10.1115/96-GT-520.
FREE TO VIEW

Very little property data on this common turbine blade alloy has been published. As longer hours of service are accumulated, maintenance considerations such as developing optimum component life strategies and repair processes become important. The lack of specific material data hampers the effort of users and repair facilities to achieve optimum service from this alloy.

This study measured some of the basic mechanical and metallurgical characteristics of this poly-crystalline nickel base superalloy. Tensile and short term creep rupture properties as well as microstructural and fracture characteristics are presented. Both the as-heat-treated and thermally exposed characteristics at two different temperatures are examined.

Topics: Alloys
Commentary by Dr. Valentin Fuster
1996;():V005T12A018. doi:10.1115/96-GT-521.
FREE TO VIEW

The feasibility of processing less-expensive alternative coatings to platinum aluminide was examined. Three approaches were followed: 1) enhancement of nickel-aluminide coatings by application of sol-gel derived two-phase-glass (TPG) overlayers, 2) evaluation of TPG coatings on bare IN 738LC, and 3) substitution of Pt with a less expensive platinum group metal (palladium). Accordingly, IN 738LC coupons were tested with several coatings including TPG, aluminide coatings (platinum aluminide, palladium aluminide, and conventional nickel aluminide), and TPG overlayers on the aluminide coatings. Isothermal-oxidation, cyclic-oxidation, and hot-corrosion tests were conducted at 900°C for 500 hours to evaluate the coatings. The results showed that the TPG by itself provided superior protection compared to the platinum-aluminide coatings under both oxidation and hot-corrosion conditions. The TPG coating also showed promise as an overcoat on aluminide coatings.

Topics: Coatings , Glass , Platinum
Commentary by Dr. Valentin Fuster
1996;():V005T12A019. doi:10.1115/96-GT-525.
FREE TO VIEW

MCrAlY (M=Co, Ni) coatings have long been used in gas turbines to protect hot section components from hot corrosion and oxidation. Their complex chemistries, along with the thickness of the deposits, makes them reliable over long periods of time in engines used in power generation and in aircraft applications. A variety of techniques such as Electron Beam Physical Vapor Deposition (EBPVD), Low Pressure Plasma Spray (LPPS), and Argon Shrouded Plasma Spray have been used to deposit MCrAlYs over the years. Until recently, OEM’s have been reluctant to consider alternate deposition methods, but the prospect of lower cost, high quality coatings has merited reconsideration of some long-standing assumptions. In recent efforts, Howmet Thermatech has gained production approval to deposit a proprietary MCrAlY composition on an OEM’s latest generation gas turbine engine components using a TAFA JP-5000 High Pressure/High Velocity Oxygen Fuel (HP/HVOF) coating system. Tensile bond and hardness data are presented which substantiate the equivalence of this coating to its VPX® forerunner. Photomicrographs illustrate the levels of porosity, oxide content and interface quality currently in revenue service. Unpublished cost data from an established coating supplier reveal that MCrAlYs can be deposited by HP/HVOF at a fraction of the cost of VPX® alternatives.

Topics: Coatings , Turbines
Commentary by Dr. Valentin Fuster

Ceramics

1996;():V005T13A001. doi:10.1115/96-GT-232.
FREE TO VIEW

At the Institute for Thermal Turbomachinery, University of Karlsruhe (ITS) a new methodology for designing thermally loaded ceramic components based on numerical analyses of the thermal and structural behaviour is developed. Major issues are an effective optimization of the geometry and an assessment of the load.

To reduce thermally induced stresses in shell structures exposed to hot gas flow a precise description of the major influencing parameters is required. Based on these relations a thermal optimization of the component is possible. Moreover, the compensation of thermal strains and the adjustment of local stiffnesses for reducing the tensile stresses and the failure probability has to be considered.

In designing a first stage vane of a stationary gas turbine the proposed guidelines are verified. The advantages of a systematic approach are demonstrated. The multi-purpose numerical procedures using the guidelines allow an adjustment of the shape to thermally and mechanically induced loads and offer new possibilities to meet the specific demands of ceramic materials.

Topics: Ceramics , Nozzles
Commentary by Dr. Valentin Fuster
1996;():V005T13A002. doi:10.1115/96-GT-269.
FREE TO VIEW

A critical phase in the life-cycle of a material is the transition from the development stage to that of design and industrial application. Advanced technical ceramics are currently in this phase and their future market penetration and acceptance as engineering materials critically depend on the availability of a robust set of standards for test methods which allow one to determine their mechanical and thermal properties in a reliable and reproducible way. In Europe, standardisation on continuous fibre reinforced ceramic matrix composites (CFCCs) officially started in 1989, when CEN TC 184 on Advanced Technical Ceramics was launched. The scope of TC 184 covers testing methods, and no efforts are paid to the development of material and product standards as yet. The work programme of TC 184 in the area of CFCCs, aimed at the establishment of testing methods for the generation of reliable data for design purposes, is reviewed.

Commentary by Dr. Valentin Fuster
1996;():V005T13A003. doi:10.1115/96-GT-270.
FREE TO VIEW

Monolithic and composite advanced ceramics have reached sufficient levels of material development to warrant serious consideration for applications in advanced heat engines. These applications require optimum material behavior with physical and mechanical property reproducibility, component reliability, and well-defined methods of data treatment and materials analysis. As new materials are introduced into the market place, these issues are best dealt with via standard methods. The primary standards writing organization in the U.S. is the American Society for Testing and Materials (ASTM), a private, nonprofit corporation which relies upon the voluntary cooperation of industry, government, and academe to develop standards by consensus. ASTM Committee C28 “Advanced Ceramics” has been active since 1986 and currently has 16 standards “on the books” with 17 standards in the balloting process. Overviews of the five subcommittees of C28 are presented. Accomplishments to date are discussed, as well as future activities, including a brief summary of joint cooperative efforts with international standards formulating organizations.

Commentary by Dr. Valentin Fuster
1996;():V005T13A004. doi:10.1115/96-GT-271.
FREE TO VIEW

This paper presents the results of a long term program initiated in December of 1984 to investigate the effects of environment (temperature, atmosphere, and stress) on the mechanical behavior of eight Si3N4 and three SiC ceramics being considered for heat engine applications. Microstructure, chemistry, and physical properties were determined. The mechanical behavior of these materials was investigated from room temperature to 1400°C by employing tests for flexural and tensile strength, dynamic, static and cyclic fatigue, and fracture toughness. The results obtained from these evaluations showed that the thermal mechanical behavior was quite varied, depending on the composition and processing methods employed. Batch to batch differences were also found to cause variances in the property values measured. Insights gained from this work about the failure mechanisms and potential service life are also discussed.

Commentary by Dr. Valentin Fuster
1996;():V005T13A005. doi:10.1115/96-GT-281.
FREE TO VIEW

A method for measuring surface temperature of semitransparent materials was developed and used to measure temperature gradients through a yttria-stabilized zirconia thermal barrier coating (TBC) heated on the front surface with a natural gas flame. A Pt stripe was deposited on the TBC surface, and its temperature was measured using infrared pyrometry at 10.6 μm. A model was developed to determine the radiant intensity emitted from a TBC on a metal substrate with a thermal gradient. The radiant intensity from TBCs was measured as a function of thermal gradient and compared with results of the model.

Commentary by Dr. Valentin Fuster
1996;():V005T13A006. doi:10.1115/96-GT-282.
FREE TO VIEW

Thermal barrier coatings (TBCs) applied to the hot gas components of turbine engines lead to enhanced fuel efficiency and component reliability. Understanding the mechanisms which control the thermal transport behavior of the TBCs is of primary importance. Electron beam-physical vapor deposition (EB-PVD) and air plasma spraying (APS) are the two most commonly used coating techniques. These techniques produce coatings with unique microstructures which control their performance and stability. The density of the APS coatings was controlled by varying the spray parameters. The low density APS yttria-partially stabilized zirconia (yttria-PSZ) coatings yielded a thermal conductivity that is lower than both the high density APS coatings and the EB-PVD coatings. The thermal aging of both fully and partially stabilized zirconia are compared. The thermal conductivity of the coatings permanently increases upon exposure to high temperatures. These increases are attributed to microstructural changes within the coatings. This increase in thermal conductivity can be modeled using a relationship which depends on both the temperature and time of exposure. Although the EB-PVD coatings are less susceptible to thermal aging effects, results suggest that they typically have a higher thermal conductivity than APS coatings before thermal aging. The increases in thermal conductivity due to thermal aging for plasma sprayed partially stabilized zirconia have been found to be less than for plasma sprayed fully stabilized zirconia coatings.

Commentary by Dr. Valentin Fuster
1996;():V005T13A007. doi:10.1115/96-GT-283.
FREE TO VIEW

Thermal barrier coatings, (TBCs) play a crucial role in the performance of advanced aircraft gas turbine engines that power the commercial and military fleets. The same technology is currently being applied to the industrial gas turbines. However the task is more challenging. The environment of the industrial gas turbine is far more demanding. Studies are in progress delineating the relationships between time, temperature and the sinterability of candidate ceramics for use in industrial gas turbine engines. Typical sintering aids include the oxides and alkali salts of silicon, iron, magnesium and calcium. Other experiments focus on the role of the alkali compounds as they affect the mechanical and chemical properties of candidate materials.

Commentary by Dr. Valentin Fuster
1996;():V005T13A008. doi:10.1115/96-GT-284.
FREE TO VIEW

Over the last twenty years, significant performance improvements of turbojet engines have been achieved by optimizing engine thermodynamic cycle along with the introduction of new materials providing higher temperature capability and weight reduction.

Metal Matrix Composites (MMC) and Ceramic Matrix Composites (CMC) are candidate material systems to meet the required thrust-to-weight ratio of 15 or higher.

Continuous fiber reinforced ceramic composites, which have been developed by SEP for more than 15 years for thermostructural applications in oxidative environment, aim at increased operating temperature over superalloys and intermetallic alloys.

This paper is a review of the main CMC component demonstrations performed by SEP over the last 10 years for turbojet engines along with an analysis of consequences on materials development and design methodology.

The development status of a new thermostructural material specifically developed for turbojet environment with the prospect of higher design stress allowables and longer operating life at high temperature is presented.

Commentary by Dr. Valentin Fuster
1996;():V005T13A009. doi:10.1115/96-GT-285.
FREE TO VIEW

Thermal barrier coating (TBC) spallation on power generation combustors was compared with TBC spallation observed both in military turboshaft engines, and in commercial turboprop engines. In each case, irrespective of operating conditions or geographic location, spallation was linked to the presence and infiltration of high temperature molten phases of similar composition. Electron microprobe analysis found that, from all the possible oxides available in the external environment, only CaO, MgO, Al2O3 and SiO2 (CMAS) are incorporated in the molten phase that infiltrates the TBC microstructure. Iron and nickel oxides from turbine components and zirconia and yttria from the TBC were also found in varying amounts in the molten phase.

Melting of environmental deposits in conjunction with infiltration was found to result in: densification of the TBC, an increase in its Young’s modulus and an increase in the room temperature compressive stress in the TBC. Delamination of the TBC during thermal cycling is, thereby, attributed to changes in the mechanical properties and associated changes in the stress state of the coating due to infiltration of the environmental deposit.

Commentary by Dr. Valentin Fuster
1996;():V005T13A010. doi:10.1115/96-GT-286.
FREE TO VIEW

Thermogravimetric methods for evaluating bond coat oxidation in plasma-sprayed thermal barrier coating (TBC) systems were assessed by high-temperature testing of TBC systems with air plasma-sprayed (APS) Ni-22Cr-10Al-1Y bond coatings and yttria-stabilized zirconia top coatings. High-mass thermogravimetric analysis (at 1150°C) was used to measure bond coat oxidation kinetics. Furnace cycling was used to evaluate APS TBC durability. This paper describes the experimental methods and relative oxidation kinetics of the various specimen types. Characterization of the APS TBCs and their reaction products are discussed.

Commentary by Dr. Valentin Fuster
1996;():V005T13A011. doi:10.1115/96-GT-287.
FREE TO VIEW

The European EUREKA project, EU 209, otherwise known as AGATA (Advanced Gas Turbine for Automobiles), is a programme dedicated to the development of three critical ceramic components — a catalytic combustor, a radial turbine wheel and a static heat exchanger — for a 60 kW turbogenerator in an hybrid electric vehicle. These three components, which are of critical importance to the achievement of low emissions and high efficiency, will be designed, developed, manufactured and tested as part of a full scale feasibility study. AGATA is a joint project conducted by eight commercial companies and four research institutes in France and Sweden. Silicon nitride ceramics play an important role both in the development of the catalytic combustor and for the radial turbine wheel. This paper outlines the main results of the AGATA project with special emphasis to the development of Si3N4 combustor and turbine wheel.

Commentary by Dr. Valentin Fuster
1996;():V005T13A012. doi:10.1115/96-GT-295.
FREE TO VIEW

NGK Insulators, Ltd. (NGK) fabricates ceramic turbine rotors and nozzle components for the Automotive Ceramic Gas Turbine development program (CGT program), conducted by the Petroleum Energy Center (PEC) of Japan. Both the rotor and nozzle components are made of silicon nitride. The turbine rotor is a radial type having a 127 mm tip diameter, and the segment type nozzle is formed integrally with either three or four vanes and shrouds. This paper discusses various processes of fabricating the ceramic components, dependence of the fabrication processes on dimensional accuracy, and the performance of each of the processes.

Commentary by Dr. Valentin Fuster
1996;():V005T13A013. doi:10.1115/96-GT-296.
FREE TO VIEW

MAN Technologie AG has been engaged in the field of ceramic matrix composites (C/SiC, SiC/SiC) for approximately 10 years. Two manufacturing methods have been developed with respect to economical series production:

- Chemical Vapour Infiltration Process (CVI) with temperature and pressure gradients

- Liquid Polymer Impregnation (LPI) and Pyrolysis

The advantage of these methods is an essential reduction of process time compared to the conventional isothermal-isobaric CVI method, with equivalent material performance. The main material characterization data, such as strength/strain, thermal shock, dynamic fatigue and high temperature behaviour are described. Several components under development or in low volume preseries production are presented.

Commentary by Dr. Valentin Fuster
1996;():V005T13A014. doi:10.1115/96-GT-318.
FREE TO VIEW

Three subscale, cylindrical combustors were rig tested on natural gas at typical industrial gas turbine operating conditions. The intent of the testing was to determine the effect of combustor liner cooling on NOx and CO emissions.

In order of decreasing liner cooling, a metal louvre-cooled combustor, a metal effusion-cooled combustor, and a backside-cooled ceramic (CFCC) combustor were evaluated. The three combustors were tested using the same lean-premixed fuel injector.

Testing showed that reduced liner cooling produced lower CO emissions as reaction quenching near the liner wall was reduced. A reduction in CO emissions allows a reoptimization of the combustor air flow distribution to yield lower NOx emissions.

Commentary by Dr. Valentin Fuster
1996;():V005T13A015. doi:10.1115/96-GT-319.
FREE TO VIEW

The U.S. Department of Energy and most U.S. manufacturers of stationary gas turbines are participating in a major national effort to develop advanced turbine systems (ATS). The ATS program will achieve ultra-high efficiencies, environmental superiority, and cost competitiveness compared with current combustion turbine systems. A major factor in the improved efficiencies of simple cycle ATS gas turbines will be higher operating temperatures than current engines. These temperatures strain the limits of metallic alloy and flow-path cooling technologies.

Ceramics materials offer a potential alternative to cooled turbine alloys for ATS turbines due to higher melting points than metallics. This paper evaluates ceramics technology and plant economic issues for ATS industrial turbine systems. A program with the objective of demonstrating first-stage ceramic vanes in a commercial industrial turbine is also described.

Topics: Ceramics , Turbines
Commentary by Dr. Valentin Fuster
1996;():V005T13A016. doi:10.1115/96-GT-320.
FREE TO VIEW

To support the advanced ceramics infrastructure in terms of ability to acquire reliable property and performance data, and to define tests for specifications, the development of testing standards has had priority in Europe through CEN (Comité Européen de Normalisation) over the past six years. The maintenance of close links with other standardisation bodies, including ASTM, JIS, and now ISO, has been important, and has been achieved through the international ceramics community. This has minimized technical differences between standards produced in different parts of the world. A review of progress in the preparation of formal CEN standards for monolithic ceramics is presented, which comprises a programme with a target of about eighty standards to be available by 1998. Some of the outstanding problems and requirements are identified. The role of pre-standardization research programmes and interlaboratory comparison is emphasized, including national and international programmes within the European Union. In particular, collaborative work under the auspices of VAMAS has proved to be a valuable method of obtaining internationally agreed positions on some of the technical issues.

Topics: Ceramics
Commentary by Dr. Valentin Fuster
1996;():V005T13A017. doi:10.1115/96-GT-321.
FREE TO VIEW

In response to a proposal of “Early-Stage Standardization” by the ISO/IEC Presidents’ Advisory Board on Technological Trends, the Technical Committee ISO/TC 206 on Fine Ceramics was established in 1992. The scope of ISO/TC 206 is a standardization in the field of fine ceramic materials and products in all forms: powders, monoliths, coatings and composites, intended for specific functional applications including mechanical, thermal, chemical, electrical, magnetic, optical and combinations thereof. The term “fine ceramics” is defined as “a highly engineered, high performance, predominantly nonmetallic, inorganic material having specific functional attributes”. A summary overview is given of membership, organization structure, work program and future work of ISO/TC 206.

Topics: Ceramics
Commentary by Dr. Valentin Fuster
1996;():V005T13A018. doi:10.1115/96-GT-341.
FREE TO VIEW

A technique to achieve stable and uniform uniaxial compression is offered for creep testing of advanced ceramic materials at elevated temperatures, using an innovative self-aligning load-train assembly. Excellent load-train alignment is attributed to the inherent ability of a unique hydraulic universal coupler to maintain self-aligning. Details of key elements, design concept, and principles of operation of the self-aligning coupler are described. A method of alignment verification using a strain-gaged specimen is then discussed. Results of verification tests indicate that bending below 1.5% is routinely achievable with the use of the load-train system. A successful compression creep test is demonstrated using a dumb-bell shaped silicon nitride specimen tested at 1300°C for a period in excess of 4000 h.

Commentary by Dr. Valentin Fuster
1996;():V005T13A019. doi:10.1115/96-GT-342.
FREE TO VIEW

The solution of fatigue strength as a function of preloading in dynamic fatigue testing was obtained analytically and numerically. The effect of preloading on dynamic fatigue strength decreases with increasing fatigue parameter (n), and for n ≥ 20 the effect is negligible up to a preloading of 90 %. The solution was verified by dynamic fatigue experiments conducted with soda-lime glass and alumina specimens in room-temperature distilled water. This result showed that one can apply a preloading corresponding up to 90 % of fatigue strength for most glass and ceramic materials, resulting in a dramatic saving of testing time in dynamic fatigue testing. The key feature that makes this technique feasible is that most of the slow crack growth under dynamic fatigue loading occurs close to failure time where a dynamic fatigue strength is defined.

Commentary by Dr. Valentin Fuster
1996;():V005T13A020. doi:10.1115/96-GT-346.
FREE TO VIEW

A ceramic combustor for a 1500°C, 20MW class industrial gas turbine was developed and tested. This combustor has a hybrid ceramic/metal structure. To improve the durability of the combustor, the ceramic parts were made of silicon carbide (SiC), which has excellent oxidation resistance under high temperature conditions as compared to silicon nitride (Si3N4) although the fracture toughness of SiC is lower than that of Si3N4. Structural improvements to allow the use of materials with low fracture toughness, were made to the fastening structure of the ceramic parts. Also the combustion design of the combustor was improved. Combustor tests using low BTU gaseous fuel of a composition that simulated coal gas were carried out under high pressure. The test results demonstrated that the structural improvements were effective because the ceramic parts exhibited no damage even in the fuel cutoff tests from rated load conditions. It also indicated that the combustion efficiency was almost 100% even under part load conditions.

Commentary by Dr. Valentin Fuster
1996;():V005T13A021. doi:10.1115/96-GT-347.
FREE TO VIEW

The development and refinement of high performance silicon nitride structural ceramic materials over the last five years is leading to evaluation and implementation of components in aircraft, space, industrial, and automotive turbomachinery applications. Current material properties, status of component fabrication technologies, and status of applications being evaluated and commercialized at AlliedSignal Ceramic Components is discussed. Currently achievable properties of in-situ reinforced monolithic silicon nitride materials are presented. The development and status of component forming processes is also discussed, including their potential as manufacturing processes. The processes discussed include slipcasting, green machining, gelcasting, and injection molding. Finally, status of silicon nitride component fabrication and evaluation in a number of applications is discussed, including nozzle, blade, and wheel components for the U.S. Department of Energy (DOE) automotive turbine technology programs, the DOE Advanced Turbine Systems industrial turbine technology program, and a NASA-funded program to develop advanced rocket engine turbopumps.

Commentary by Dr. Valentin Fuster
1996;():V005T13A022. doi:10.1115/96-GT-348.
FREE TO VIEW

We are conducting the development of ceramic matrix composites (CMC) and components made of CMC for a 100 kW automotive ceramic gas turbine (CGT) as shown in Fig.1. When compared to monolithic ceramics (MC), CMC that we have developed demonstrate superior strength characteristics in terms of resistance to particle impact and thermal shock. We have conducted evaluation tests on the strength of CMC components in which MC such as silicon nitride and silicon carbide were used as a reference for comparison with CMC in the same testing process as employed for components made of MC such as silicon nitride and silicon carbide. It was confirmed that actual components made of CMC realized approximately the same strength as the test pieces. Furthermore, some CMC components have already passed screening tests that evaluated the strength of the components. It was therefore confirmed that the potential exists for the possibility of testing these components in high temperature assembly tests and engine tests.

Commentary by Dr. Valentin Fuster
1996;():V005T13A023. doi:10.1115/96-GT-385.
FREE TO VIEW

The tensile creep-rupture performance of a commercially available gas pressure sintered silicon nitride (Si3N4) and a sintered silicon carbide (SiC) is examined at 1038, 1150, and 1350°C. These two ceramic materials are candidates for nozzles and combustor tiles that are to be retrofitted in land-based gas turbine engines, and interest exists to investigate their high temperature mechanical performance over service-times up to, and in excess of, 10000 hours (≈ 14 months). To achieve lifetimes approaching 10000 hours for the candidate Si3N4 ceramic, it was found (or it was estimated based on ongoing test data) that a static tensile stress of 300 MPa at 1038 and 1150°C, and a stress of 125 MPa at 1350°C cannot be exceeded. For the SiC ceramic, it was estimated from ongoing test data that a static tensile stress of 300 MPa at 1038°C, 250 MPa at 1150°C, and 180 MPa at 1350°C cannot be exceeded. The creep-stress exponents for this Si3N4 were determined to be 33, 17, and 8 for 1038, 1150, and 1350°C, respectively. The fatigue-stress exponents for the Si3N4 were found to be equivalent to the creep exponents, suggesting that the fatigue mechanism that ultimately causes fracture is controlled and related to the creep mechanisms. Little success was experienced at generating failures in the SiC after several decades of time through exposure to appropriate tensile stress; it was typically observed that if failure did not occur on loading, then the SiC specimens most often did not creep-rupture. However, creep-stress exponents for the SiC were determined to be 57, 27, and 11 for 1038, 1150, and 1350°C, respectively. For SiC, the fatigue-stress exponents did not correlate as well with creep-stress exponents. Failures that occurred in the SiC were a result of slow crack growth that initiated from the specimen’s surface.

Topics: Creep , Gas turbines
Commentary by Dr. Valentin Fuster
1996;():V005T13A024. doi:10.1115/96-GT-446.
FREE TO VIEW

Kyocera has been developing various ceramic components for gas turbines under the Ceramic Gas Turbine (CGT) Project funded by the Japanese Government. This project has set a turbine inlet temperature (TIT) of 1350°C as a final target.

For 1350°C TIT, we have developed a new silicon nitride material SN281, which has high stress rupture strength at elevated temperatures up to 1500°C. This material has excellent oxidation resistance as well.

We have also developed improved sintering and inspection technologies for the use of SN281 as engine components. We are able to fabricate rotors and nozzles of the gas generator turbine (GGT) in good agreement with design geometry requirements, by optimizing sintering conditions. Small defects were also successfully detected by microfocus X-ray radiography. The SN281 rotors have attained 120% of design rotating speed at room temperature.

Commentary by Dr. Valentin Fuster
1996;():V005T13A025. doi:10.1115/96-GT-458.
FREE TO VIEW

As part of the U.S. Department of Energy Advanced Turbine Systems Program, the performance of Chromalloy RT122, RT122 over RT69 and the Howmet 150L bond coats were evaluated for use in the next generation of Westinghouse combustion turbines. Air plasma sprayed and electron beam physical vapor deposition 8% yttria stabilized zirconia thermal barrier coatings were applied to the bond coats. The coating systems were evaluated in air at 2102°F (1150°C), cooling to room temperature once per day. The life-limiting failure mode in both air plasma sprayed (APS) and electron beam - physical vapor deposition (EB-PVD) coating systems is the oxidation of the bond coat. The coating life is related to the growth rate and morphology of the thermally grown oxide. The superior performance of RT122 on MarM-002, the duplex bond coat system of RT122 over RT69 on MarM-002 and Howmet 150L on MarM-002 can be related to the development of a uniform, slow growing oxide scale. The development of a non-uniform oxidation front contributes to the reduced life of RT122 on IN-939 and CM-247.

Topics: Coatings
Commentary by Dr. Valentin Fuster
1996;():V005T13A026. doi:10.1115/96-GT-459.
FREE TO VIEW

Employing ceramic materials for the critical components of industrial gas turbines is anticipated to improve the thermal efficiency of power plants. We have developed a first stage ceramic stator vane for a 1500°, 20MW class industrial gas turbine by improving our original one for a 1300°C class gas turbine. Our stator vane has a hybrid ceramic/metal structure composed of a ceramic shell, a metal core and a heat insulating layer. This composition increases the strength of the brittle ceramic parts and reduces the amount of cooling air. To improve the durability and reliability of the stator vane in 1500°C combustion gas, the ceramic shell uses silicon carbide instead of silicon nitride, and its configuration is improved. Furthermore, we use an internal cooling system to control the temperature of the metal core. Thermal loading cascade tests are conducted to prove the reliability and cooling performance of the stator vane.

Commentary by Dr. Valentin Fuster
1996;():V005T13A027. doi:10.1115/96-GT-460.
FREE TO VIEW

The goal of the Ceramic Stationary Gas Turbine (CSGT) Development Program, under the sponsorship of the United States Department of Energy (DOE), Office of Industrial Technologies (OIT), is to improve the performance (fuel efficiency, output power, exhaust emissions) of stationary gas turbines in cogeneration through the selective replacement of hot section components with ceramic parts. The program, currently in Phase II focuses on detailed engine and component design, ceramic component fabrication and testing, establishment of a long term materials property data base, the development of supporting nondestructive evaluation (NDE) technologies, and the application of ceramic component life prediction. A 4000 hr engine field test is planned for Phase III of the program. This paper summarizes progress from January 1995 through January 1996.

First generation designs of the primary ceramic components (first stage blades and nozzles, combustor liners) for the program engine, the Solar Centaur 50S, and of the secondary metallic components interfacing with the ceramic parts were completed. The fabrication of several components has been completed as well. These components were evaluated in rigs and the Centaur 50S test engine. NTI64 (Norton Advanced Ceramics) and GN-10 (AlliedSignal Ceramic Components) silicon nitride dovetail blades were cold and hot spin tested and engine tested at the baseline nominal turbine rotor inlet temperature (TRIT) of 1010°C. Full scale SiC/SiC continuous fiber-reinforced ceramic matrix composite (CFCC) liners (B.F. Goodrich Aerospace) were also rig tested and engine tested at the nominal baseline TRIT of 1010°C. One of the engine tests, incorporating both the GN-10 blades and the full scale SiC/SiC CFCC liners, was performed for 21.5 hrs (16 hrs at 100% load) with six start/stop cycles. A cumulative 24.5 hrs of engine testing was performed at the end of January, 1996. The ceramic components were in good condition following completion of the testing.

Subscale Hexoloy® SA silicon carbide (Carborundum) and enhanced SiC/SiC CFCC (DuPont Lanxide Composites) and Al2O3/Al2O3 CFCC (Babcock & Wilcox) combustor liners were tested to evaluate mechanical attachment, durability and/or emissions reduction potential. The enhanced SiC/SiC CFCC of DuPont Lanxide Composites demonstrated superior durability in subscale combustor testing and this material was subsequently selected for the fabrication of full scale combustor liners for final engine rig testing in Phase II and field testing in Phase III of the program. Enhanced SiC/SiC CFCC liners also showed significantly reduced emissions of NOx and CO when compared with conventionally cooled subscale metallic liners. This observation is believed to apply generally to “hot wall” combustor substrates. The emissions results for the enhanced SiC/SiC CFCC liners were paralleled by similar emissions levels of NOx and CO monitored during engine testing with B.F. Goodrich Aerospace SiC/SiC CFCC combustor liners. NOx levels below 25 ppmv and CO levels below 10 ppmv were measured during the engine testing.

Short term (1,000 hrs) creep testing of candidate ceramic materials under approximate nozzle “hot spot” conditions was completed and long term (5000–10,000 hrs) creep testing is in progress. The selected nozzle material, SN-88 silicon nitride, has survived over 5,500 hrs at 1288°C and 186 MPa stress at the end of January, 1996.

Commentary by Dr. Valentin Fuster
1996;():V005T13A028. doi:10.1115/96-GT-532.
FREE TO VIEW

The high temperature fatigue behaviour of two 2D reinforced ceramic matrix composites (CMCs) Is studied under high vacuum conditions. The mechanical loads imposed result in matrix cracking upon first loading, so that continued cyclic loading results in progressive interfacial debonding and/or matrix crack multiplication, as well as fibre failure. In order to investigate whether the fatigue life is mainly governed by time-dependent creep or by cyclically induced fatigue damage, a range of frequencies and two stress ratios are explored in stress controlled fatigue tests. The results obtained indicate that under pulsating fatigue (positive stress ratios or tension-tension) the material response is affected by both creep and fatigue mechanisms. The cyclic damage component gains in relative importance with increasing test frequency. Under reversed loading conditions (negative stress ratios), and depending on the creep strength mismatch between the fibres and the matrix, the time-dependent damage component can be largely suppressed, and the composite fatigue behaviour can become close to purely cycle-dependent. In both cases and for both composites fatigue failure is triggered by fibre failure.

Commentary by Dr. Valentin Fuster
1996;():V005T13A029. doi:10.1115/96-GT-533.
FREE TO VIEW

Currently available Ceramic Matrix Composites (CMCs) have very low stress carrying capability if they are to achieve the service life required for application in gas turbine engines. As such, they are most likely to find their first applications in non-structural components with low mechanical loads, where the majority of the stress is thermally induced. The thermal cycling experienced in gas turbine engines, coupled with the necessary interfaces with surrounding metal components and other geometric features, means that these thermal stresses are often localised, but in order to produce a valid component design they may significantly exceed the maximum design stress.

The aim of this paper is to discuss the implications for the life of the component of these excess stresses. This will cover the mechanisms for the propagation of localised damage in a strain controlled environment, and the effect of this damage on the thermal conductivity and hence on the induced thermal gradients and thermal strains. Strains corresponding to stresses considerably above the normally accepted design stress can be sustained for a considerable number of cycles, but the influence of extended time periods with damage at elevated temperatures remains unexplored.

Commentary by Dr. Valentin Fuster

Structures and Dynamics

1996;():V005T14A001. doi:10.1115/96-GT-012.
FREE TO VIEW

The shaft misalignment, even being a common fault in rotating machinery, is not sufficiently studied. The present work addresses effects of misalignment in rotating machinery. An attempt to give a theoretical model for a rotor-coupling-bearing system has been done. The rotor-bearing system including the flexible coupling is modelled using the finite elements. The reaction forces and moments developed due to flexible coupling misalignment both for parallel and angular are derived and introduced in the model. Vibration analyses such as eigen value analysis and unbalance response are carried out for the rotor system with misaligned shafts.

Commentary by Dr. Valentin Fuster
1996;():V005T14A002. doi:10.1115/96-GT-017.
FREE TO VIEW

An investigation of the influences of running speeds to the low speed balance of flexible rotors is made in this paper and shows that the balancing weights obtained at the balance speeds can be corrected and nearly equal to that at the operation speed. So the theory of low speed balancing of flexible rotors can be used not only for flexible rotors with natural properties independent of running speeds, but also for those with natural properties dependent on running speeds.

Topics: Rotors
Commentary by Dr. Valentin Fuster
1996;():V005T14A003. doi:10.1115/96-GT-018.
FREE TO VIEW

A 2-dimensional fluid model is developed to investigate the hydrodynamic forces exerted on a rotating impeller caused by the impeller-fluid-volute interaction in a centrifugal pump. In this model, the impeller periphery and the volute contour are replaced by a distribution of unsteady vortices. The impeller center is assumed to execute a whirling motion about the rotor center. This is an improvement of the earlier quasisteady flow model of Colding-Jørgensen (1980) where the impeller was taken as a single vortex-source point. The forces can be presented as a sum of a steady and an unsteady part. The rotordynamic coefficients are deduced from the unsteady forces decomposed into radial and tangential components relative to the orbit described by the impeller center. In comparison to most of the theoretical and experimental results found in the literature, the model seems to give good prediction. It appears clearly from this analysis that, under certain operating conditions, the fluid forces on the impeller have a destabilizing effect on the pump rotor.

Commentary by Dr. Valentin Fuster
1996;():V005T14A004. doi:10.1115/96-GT-019.
FREE TO VIEW

An improvement is made to the design of epicyclic gearbox that limits the dynamic eccentricities of gearbox components. Gearboxes manufactured to the new design and tested, showed reduced levels of sub-synchronous vibrations at the turbine bearings. The in-situ balancing feature of the gearbox is also found to be very effective in reducing these vibrations further. The improved gearboxes not only assured the service life of the turbine bearings, but also increased the success rate of package testing the first time with substantial cost savings.

Commentary by Dr. Valentin Fuster
1996;():V005T14A005. doi:10.1115/96-GT-022.
FREE TO VIEW

The free vibration of rotating disk–blade coupled system is investigated by the Ritz method. Centrifugal effects due to rotation are taken into account for both of the disk and blades. The boundary and continuity conditions between the disk and blades are satisfied by means of artificial springs introduced at their joints, and the orthogonal polynomials generated by using the Gram–Schmidt process are employed as admissible functions for both of the disk and blades. Frequency parameters and mode shapes of vibration are obtained to investigate the vibration of the disk–blade coupled system.

Commentary by Dr. Valentin Fuster
1996;():V005T14A006. doi:10.1115/96-GT-023.
FREE TO VIEW

This paper describes a fast algorithm to obtain the steady state unbalance response of a multi-mode rotor supported on short squeeze film dampers (SFDs). The presented algorithm is developed based on planar modal theory. Undamped critical speed analysis is first performed to obtain the rotor critical speeds and their associated mode shapes. The modal analysis technique is then applied to the linear part of the rotor-SFD assembly to obtain the system differential equations. The rotor is assumed to execute circular centered orbits, hence all differential equations are reduced to algebraic ones. The resulting equations are manipulated algebraically to form a polynomial in rotor rotational speed. The roots of the polynomial are found and the full unbalance response is obtained. A conventional rotor is used to describe the developed algorithm numerically. Results show that the proposed algorithm gives accurate response in comparison to that obtained by integrating the system differential equations numerically. The great advantage of the proposed algorithm is the saving in the execution time which is extremely dramatic with respect to numerical integration, in addition to other advantages such as the possibility of obtaining all solutions occurring in regions of multiple steady state. Accuracy and speed of execution are quite advantageous regarding parametric studies on multi-mode rotors. These parametric studies can help in the optimization of SFDs design.

Topics: Dampers , Rotors
Commentary by Dr. Valentin Fuster
1996;():V005T14A007. doi:10.1115/96-GT-024.
FREE TO VIEW

A major challenge during the design process of a modern low aspect ratio high speed axial compressor is to find rotor blade geometries that meet both, aerodynamic and mechanical requirements. This paper deals with the mechanical design of a transonic compressor blade. In order to meet the mechanical requirements in a short development time, new methods were used: A numerical optimization tool and an optical blade vibration measurement method:

The numerical resonance tuning took advantage of a semi-automatic optimization technique, based on a Finite Element vibration anlysis tool. The intention was to find a geometry which has no critical resonances (with fundamental engine orders) within the operation range.

To verify the calculated blade natural frequencies and eigen-values standard shaker tests using a laser holography system were carried out. Blades under g-load in the running compressor were investigated with an in-house developed vibration measurement system. This system is able to measure frequencies and amplitudes of the rotor blade vibrations without blade instrumentation but small optical probes, mounted in the compressor casing. The measured resonance points are in good agreement with the predictions. All amplitudes are far below the blade fatigue limits.

Commentary by Dr. Valentin Fuster
1996;():V005T14A008. doi:10.1115/96-GT-026.
FREE TO VIEW

The paper presents a numerical study on turbine clocking using a stator/rotor/stator model configuration of equal pitches. The results of time accurate flow calculations for four clocking positions are compared. The unsteady, two dimensional Reynolds averaged equations are solved using a one-equation turbulence model by Spalart and Allmaras. In absence of a transition model, the flow is assumed to be fully turbulent.

In a first analysis of the unsteady data, emphasis has been put on the loss production of the clocked stator. A Fourier decomposition of the specific entropy at the stator entry and exit is discussed. It is shown that clocking determines the degree of interaction of a stator with the wake of another upstream stator for a given rotor speed. Further, clocking has little upstream influence.

Topics: Turbines
Commentary by Dr. Valentin Fuster
1996;():V005T14A009. doi:10.1115/96-GT-078.
FREE TO VIEW

Aeroelasticity phenomena are characterised by the interaction of fluid and structural domains, whose describing equations are nonlinear. Classical prediction methods are generally based on treating the two domains separately while integrated (or coupled) approaches link them via boundary conditions throughout the solution phase. In turbomachinery environments, the aeroelasticity problem is further compounded by the fact that blades vibrate with a relative phase with respect to each other, the value of which is not necessarily known. Using a 3D thin-layer Reynolds-averaged Navier-Stokes solver and a 3D structural model, various coupled and uncoupled flutter analysis methods are compared with particular emphasis on inter-blade phase angle. A typical fan geometry, the NASA Rotor 67 blade, was chosen as the test case since steady-flow measurements are available for this particular structure. Two flow conditions, near peak-efficiency and near stall, were investigated for inter-blade phase angles of −90°, 0°, 90° & 180°. The performance of the uncoupled analysis with shape correction was first compared with that of the uncoupled multi-passage analysis. A coupled multi-passage analysis was performed next in order to highlight the importance of fluid/structure interaction. It was found that significant natural frequency shifts could exist between the structural and aeroelastic modes of the system, which suggests that coupled analyses may be more appropriate for such cases.

Commentary by Dr. Valentin Fuster
1996;():V005T14A010. doi:10.1115/96-GT-079.
FREE TO VIEW

The transfer matrix–component mode synthesis has been developed for the analysis of critical speed, response to imbalance and rotordynamic optimal design of multi–spool rotor system. This method adopted the advantages of the transfer matrix method for the train structure and the component mode synthesis for reducing degrees of freedom. In this method, the whole system is divided into several subsystems at the boundary coordinates. The constrained vibration modes and the static deflection curves of the constrained rotor subsystems are analysed by the improved transfer matrix method. The whole system is connected together by the component mode synthesis in accordance with the coordinate transformation. Numerical examples show that this method is superior to the traditional transfer matrix method and the component mode synthesis by FEM. This method has been successfully used for the rotordynamic analysis and optimal design of the compressors and the gas turbine aeroengines.

Commentary by Dr. Valentin Fuster
1996;():V005T14A011. doi:10.1115/96-GT-081.
FREE TO VIEW

In order to verify an inverse design concept for modern compressor bladings, a subsonic compressor front stage with IGV was investigated.

One objective of the design was to survey the flow field in detail, with emphasis on 3D viscous and unsteady aspects of the flow. Therefore, the compressor was equipped with various steady and unsteady measurement techniques. Additionally, a compressor design was chosen that allows an extension up to three stages with regard to the investigation of multistage axial compressor flow behavior.

Test results of the steady measurements are discussed for IGV, rotor, and stator flow at design conditions as well as the overall stage performance. The measurements of the steady flow behavior confirm the expected design performance and show the high potential of the controlled diffusion airfoil concept. Only at the side walls near hub and casing there are some differences between design and measurement due to the complex three dimensional flow.

For the study of unsteady effects, detailed measurements using hot-wire probes, glue-on hot-films, and semiconductor pressure transducers were performed. All measurements are evaluated using the ensemble-average technique. The results show how the boundary layers of the inlet guide vanes and stator blades develop in a flow that is periodically disturbed by the rotor. Time-dependent pressure distributions at midspan of both stators are described. In addition, the unsteady pressure field at the casing above the rotor was investigated. The minimum wall pressure is located away from the blade suction surface. The effects of tip clearance flow on the performance are presented. The radial extent covers 15% span from the tip. At rotor exit, the unsteady pressure field and the time-dependent three-dimensional velocity vectors illustrate the salient features of the viscous flow associated with the rotor.

Commentary by Dr. Valentin Fuster
1996;():V005T14A012. doi:10.1115/96-GT-109.
FREE TO VIEW

The quest for higher performance engines in conjunction with the requirement for lower life cycle costs has resulted in stage configurations that are more susceptible to high cycle fatigue. One solution is the use of innovative approaches that introduce additional mechanical damping. The present paper describes an approach that may be used to assess the benefits of friction dampers located within internal cavities of a hollow structure.

The friction dampers used in this application are often relatively thin devices that, if unconstrained, have natural frequencies in the same range as the natural frequencies of the hollow airfoil. Consequently, the analytical approach that is developed is distinct in that it has to take into account the dynamic response of the damper and how it changes as the amplitude of the vibration increases. In this paper, results from the analytical model are compared with independently generated results from a time integration solution of a three mass test problem. Results from the analytical model are compared with experimental data in a companion paper.

Topics: Friction , Damping , Airfoils
Commentary by Dr. Valentin Fuster
1996;():V005T14A013. doi:10.1115/96-GT-110.
FREE TO VIEW

The use of hollow airfoils in turbomachinery applications, in particular fans and turbines, is an essential element in reducing the overall engine weight. However, state–of–the–art airfoil geometries are of low aspect ratio and exhibit unique characteristics associated with plate–like modes. These modes are characterized by a chordwise form of bending and high modal density within the engine operating speed range. These features combined with the mistuning effects resulting from manufacturing tolerances make accurate frequency and forced response predictions difficult and increase the potential for High Cycle Fatigue (HCF) durability problems. The present paper summarizes the results of an experimental test program on internal damping of hollow blade–like specimens. Friction damping is provided via sheet metal devices configured to fit within a hollow cavity with various levels of preload. The results of the investigation indicate that such devices can provide significant levels of damping provided the damper location and preload is optimized for the modes of concern. The transition of this concept to actual engine hardware would require further optimization with regard to wear effects and loss of preload particularly in applications where the preload is independent of rotational speed. Excellent agreement was achieved between the experimental results and the analytical predictions using a micro–slip friction damping model.

Topics: Friction , Damping , Airfoils
Commentary by Dr. Valentin Fuster
1996;():V005T14A014. doi:10.1115/96-GT-111.
FREE TO VIEW

Modal interaction refers to the way that the modes of a structure interact when its geometry and material properties are perturbed. The amount of interaction between the neighboring modes depends on the closeness of the natural frequencies, the mode shapes, and the magnitude and distribution of the perturbation. By formulating the structural eigenvalue problem as a normalized modal eigenvalue problem, it is shown that the amount of interaction in two modes can be simply characterized by six normalized modal parameters and the difference between the normalized frequencies. In this paper, the statistical behaviors of the normalized frequencies and modes are investigated based on a perturbation analysis. The results are independently verified by Monte Carlo simulations.

Topics: Eigenvalues
Commentary by Dr. Valentin Fuster
1996;():V005T14A015. doi:10.1115/96-GT-120.
FREE TO VIEW

Net radial loading arising from asymmetric pressure fields in the volutes of centrifugal pumps during off-design operation is well known and has been studied extensively. In order to achieve a marked improvement in overall efficiency in centrifugal gas compressors, vaneless volute diffusers are matched to specific impellers to yield improved performance over a wide application envelope. As observed in centrifugal pumps, nonuniform pressure distributions that develop during operation above and below the design flow create static radial loads on the rotor. In order to characterize these radial forces, a novel experimental measurement and post-processing technique is employed that yields both the magnitude and direction of the load by measuring the shaft centerline locus in the tilt-pad bearings. The method is applicable to any turbomachinery operating on fluid film radial bearings equipped with proximity probes. The forces are found to be a maximum near surge and increases with higher pressures and speeds. The results are nondimensionalized allowing the radial loading for different operating conditions to be predicted.

Topics: Gas compressors
Commentary by Dr. Valentin Fuster
1996;():V005T14A016. doi:10.1115/96-GT-121.
FREE TO VIEW

Force versus current and air gap measurements were obtained for the actuator component of a double acting magnetic thrust bearing constructed from a powdered metal. Static force measurements were made for various air gap settings and bearing current combinations. The resulting data was reduced and an optimized expression representing the force versus current and air gap relationship of the actuator was found. In addition, a theoretical force model was developed using simple magnetic circuit theory and constant material magnetic properties. The theoretical model predicted force magnitudes approximately 20 percent greater than the experimentally measured values. Hysteresis tests were conducted with the thrust disk in the centered position for various current perturbation amplitudes about the design bias current. Hysteresis effects were shown to cause a difference between the measured force as the current was increasing as compared to when the current was decreasing. This difference varied from 10% to 7% of the peak force from each respective hysteresis loop. A second order polynomial expression was developed to express the coercive force as a function of the perturbation current amplitude. The bearing frequency response was examined by injecting sinusoidal currents of varying frequencies into the bearing. An actuator bandwidth of at least 700 Hz was determined. Above 700 Hz the bearing frequency response could not be distinguished from the test fixture frequency response.

Commentary by Dr. Valentin Fuster
1996;():V005T14A017. doi:10.1115/96-GT-122.
FREE TO VIEW

A two-control-volume model is employed for honeycomb-stator/smooth-rotor seals, with a conventional control-volume used for the through flow and a “capacitance-accumulator” model for the honeycomb cells. The control volume for the honeycomb cells is shown to cause a dramatic reduction in the effective acoustic velocity of the main flow, dropping the lowest acoustic frequency into the frequency range of interest for rotordynamics. In these circumstances, the impedance functions for the seals can not be modeled with conventional (frequency-independent) stiffness, damping, and mass coefficients. More general transform functions are required to account for the reaction forces, and the transfer functions calculated here are a lead-lag term for the direct force function and a lag term for the cross-coupled function. Experimental measurements verify the magnitude and phase trends of the proposed transfer functions.

These first-order functions are simple, compared to transfer functions for magnetic bearings or foundations. For synchronous response due to imbalance, they can be approximated by running-speed-dependent stiffness and damping coefficients in conventional rotordynamics codes. Correct predictions for stability and transient response will require more general algorithms, presumably using a state-space format.

Commentary by Dr. Valentin Fuster
1996;():V005T14A018. doi:10.1115/96-GT-123.
FREE TO VIEW

This paper describes the proportional integral (PI) control of hybrid squeeze film dampers (HSFDs) for active control of vibrations of rotors. Recently it was shown that the automatically controlled HSFD based on feedback of rotor speed can be a very efficient device for active control of rotor vibration when passing through critical speeds. Although considerable effort has been put into the study of steady state vibration control, there are few methods in the literature applicable to transient vibration control of rotor-bearing systems. Rotating machinery may experience dangerously high dynamic loading due to the sudden mass unbalance that could be associated with blade loss. Transient run-up and coast down through critical speeds when starting up or shutting down rotating machinery induces excessive bearing loads at criticals. In this paper, PI control is proposed as a regulator for the HSFD system to attenuate transient vibration for both sudden unbalance and transient run-up through critical speeds. A complete mathematical model of this closed-loop system is simulated on a digital computer. Results show an overall enhanced behavior for the closed-loop rotor system. Gain scheduling of both the integral gain and the reference input is incorporated to the closed-loop system with the PI regulator and results in an enhanced behavior of the controlled system.

Topics: Bearings , Rotors
Commentary by Dr. Valentin Fuster
1996;():V005T14A019. doi:10.1115/96-GT-124.
FREE TO VIEW

The prediction capabilities of a linearized unsteady potential analysis have been extended to include supersonic cascades with subsonic axial flow. The numerical analysis of this type of flows presents several difficulties. First, complex oblique shock patterns exist within the cascade passage. Second, the acoustic response is discontinuous and propagates upstream and downstream of the blade row. Finally, a numerical scheme which is based on the domain of dependence is required for numerical stability. These difficulties are addressed by developing a discontinuity capturing scheme and matching the numerical near-field solution to an analytical far-field solution. Comparisons with semi-analytic results for flat plate cascades show that reasonable predictions of the unsteady aerodynamic response at the airfoil surfaces are possible, but aeroacoustic response calculations are difficult. Comparisons between flat plate and real blade cascade results show that one effect of real blades is the impulsive loads due to motion of finite strength shocks.

Topics: Blades
Commentary by Dr. Valentin Fuster
1996;():V005T14A020. doi:10.1115/96-GT-125.
FREE TO VIEW

In 1966 it was shown that the maximum factor by which the amplitude of forced vibration of blades can increase due 10 mistuning is, with certain assumptions, 1/2(1+√N), where N is the number of blades in the row. This report gives a further investigation of the circumstances when this factor can be obtained. These are, small damping, and a relationship must hold between the mistuning distribution γ(s) and the interblade coupling function c(r), where r is the mode number. The mistuning distribution must be symmetrical about the blade on which maximum amplitude is to occur, s=0. The coupling function must be symmetrical about r=R, where R is the mode number of the excitation.

If the coupling is purely mechanical, additional conditions apply. The coupling function c(r) must consist of a number of identical symmetrical sub-strips. A 1976 result for mechanical coupling is amended.

Topics: Vibration , Blades
Commentary by Dr. Valentin Fuster
1996;():V005T14A021. doi:10.1115/96-GT-218.
FREE TO VIEW

The first offshore turboexpanders with magnetic bearings have now been in operation on the Sleipner A platform in the North Sea since October 1993.

Four machines are installed, each at 3.3 MW, running at 16,500 RPM for natural gas dew point control.

During the commissioning phase and the first year of operation a number of problems were discovered, mainly due to the application of magnetic bearings in a new environment, and some unexpected bearing / process interaction phenomenon.

To solve the problems it was necessary to do modifications to electrical systems, mechanical parts, operational procedures and buffer gas systems.

After a 1.5 year period of modifications, all 4 machines are now running with 99 % availability.

Commentary by Dr. Valentin Fuster
1996;():V005T14A022. doi:10.1115/96-GT-260.
FREE TO VIEW

A parametric study of the unstalled flutter stability characteristics of compressor and turbine cascades in subsonic and supersonic flow is carried out. Based on typical section two-dimensional cascade models, stability boundaries and dominant trends in flutter behaviour are outlined with emphasis on the effects of (a) single mode structural coupling in bending and torsion, (b) coupling among multiple blade degrees of freedom, (c) mass ratio and structural damping, (d) compressibility, (e) cascade solidity and stagger, and (f) steady aerodynamic blade loading. Practical design aspects are in the foreground of all of these investigations.

Commentary by Dr. Valentin Fuster
1996;():V005T14A023. doi:10.1115/96-GT-264.
FREE TO VIEW

The analytical solution of whirl speed and mode shape of a rotating shaft in six boundary conditions is presented in this paper. The shaft is modelled by a Rayleigh beam with rotatory inertia and gyroscopic effects, and the boundary conditions are (1) short-short, (2) long-long, (3) long-free, (4) free-free, (5) long-short, and (6) short-free bearings. It is shown that the whirl speed can be written analytically by a function of the whirl ratio (λ) defined by the rotating speed over the whirl speed and the slenderness ratio (l) defined by the length of the shaft over its radius. The number of whirl speeds, contrary to common belief, is finite when λ > 1/2. For the first time, the rotating system’s unbalanced response can be written analytically in an exact form by a finite number of vibration modes with the corresponding generalized coordinates.

Topics: Whirls , Mode shapes
Commentary by Dr. Valentin Fuster
1996;():V005T14A024. doi:10.1115/96-GT-265.
FREE TO VIEW

This paper describes a computer program which can determine the probability of failure of gas turbine structures as a function of time due to creep fatigue crack growth. The probability of failure is computed by combining stress analysis and creep fatigue analysis with probabilistic analysis methods. The creep fatigue analysis is based on a reference stress approach which provides a simple, accurate, and efficient method for determining the steady state component, C*, of the time dependent fracture mechanics parameter C(t). Stress intensity factors are computed from stress distributions derived from a linear elastic finite element analysis of the uncracked structure and weight functions. Several probabilistic methods are available such as efficient approximate methods, importance sampling and Monte Carlo sampling. Efficient approximate methods and importance sampling methods are typically one to two orders of magnitude more efficient than Monte Carlo sampling. Probabilistic sensitivity measures are generated as a byproduct of the probabilistic analysis and indicate the importance of the random variables to the reliability of the structure. The theoretical background, computer code and an example problem are presented.

Commentary by Dr. Valentin Fuster
1996;():V005T14A025. doi:10.1115/96-GT-266.
FREE TO VIEW

Availability and Reliability have become a major concern for operating companies; especially for those where unscheduled outages due to purchased power agreements may be highly penalized. On the other hand, high revenues by the sale of electricity can be achieved if the power generating set performs according to its demand.

Economic calculations demonstrate that the bottom line between profit and loss lies within the high 90th% of availability. This paper summarizes the results of an analysis based on the regulations in France with its so-called EJP days (Effacement Jour Pointe). Any loss of highly penalized EJP days is quantified based on known RAM (Reliability, Availability, Maintainability) values of the equipment, as these are the Mean Time Between Failures MTBF’s and the Mean Times To Repair MTTR’s of the overall equipment as well as its subassemblies. In formulating the demand, the past 12 years of assigned EJP days by the Electricité de France, EdF, was analyzed to derive probability ratings of seasonal distributions, weekly distributions and day block distributions.

The mathematics of this simulation model are based on well proven statistical procedures (i.e. the Monte Carlo Method). By performing parameter variations, the model can also quantitatively predict how much the Mean Time Between Failures of a heavy duty gas turbine must usually be better for this application when compared to an aeroderivative gas turbine. This is because it normally takes longer to repair or replace a heavy duty gas turbine versus an aeroderivative unit in case of a major unscheduled or forced outage.

Commentary by Dr. Valentin Fuster
1996;():V005T14A026. doi:10.1115/96-GT-267.
FREE TO VIEW

For gas turbines used in power generation, reliability is one of the most important considerations. In the past, reliability was determined qualitatively (at best). When feedback from operation periods was available, reliability was recorded and rated statistically. However, in the present situation it is more and more necessary to quantify reliability quantities in the design stage of a new gas turbine (“built-in reliability”) with a sufficient accuracy.

Commentary by Dr. Valentin Fuster
1996;():V005T14A027. doi:10.1115/96-GT-268.
FREE TO VIEW

The stochastic Finite-Element-Method (FEM) is a suitable tool to assess the reliability of lifetime prediction models for complex components. Due to its demands on the number of FE analysis it has been rarely used on FE models of real components under realistic operational conditions. In the following a response surface approach is suggested that minimizes the number of FE calculations. Based on a first order method the sensitivity of the failure probability with respect to the random input quantities is evaluated. Subsequently a response surface of higher order, weighting the important and unimportant input uncertainties appropriately, is used to assess the reliability of the lifetime prediction. The method is used to analyze the reliability of typical rotor disks concerning failure due to creep rupture and low cycle fatigue (LCF) during static and transient operation.

Commentary by Dr. Valentin Fuster
1996;():V005T14A028. doi:10.1115/96-GT-306.
FREE TO VIEW

The present methods for life prediction in rotating parts rely heavily on an assumed distribution of potential crack initiation sites, corresponding to a distribution of properties which is determined from laboratory testing of samples of the alloy. However, as the expectations of the designer extend into increasing reliable life and also come closer to the limits of alloy performance, we also approach the limits of validity of the statistical methods used in property specification, raising the problem of the rare random defect which may be present in spite of the installed production control methodologies. This type of defect may be found in all alloys, but the specific form depends on the alloy and on the manufacturing method. In this work we give examples for high-strength steels, superalloys and titanium alloys.

We propose that the most effective solution to this problem lies in the more vigorous application of process control within manufacturing fixed practices, ensuring the absence of this type of defect through control and data acquisition techniques which can identify a process anomaly and trigger rejection or down-grading of the product.

Commentary by Dr. Valentin Fuster
1996;():V005T14A029. doi:10.1115/96-GT-307.
FREE TO VIEW

Subject of this paper is a numerical method for the simulation of flutter in three dimensional oscillating cascades. Unsteadiness can be caused by bending and torsional oscillation modes simultaneously. The goal of the investigation is the evaluation of the resulting blade forces and moments. The flow is assumed to be time-dependent and inviscid. By solving the Euler equations in a nonlinear way, large oscillations as well of the airfoil as of existing shocks can be treated.

The numerical solution follows a Godunov-type upwind scheme, formulated in node centered finite volume technique. An approximative Riemann solver proposed by Roe is used to determine the fluxes over the surfaces of the control volume. Since unphysical expansion shocks have to be suppressed, a modification of the transonic characteristic speeds is included. The extrapolation of the flow values onto the control volumes’ surfaces is done by means of the MUSCL technique, embedded in a TVD-scheme with the flux limiter by van Albada. The computational domain is restricted to only one channel and the periodic values are stored over one period of oscillation. A special technique is introduced, which reduces both the effort in CPU-time and in computer memory. Results are included for compressor and turbine geometries in sub- and transonic flow.

Commentary by Dr. Valentin Fuster
1996;():V005T14A030. doi:10.1115/96-GT-311.
FREE TO VIEW

This paper describes the development of an aeroelastic code (TURBO-AE) based on an Euler / Navier-Stokes unsteady aerodynamic analysis. A brief review of the relevant research in the area of propulsion aeroelasticity is presented. The paper briefly describes the original Euler / Navier-Stokes code (TURBO) and then details the development of the aeroelastic extenstons. The aeroelastic formulation is described. The modeling of the dynamics of the blade using a modal approach is detailed, along with the grid deformation approach used to model the elastic deformation of the blade. The work-per-cycle approach used to evaluate aeroelastic stability is described. Representative results used to verify the code are presented. At the present stage of development, the aeroelastic code is limited to in-phase blade motions. The paper concludes with an evaluation of the development thus far, and some plans for further development and validation of the TURBO-AE code.

Commentary by Dr. Valentin Fuster
1996;():V005T14A031. doi:10.1115/96-GT-334.
FREE TO VIEW

An experimental investigation of the steady-state and time-dependent aerodynamic behaviour of a compressor cascade in a ring channel was conducted at the Laboratoire de thermique appliquée et de turbomachines (LTT) at the Swiss Federal Institute of Technology in Lausanne. The cascade consisted of 20 blades with a NACA-3506 profile, stagger angle of 40°, and solidity of 0.72 at midspan. Measurements were done for a number of incidence angles over a small range of inlet Mach numbers between ∼0.75 and ∼0.8 in order to examine the influence of an increasing angle of attack on the steady-state and time-dependent pressures. As the angle of attack increased a growing corner stall was observed at the hub and a supersonic zone appeared at the leading edge.

The cascade was vibrated in bending mode with a constant amplitude at a reduced frequency of ∼0.42 at imposed interblade phase angles ranging from 0° to 324°, but also with each blade vibrating in a single blade vibration mode. The unsteady data showed that the cascade was in general damped with the minimum damping between ∼−36° to ∼+36° interblade phase angle for all examined incidence angles. The influence coefficient technique was used to identify the damping influence of each of the blades on itself (eigeninfluence) and of blades up and down the cascade (positive- and negative-sided) for different inlet incidence angles.

Commentary by Dr. Valentin Fuster
1996;():V005T14A032. doi:10.1115/96-GT-335.
FREE TO VIEW

The paper presents an approach for the modal aeroelastic analysis of three-dimensional turbomachinery bladings with several fluid and structure analyzers. Structure analyzers are three-dimensional solvers for static and dynamic analyses of axisymmetric/cyclic-symmetric blade-shroud-disk-shaft assemblies with/without elastic coupling between blades. Fluid analyzers are two-dimensional/three-dimensional solvers for single/multi-stage steady/unsteady turbomachinery flows. An automatic interfacing procedure for exchanging data at the incompatible fluid-structure boundary and the development of a multi-model interfacing software are discussed. The modal aeroelastic analysis of a first stage shrouded fan is carried out to illustrate the main issues of the paper. In particular, two structural models for the elastic coupling of the part-span shrouds are discussed. The results show the strong dependence of the structure dynamics and aeroelastic analysis on this modelling.

Commentary by Dr. Valentin Fuster
1996;():V005T14A033. doi:10.1115/96-GT-338.
FREE TO VIEW

This paper presents the numerical results of a code for computing the unsteady transonic viscous flow in a two-dimensional cascade of harmonically oscillating blades. The flow field is calculated by a Navier-Stokes code, the basic features of which are the use of an upwind flux vector splitting scheme for the convective terms (Advection Upstream Splitting Method), an implicit time integration and the implementation of a mixing length turbulence model.

For the present investigations two experimentally investigated test cases have been selected in which the blades had performed tuned harmonic bending vibrations. The results obtained by the Navier-Stokes code are compared with experimental data, as well as with the results of an Euler method.

The first test case, which is a steam turbine cascade with entirely subsonic flow at nominal operating conditions, is the fourth standard configuration of the “Workshop on Aeroelasticity in Turbomachines”. Here the application of an Euler method already leads to acceptable results for unsteady pressure and damping coefficients and hence this cascade is very appropriate for a first validation of any Navier-Stokes code. The second test case is a highly-loaded gas turbine cascade operating in transonic flow at design and off-design conditions. This case is characterized by a normal shock appearing on the rear part of the blades’s suction surface, and is very sensitive to small changes in flow conditions. When comparing experimental and Euler results, differences are observed in the steady and unsteady pressure coefficients. The computation of this test case with the Navier-Stokes method improves to some extent the agreement between the experiment and numerical simulation.

Commentary by Dr. Valentin Fuster
1996;():V005T14A034. doi:10.1115/96-GT-339.
FREE TO VIEW

The aerodynamics of a cascade of airfoils oscillating in torsion about the midchord is investigated experimentally at a large mean incidence angle and, for reference, at a low mean incidence angle. The airfoil section is representative of a modern, low aspect ratio, fan blade tip section. Time-dependent airfoil surface pressure measurements were made for reduced frequencies of up to 1.2 for out-of-phase oscillations at a Mach number of 0.5 and chordal incidence angles of 0° and 10°; the Reynolds number was 0.9×106. For the 10° chordal incidence angle, a separation bubble formed at the leading edge of the suction surface. The separated flow field was found to have a dramatic effect on the chordwise distribution of the unsteady pressure. In this region, substantial deviations from the attached flow data were found with the deviations becoming less apparent in the aft region of the airfoil for all reduced frequencies. In particular, near the leading edge the separated flow had a strong destabilizing influence while the attached flow had a strong stabilizing influence.

Commentary by Dr. Valentin Fuster
1996;():V005T14A035. doi:10.1115/96-GT-340.
FREE TO VIEW

A time-accurate Euler/Navier-Stokes analysis is applied to predict unsteady subsonic and transonic flows through a vibrating cascade. The intent is to validate this nonlinear analysis along with an existing linearized inviscid analysis via result comparisons for unsteady flows that are representative of those associated with blade flutter. The time-accurate analysis has also been applied to determine the relative importance of nonlinear and viscous effects on blade response. The subsonic results reveal a close agreement between inviscid and viscous unsteady blade loadings. Also, the unsteady surface pressure responses are essentially linear, and predicted quite accurately using a linearized inviscid analysis. For unsteady transonic flows, shocks and their motions cause significant nonlinear contributions to the local unsteady response. Viscous displacement effects tend to diminish shock strength and impulsive unsteady shock loads. For both subsonic and transonic flows, the energy transfer between the fluid and the structure is essentially captured by the first-harmonic component of the nonlinear unsteady solutions, but in transonic flows, the nonlinear first-harmonic and the linearized inviscid responses differ significantly in the vicinity of shocks.

Topics: Blades
Commentary by Dr. Valentin Fuster
1996;():V005T14A036. doi:10.1115/96-GT-373.
FREE TO VIEW

Rotors with three or more fluid-film bearings (or fluid seals) have ‘redundant’ supports, and therefore interdependent bearing loads which are generally unknown both in magnitude and direction. The steady-state bearing eccentricities and the dynamic stiffness and damping coefficients of the bearings are therefore also unknown, since both are functions of the bearing loads. Thus, the dynamic behaviour of multi-bearing rotors generally cannot be predicted with good accuracy without access to a procedure for calculating the steady-state bearing loads and eccentricities. This paper outlines such a procedure in terms of both the influence coefficient method, the transfer matrix method, and the finite element method. Radial bearing misalignment and flexibility of the bearing back-up structures are accounted for. Once the eccentricities are available, the bearing stiffness and damping coefficients can be calculated in the usual way and used to predict critical speeds, instability threshold speed and rotor response to imbalance. A numerical example is presented which illustrates some of the non-linear effects of bearing support redundancy, notably the large variations in instability threshold speed with radial bearing misalignment. The example shows how the method can be used to determine the level of bearing misalignment which leads to optimum rotor stability. It is concluded that no simple guide lines exist by which optimum stability can be achieved. Neither perfect bearing alignment nor equal load sharing between bearings necessarily lead to optimum stability.

Commentary by Dr. Valentin Fuster
1996;():V005T14A037. doi:10.1115/96-GT-374.
FREE TO VIEW

An experimental and computational study has been carried out on a linear cascade of low pressure turbine blades with the middle blade oscillating in a torsion mode. The main objectives of the present work were to enhance understanding of the behaviour of bubble type of flow separation and to examine the predictive ability of a computational method. In addition, an attempt was made to address a general modelling issue: was the linear assumption adequately valid for such kind of flow?

In Part 1 of this paper, the experimental work was described. Unsteady pressure was measured along blade surfaces using off-board mounted pressure transducers at realistic reduced frequency conditions. A short separation bubble on the suction surface near the trailing edge and a long leading-edge separation bubble on the pressure surface were identified. It was found that in the regions of separation bubbles, unsteady pressure was largely influenced by the movement of reattachment point, featured by an abrupt phase shift and an amplitude trough in the 1st harmonic distribution. The short bubble on the suction surface seemed to follow closely a laminar bubble transition model in a quasi-steady manner, and had a localized effect. The leading-edge long bubble on the pressure surface, on the other hand, was featured by a large movement of the reattachment point, which affected the surface unsteady pressure distribution substantially. As far as the aerodynamic damping was concerned, there was a destabilizing effect in the separated flow region, which was however largely balanced by the stabilizing effect downstream of the reattachment point due to the abrupt phase change.

Commentary by Dr. Valentin Fuster
1996;():V005T14A038. doi:10.1115/96-GT-375.
FREE TO VIEW

Unsteady flow around a linear oscillating turbine cascade has been experimentally and computationally studied, aimed at understanding the bubble type of flow separation and examining the predictive ability of a computational method. It was also intended to check the validity of the linear assumption under an unsteady viscous flow condition.

Part 2 of the paper presents a computational study of the experimental turbine cascade as discussed in Part 1. Numerical calculations were carried out for this case using an unsteady Navier-Stokes solver. The Baldwin-Lomax mixing length model was adopted for turbulence closure. The boundary layers on blade surfaces were either assumed to be fully turbulent or transitional with the unsteady transition subject to a quasi-steady laminar separation bubble model. The comparison between the computations and the experiment were generally quite satisfactory, except in the regions with the flow separation. It was shown that the behaviour of the short-bubble on the suction surface could be reasonably accounted for by using the quasi-steady bubble transition model. The calculation also showed that there was a more apparent mesh dependence of the results in the regions of flow separation.

Two different kinds of numerical tests were carried out to check the linearity of the unsteady flow and therefore the validity of the Influence Coefficient method. Firstly calculations using the same configurations as in the experiment were performed with different oscillating amplitudes. Secondly calculations were performed with a tuned cascade model and the results were compared with those using the Influence Coefficient method. The present work showed that nonlinear effect was quite small, even though for the most severe case in which the separated flow region covered about 60% of blade pressure surface with a large movement of the reattachment point. It seemed to suggest that the linear assumption about the unsteady flow behaviour should be adequately acceptable for situations with bubble type flow separation similar to the present case.

Commentary by Dr. Valentin Fuster
1996;():V005T14A039. doi:10.1115/96-GT-376.
FREE TO VIEW

This paper describes a portion of an experimental and computational program (ADLARF) which incorporates, for the first time, measurements of all aspects of the forced response of an airfoil row, i.e., the flow defect, the unsteady pressure loadings and the vibratory response. The purpose of this portion was to extend the knowledge of the unsteady aerodynamics associated with a low aspect ratio transonic fan where the flow defects were generated by inlet distortions. Measurements of screen distortion patterns were obtained with total pressure rakes and casing static pressures. The unsteady pressure loadings on the blade were determined from high response pressure transducers. The resulting blade vibrations were measured with strain gages. The steady flow was analyzed using a 3D Navier–Stokes solver while the unsteady flow was determined with a quasi–3D linearized Euler solver.

Experimental results showed that the distortions had strong vortical, moderate entropic and weak acoustic parts. The 3D Navier–Stokes analyses showed that the steady flow is predominantly two–dimensional, with radially outward flow existing only in the blade surface boundary layers downstream of shocks and in the aft part of the suction surface. At near resonance conditions, the strain gage data showed blade–to–blade motion variations and thus, linearized unsteady Euler solutions showed poorer agreement with the unsteady loading data than comparisons at off–resonance speeds. Data analysis showed that entropic waves generated unsteady loadings comparable to vortical waves in the blade regions where shocks existed.

Commentary by Dr. Valentin Fuster
1996;():V005T14A040. doi:10.1115/96-GT-377.
FREE TO VIEW

There has been much research work carried out on various aspects of individual squeeze-film dampers (SFDs) but very little on the interplay between a damper and the rotating assembly of which it forms a part. In this paper, a flexible rotor-bearing assembly in a configuration, typical of a small centrifugal pump and incorporating an SFD is investigated theoretically and experimentally from the points of view of forced vibration control and stability control. It is found that change in rotor unbalance, SFD static eccentricity ratio and SFD supply pressure can cause significant movement of system resonances and vibration resulting from excessive damping. The provision of an SFD also delays the onset of instability and because of its nonlinearity, the SFD contributes more damping than can a linear damper when the vibration amplitude becomes large as instability develops. It is shown that this instability is curbed at some limit cycle, whose frequency is a system natural frequency.

Topics: Stability , Dampers , Rotors
Commentary by Dr. Valentin Fuster
1996;():V005T14A041. doi:10.1115/96-GT-378.
FREE TO VIEW

During full load shop tests under natural gas, two multistage centrifugal compressors exhibited subsynchronous vibrations. Both of them are low flow, high pressure, high rotational speed compressors, and are fitted with tilting pad bearings, and dry gas seals. A rotating stall problem was firstly eliminated by a modification of the diffuser geometry. Then, aerodynamic excitations caused the rotors to operate at their stability limit, and high vibration levels were observed at the first natural frequency. A complete rotordynamics analysis was performed in order to model precisely all the fluid-structure interactions. Modifications of the rotor designs were implemented, consisting in optimizing conveniently the bearing pads, replacing the toothed labyrinth seals of the balance pistons by damping honeycomb seals, fitting them with improved shunt hole systems. In addition, the dry gas seals were found to have been damaged, due to thermal effects, and further modifications were implemented to eliminate this problem. Final full load tests demonstrated a satisfactory behaviour of both centrifugal compressors.

Commentary by Dr. Valentin Fuster
1996;():V005T14A042. doi:10.1115/96-GT-379.
FREE TO VIEW

Experience has shown that energy dissipation due to rubbing at interfaces of engine blade shrouds provides damping of bladed disk assemblies. An accurate estimate of such damping has been a subject of study for more than a decade. The most difficult parameter that influences its accurate calculation pertains to an appropriate definition of the laws of friction that relate the rubbing motion with the forces induced.

This paper develops an analysis which leads to a mathematical relationship between the forces of friction damping at a vibrating interface and the corresponding relative motion. The analysis permits calculation of forces at the interface if the displacements are known or vice versa. The equations are cast in terms of relative motion between mating shrouds so that degenerate cases of fully locked and freely slipping can be calculated. Examples are given showing simulation results obtained using discrete structural models. Extension of the analysis to the case of a full assembly is discussed.

Commentary by Dr. Valentin Fuster
1996;():V005T14A043. doi:10.1115/96-GT-398.
FREE TO VIEW

This paper discusses the methods used to derive an improved design for gas turbine exhaust duct expansion joints. Typically these joints are subjected to very rapid increase in internal exhaust gas temperatures that result in large temperature differentials within the joint structures. The thermal gradients can cause stress levels in excess of yield and when the turbine is used intermittently, such as peaking power units would be, the net result is crack propagation and gas leakage.

Commentary by Dr. Valentin Fuster
1996;():V005T14A044. doi:10.1115/96-GT-399.
FREE TO VIEW

The cogeneration facility of interest contains four identical cogeneration units which produce about one half million Kg per hour of steam for an adjacent refinery and 385 megawatts of electricity. To supplement the production of steam, burners are used to heat the gas turbine exhaust. These burners incorporate shields to deflect exhaust gas flow around the flame base.

In an effort to improve burner emissions of the units, the burner shield design was modified. This alteration resulted in gross deformation of the shields which interfered with combustion. A failure analysis of these components was conducted to ascertain the root cause of the observed behavior. Loads were estimated based upon operational conditions and material properties were obtained from the open literature. An evaluation was conducted to determine the temperature distribution first. This temperature distribution was then coupled with mechanical loading to obtain total operational stress levels. The stress levels at the observed temperatures clearly placed the material in the high strain rate (creep) region. The computed stress distribution confirmed the observed failure configuration.

A new design was proposed to eliminate this failure mechanism. Detailed evaluations revealed that the new design, while a significant improvement, still operated near the creep region for the material.

Commentary by Dr. Valentin Fuster
1996;():V005T14A045. doi:10.1115/96-GT-401.
FREE TO VIEW

An experimental investigation of the aerodynamic response of the rotor blading of a single stage low speed axial research compressor, which was exposed to wake type inlet flow disturbances was carried out at the Institute’s low speed compressor and fan research facility. The investigation comprised 5 hole pressure probe surveys of the nonuniform flow field at the rotor inlet and exit planes as well as direct recording of the unsteady surface pressure response by rotor blade mounted ‘fast response’ miniature pressure transducers. Forcing functions in terms of streamwise and transverse gust velocities were determined from the circumferential and radial distribution of the inlet flow properties. The blade surface pressure history of the unsteady aerodynamic response was recorded at several locations along span and along chord. The influence of steady stage loading on the forcing function as well as on the unsteady rotor blade response was investigated. At midspan the recorded local blade surface pressure histories were integrated to a spanwise unsteady blade force coefficient, which was correlated to a calculated force coefficient determined by a two-dimensional linearized unsteady aerodynamic analysis.

Commentary by Dr. Valentin Fuster
1996;():V005T14A046. doi:10.1115/96-GT-402.
FREE TO VIEW

A computer code, BDAMPER, was recently developed by Ohio State University researchers and can be used to predict the vibratory response of shrouded blades that contain friction interfaces. This paper discusses some modeling issues that arose in applying BDAMPER to actual blades and outlines a procedure for optimizing the shroud’s design in order to minimize the blade’s resonant response. A comparisons are made between BDAMPER predictions of blade response and experimental data taken from spin pit and engine tests.

Commentary by Dr. Valentin Fuster
1996;():V005T14A047. doi:10.1115/96-GT-403.
FREE TO VIEW

The effects of support/foundation dynamics are often significant in high speed turbomachinery, and can affect the stability and response to unbalance. In some cases additional critical speeds are introduced, related to resonances in the foundation or interaction with rotor resonances of foundation resonances.

This paper reviews several methods for representing these effects, including (1) reduced matrices from finite element substructures (ANSYS, for example), (2) matrices generated from modal data, and (3) direct use of experimental transfer functions. These methods are implemented in a finite element rotor program in a PC-DOS environment.

The application of the methods to two laboratory rotor configurations described and results presented. Situations with a foundation resonance above and near the rotor critical are included. The importance of including coupling effects between supports is shown.

Commentary by Dr. Valentin Fuster
1996;():V005T14A048. doi:10.1115/96-GT-407.
FREE TO VIEW

This paper deals with the stability analysis of self-excited bending vibrations of linear symmetrical rotor-bearing systems caused by internal damping using the finite element method. The rotor system consists of uniform circular Rayleigh shafts with internal viscous damping, symmetrical rigid disks, and discrete undamped isotropic bearings. By combining the sensitivity method and the matrix representation of the rotor dynamic equations in complex form to assess stability, it is proved theoretically that the stability threshold speed and the corresponding whirling speed coincide with the first forward critical speed regardless of the magnitude of the internal damping.

Commentary by Dr. Valentin Fuster
1996;():V005T14A049. doi:10.1115/96-GT-408.
FREE TO VIEW

The paper contains the study of the lubricant fluid forces applied to the journal in a cylindrical, externally-pressurized bearing. The analytical expressions take into account the appearance of fluid cavitation at high eccentricity level.

The obtained expressions for fluid forces and torque are used for numerical simulation of a simple, one-mode rotor system. The system exhibits typical fluid whirl/whip instability which results in high fluid resistance torque values. The latter is considered as one of the causes of turbomachinery efficiency deterioration due to fluid-induced instabilities.

This paper is a continuation of the study undertaken by Bently et al. (1985), Muszynska (1986), and Petchenev et al. (1995).

Topics: Torque , Fluids , Bearings , Rotors
Commentary by Dr. Valentin Fuster
1996;():V005T14A050. doi:10.1115/96-GT-414.
FREE TO VIEW

A recently introduced perturbation technique is employed to derive a novel closed form model for the probability density function of the resonant and near-resonant, steady state amplitude of blade response in randomly mistuned disks. In its most general form, this model is shown to involve six parameters but, in the important practical case of a pure stiffness (or frequency) mistuning, only three parameters are usually sufficient to completely specify this distribution. A series of numerical examples are presented that demonstrate the extreme reliability of this three-parameter model in accurately predicting the entire probability density function of the amplitude of response, and in particular the large amplitude tail of this distribution which is the most critical effect of mistuning.

Topics: Disks
Commentary by Dr. Valentin Fuster
1996;():V005T14A051. doi:10.1115/96-GT-417.
FREE TO VIEW

The Institute of Aeroelasticity operates a test facility which enables aeroelastic investigations of plane cascades in low-speed flow. The test stand serves as a pilot facility to develop tools for analogous investigations in transonic flow. Eleven blades are elastically suspended in a windtunnel with a 1 × 0.2 m2 cross section.

This paper describes the experimental method of determining the flutter boundary by extrapolation of the results measured in subcritical flow. A 2D theoretical model of the 11 blades including the windtunnel walls permits the computation of unsteady pressures, forces and moments in close relation to the experiment. The prediction of flutter is compared with experimental results. In the present investigation, the motion of the blades is constrained to pitch around mid-chord. The vibrating blades are mechanically uncoupled. Any interaction between the blades is effected by the air stream, leading to a sensitive dependence on the aerodynamic forces.

Commentary by Dr. Valentin Fuster
1996;():V005T14A052. doi:10.1115/96-GT-429.
FREE TO VIEW

A finite volume method for blade flutter analyses, using moving grids is presented and partly validated. The method which solves the unsteady three-dimensional Euler equations is formulated in the four-dimensional time-space domain. An algebraic grid generation technique based on transfinite interpolation is used to move and deform the grid to conform to the blade motion. Fluxes are calculated using a third-order upwind-biased scheme. For time marching both an explicit three-stage Runge-Kutta scheme and a Crank-Nicolson scheme is used. Internal and external flows are calculated using the present method. Calculated results agree well with the corresponding experiments and with results obtained using other methods.

Commentary by Dr. Valentin Fuster
1996;():V005T14A053. doi:10.1115/96-GT-443.
FREE TO VIEW

In order to predict rotor-stator interactions in hydraulic turbomachinery, a new multidomain method was implemented in a turbulent flow finite element code. The formulation is based on the mortar element method, extended to sliding meshes. It allows the coupling between several domains using nonconforming discretization on the interface.

This method was used to study the unsteady flow in a cross-flow fan. Such fans are widely used in air-conditioning devices. They consist of a cylindrical rotor with forward curved blades, rotating in a housing. The air passes the blades twice at the inlet and the outlet of the impeller while a large vortex is noticeable inside the impeller which is located in the blading near the stabilizer.

The numerical results are compared to Laser Doppler Velocity measurements and to steady calculations obtained by replacing the blades by a rotating force field. The unsteady calculation shows the formation of the vortex structure at the start-up and the periodic established flow after a few revolutions of the impeller.

Commentary by Dr. Valentin Fuster
1996;():V005T14A054. doi:10.1115/96-GT-469.
FREE TO VIEW

Friction dampers are widely used to improve the performance of rotating blades. This paper is concerned with the steady stale response and stability analysis of ratating composite plates in the presence of non linear friction damping. Direct Integration Method (DIM) and Harmonic Balance Method (HBM) are used to determine the steady state response due to periodic lateral external forces. In addition, an alternate procedure, Hybrid Method (HM) is proposed for this analysis to substantiate the results from DIM and HBM. The analysis shows that the steady state response is a function of friction damping magnitude as well as its location besides the excitation frequency and the rotational speed. A stability analysis of the composite blades is also made by including periodic in-plane excitation using Floquet-Liapunov theory.

Commentary by Dr. Valentin Fuster
1996;():V005T14A055. doi:10.1115/96-GT-470.
FREE TO VIEW

The continual increase in the use of magnetic bearings in various capacities, including high speed aerospace applications such as jet engine prototypes, dictates the need to quantify power losses in this type of bearing. The goal of this study is to present experimentally measured power losses during the high speed nperatinn nf a pair of magnetic bearings. A large scale test rotor has been designed and built to obtain unambiguous power loss measurements while varying a variety of test parameters. The test apparatus consists of a shaft supported in two radial magnetic bearings and driven by two electric motors also mounted nn the shaft. The power losses of the spinning rotor are determined from the time rate of change of the kinetic energy of the rotor as its angular speed decays during free rotatinn. Measured results for the first set of magnetic bearings, a pair of 8-pole planar radial bearings, are presented here. Data from three different parameter studies including the effect of the hias flux density, the effect of the bearing pole configuration, and the effect of the motor stator on the power loss are presented. Rundown pints of the test with the bearings in the paired pole (NNSS) versus the alternating (NSNS) pole configuration shnw only small differences, with losses only slightly higher when the poles are in the alternating pole (NSNS) configuration. Loss data was also taken with the motor statnrs axially removed from the mntnr rotors for comparison with the case where the mntor stators are kept in place. No measurable difference was observed between the two cases, indicating negligible windage and residual magnetic effects. Throughout mnst of the speed range the dominant loss mechanism appears to be eddy currents.

Commentary by Dr. Valentin Fuster
1996;():V005T14A056. doi:10.1115/96-GT-471.
FREE TO VIEW

The continual increase in the use nf magnetic bearings in varinus capacities, including high speed aerospace applications such as jet engine prototypes, dictates the need to quantify power losses in this type of bearing. The goal of the present study is to develop and experimentally verify general power loss equations for the high speed operation of magnetic bearings. Experimental data from a large scale test rotor has been presented in Part 1 of this study. Analytical/empirical predictions are presented here for the test bearings, a pair of 8-pole planar radial bearings, for comparison to the experimental results from Part 1. Expressions for the four loss components, eddy current, alternating hysteresis, rotating hysteresis, and windage, are also presented. Analytical/empirical predictions for the test bearings at three different bias flux levels demonstrate good correlation with corresponding experimental data. Throughout mnst of the speed range the dominant loss mechanism appears to be eddy currents.

Commentary by Dr. Valentin Fuster
1996;():V005T14A057. doi:10.1115/96-GT-472.
FREE TO VIEW

Designers of aircraft engines frequently employ shrouds in turbine design. In this paper, a variable normal load friction force model is proposed to investigate the influence of shroud-like contact kinematics on the forced response of frictionally constrained turbine blades. Analytical criteria are formulated to predict the transitions between slick, slip, and separation of the interface so as to assess the induced friction forces. When considering cyclic loading, the induced friction forces are combined with the variable normal load so as to determine the effective stiffness and damping of the friction joint over a cycle of motion. The harmonic balance method is then used to impose the effective stiffness and damping of the friction joint on the linear structure. The solution procedure for the nonlinear response nf a two-degree-of-freedom oscillator is demonstrated. As an application, this procedure is used to study the coupling effect of two constrained forces, friction force and variable normal load, on the optimization of the shroud contact design.

Topics: Friction , Design , Modeling
Commentary by Dr. Valentin Fuster
1996;():V005T14A058. doi:10.1115/96-GT-473.
FREE TO VIEW

The effect of combustion dynamic pressure oscillations on the structural integrity of the MS 7001F dry low NOX 2 (DLN 2) combustion system has been evaluated using ANSYS [ref. 1] finite element analyses and high cycle fatigue material data. Analytical results were validated with laboratory measurements on the combustion system subjected to combustion dynamic pressure at actual gas turbine temperature and pressure operating conditions. The combustion liner, transition piece, impingement sleeve and supports were proven to have excellent durability when subjected to dynamic loads. No risk of structural failure exists at anticipated dynamic pressures using assumptions shown to be conservative.

Commentary by Dr. Valentin Fuster
1996;():V005T14A059. doi:10.1115/96-GT-511.
FREE TO VIEW

In this paper, the stall and choke flutter analyses of NACA 0006 unstaggered cascades in transonic viscous flows are presented by using a time domain approach. For the present time domain approach, a solution-adaptive finite volume method with rigid-deformable dynamic mesh treatment is adopted to solve the two-dimensional unsteady Navier-Stokes equations. The structural model equations, where each blade is treated as a typical section having plunging and pitching degrees of freedom, are integrated to obtain the blade displacements by an explicit four-stage Runge-Kutta scheme. In the present calculations, the Baldwin-Lomax turbulence model and two transition formulations are used. The instantaneous meshes, vorticity contours, pressure contours and velocity vectors around the trailing edge clearly indicate the flow phenomena, such as the vortex shedding and λ shocks with separation bubbles. From the histories of blade displacements and total energy, the flutter phenomena are studied. Furthermore, the Fast Fourier Transformation (FFT) and modal identification techniques are introduced to investigate the aeroelastic behaviors in present transonic stall and choke flutter problems.

Commentary by Dr. Valentin Fuster
1996;():V005T14A060. doi:10.1115/96-GT-523.
FREE TO VIEW

The design principles of the splined-disc-type rotor with Hirth-serration couplings used in all Siemens heavy-duty gas turbines since 1960 are explained. The benefits of this rotor construction are elaborated against the operational requirements imposed on rotors by present-day large advanced gas turbines for electric power generation. Details of the stress assessment, analysis and testing underlying the design, as well as a description of the assembly of the rotor are also presented, using mostly the 170 MW-class 3600 rpm Model V84.3A gas turbine to illustrate the most recent technology.

Commentary by Dr. Valentin Fuster
1996;():V005T14A061. doi:10.1115/96-GT-524.
FREE TO VIEW

This paper describes the design and development of a new Cylinder Bolting System to replace the Main Joint Hardware for both Combustion (and Steam) Turbine applications. The new bolts are designed to be hydraulically tensioned to the specified preloed and utilize ultrasonic verification of elongation. The new Bolting System uses a reduced number of components in each assembly and the individual components themselves are of a simplified design. The new hardware can be applied to new equipment without modification and retrofitted to customer owned equipment as a direct replacement for existing joint hardware. The prototype, production and field testing of this hardware, the installation tooling and ultrasonic elongation measuring equipment is described. This testing has shown significant savings in assembly and disassembly cycle times even after prolonged exposure to turbine operating temperatures in a corrosive environment. The new design of bolting is now standard equipment for the CW251B11/B12 Combustion Turbine manufactured by Westinghouse P.G.B.U.

Topics: Turbines , Cylinders
Commentary by Dr. Valentin Fuster
1996;():V005T14A062. doi:10.1115/96-GT-528.
FREE TO VIEW

The depth of internal oxidation and nitridation from the surface of the 16 cooling holes in a first-stage turbine blade was measured by optical microscopy after 32,000 hours of service. Maximum depth of penetration was 15.5 mils (0.4 mm) at the trailing edge hole. An effort was made to predict hole surface metal temperatures based on these measurements using the Arrhenius relationship between time and temperature with depth of oxidation assumed to be parabolic with time. Reasonable correlations were obtained between finite element analysis results and temperature estimates based on the oxidation measurements. In the thickest part of the airfoil, where metal temperature is minimum, intergranular cracks up to 12.6 mils (0.32 mm) in depth were found at the surface of the cooling holes. Measurable oxidation attack was only one to two mils (0.025–0.050mm). Based on an approximate elastic-relaxation-local inelastic stress analysis, it was calculated that inelastic local strains of over one percent occur at the points of cracking. No cracking was observed in the more heavily oxidized, lower stressed, hotter holes. However, cracking occurred in a trailing edge tip cooling hole when weld repair of the tip squealer was attempted, due to embrittlement and grain boundary oxidation from service exposure. Temperature estimates suitable for life assessment purposes using oxidation measurements appears to be a possible technique that should be further developed and validated.

Commentary by Dr. Valentin Fuster
1996;():V005T14A063. doi:10.1115/96-GT-539.
FREE TO VIEW

A surface integral formulation, based on representing a crack as a distribution of force dipoles, has been developed for modeling the propagation of a three-dimensional nonplanar fracture. The minimum strain energy density, and maximum circumferential stress theories were used to determine the direction of crack growth. The extension of the fracture surface was based on the Paris Law for fatigue. Remeshing of the fracture during growth was accomplished by adding a ring of elements to the existing mesh at the conclusion of each increment of crack growth. This promoted the efficiency of the algorithm by eliminating the need to recalculate the entire coefficient matrix.

Use of the surface integral method, coupled with growth criteria, has yielded an accurate model for three-dimensional nonplanar crack growth under mixed mode loading conditions. The study of several penny-shaped precracks under mixed mode loading conditions produced the expected growth trajectory, and compared favorably to existing 2D, 3D, and experimental results found in the literature.

Commentary by Dr. Valentin Fuster

Controls, Diagnostics and Instrumentation

1996;():V005T15A001. doi:10.1115/96-GT-002.
FREE TO VIEW

This paper describes the development of a Microwave Tip Clearance Measurement System for use in the gas turbine environment Applications for this sensor include basic tip clearance measurements, seal wear measurement and active blade tip clearance control in gas turbine engines. The system being developed was designed for useful operation to temperatures exceeding 1093°F, since only ceramic materials are directly exposed in the gas path. Other advantages of this microwave approach to blade tip clearance sensing include the existence of an inherent self-calibration in the sensor that permits accurate operation despite temperature variations and possible abrasion by the rotating blades. Earlier experiments designed to simulate this abrasion of the sensor head indicated that rubs as deep as 1 mm (40 mils) were easily tolerated. In addition, unlike methods based upon phase measurements, this method is very insensitive to cable vibration and length variations. Finally, this microwave technique is expected to be insensitive to fuel and other engine contamination, since it is based on the measurement of resonant frequencies, which are only slightly affected by moderate values of loss due to contamination.

Commentary by Dr. Valentin Fuster
1996;():V005T15A002. doi:10.1115/96-GT-003.
FREE TO VIEW

In current generations of aero-engines the need to ensure the reliability of sensor measurements is met by using two or more sensors to read the same engine parameter. This technique is known as “hardware redundancy” and although reliable it does increase the cost of the sensor suite and also the weight of the engine. With the increasing computing power being fitted to new engines the opportunity has arisen of using “analytical redundancy” instead. Here the outputs of a software model of the engine are used to validate the real sensor outputs. Analytical redundancy divides into two main parts: the detection of a fault on a sensor, and the reconfiguration of the control system in response to this fault. The work conducted at DRA Pyestock has concentrated on the detection of sensor faults. The method employed uses a correlation approach to compare the shapes of the signals from the sensor and model. The sensor is declared to be faulty if the shapes become too dissimilar. Sea-level static engine trials have been conducted at Pyestock in which different faults were added to the sensor outputs of a Spey turbofan engine. The results were encouraging and indicated that such a fault detection approach could form part of a practical analytical redundancy scheme.

Commentary by Dr. Valentin Fuster
1996;():V005T15A003. doi:10.1115/96-GT-004.
FREE TO VIEW

Non-contact eddy-current displacement transducers, or proximity probes as they are popularly known, have always played an important role in monitoring the condition of a variety of machines, and will remain doing so for years to come. Although immensely useful for detecting a wide range of machine faults by simply monitoring the shaft position and vibration, there are definite limitations to the traditional probe systems due to the aging technology. Some of these problems are the probe driver sensitivity to the probe coil parameters, the shaft’s electromagnetic properties, and the capacitance of the cable it is attached to. Since each probe driver has to be individually calibrated to a specific probe tip diameter, a specific kind of shaft material and a specific cable length, it is easy to imagine that stocking such dedicated drivers can be expensive for large applications. But what is even more costly than this is the risk of reduced monitoring reliability due to an incorrectly selected or calibrated driver. As a result of this, a new proximity probe system recently has been designed and tested that uses a simple yet innovative principle that makes it insensitive to shaft material electromagnetic properties, can accommodate different cable lengths without recalibration, and yet still can fulfil all the installation, application, and performance expectations of a traditional probe.

Topics: Probes
Commentary by Dr. Valentin Fuster
1996;():V005T15A004. doi:10.1115/96-GT-032.
FREE TO VIEW

The determination of gas turbine engine performance relies heavily on intrusive rakes of pilot tubes and thermocouples for gas path pressure and temperature measurement. For over forty years, Kiel-shrouds mounted on the rake body leading edge have been used as the industry standard to de-sensitise the instrument to variations in flow incidence and velocity. This results in a complex rake design which is expensive to manufacture, susceptible to mechanical damage, and difficult to repair.

This paper describes an exercise aimed at radically reducing rake manufacture and repair costs. A novel ’common cavity rake’ (CCR) design is presented where the pressure and/or temperature sensors are housed in a single slot let into the rake leading edge. Aerodynamic calibration data is included to show that the performance of the CCR design under uniform flow conditions and in an imposed total pressure gradient is equivalent to that of a conventional Kiel-shrouded rake.

Commentary by Dr. Valentin Fuster
1996;():V005T15A005. doi:10.1115/96-GT-033.
FREE TO VIEW

The aero-gas turbine has undergone significant development to improve the thrust to weight ratio. These developments have increased the engine’s capability to withstand elevated temperatures which corresponds to both improved performance and increases in hot section creep. In addition, as the engine’s condition degrades over time, the gas generator’s rotational speed, free body stresses, and the hot gas temperature are further increased. To maximize the potential safe life of an engine and reduce operating costs, understanding the effects of engine component degradations on component life is essential. This paper examines the effects of engine degradation on the failure mechanism of creep, and presents a method in which the metal temperatures can be tied to transient parameters. The use of transient parameters enables engine stresses and temperatures to be determined at any operating point, which is essential for military aircraft engines.

Commentary by Dr. Valentin Fuster
1996;():V005T15A006. doi:10.1115/96-GT-034.
FREE TO VIEW

In a previous paper (Hansen et al., 1995), a conceptual framework for developing a true prognostic or predictive diagnostic capability was described. The current paper expands on this framework by describing micro-mechanical and dynamic models, sensors and data fusion, signal processing, approximate reasoning, distributed architecture, and human factors research and development being conducted to provide such a capability for a broad range of applications. These include both autonomous and man-in-the-loop decision making about maintenance actions and local and geographically distributed monitoring and data analysis architectures.

Commentary by Dr. Valentin Fuster
1996;():V005T15A007. doi:10.1115/96-GT-035.
FREE TO VIEW

Condition-based maintenance (CBM) is an emerging technology which seeks to develop sensors and processing systems aimed at monitoring the operation of complex machinery such as turbine engines, rotor craft drive trains, or industrial equipment. The goal of CBM systems is to determine the state of the equipment (i.e., the mechanical health and status), and to predict the remaining useful life for the system being monitored. The success of such systems depends upon a number of factors including: (1) the ability to design or use robust sensors for measuring relevant phenomena such as vibration, acoustic spectra, infrared emissions, oil debris, etc.; (2) real time processing of the sensor data to extract useful information (such as features or data characteristics) in a noisy environment and to detect parametric changes which might be indicative of impending failure conditions; (3) fusion of multi-sensor data to obtain improved information beyond that available to a single sensor; (4) micro and macro level models which predict the temporal evolution of failure phenomena; and finally, (5) the capability to perform automated approximate reasoning to interpret the results of the sensor measurements, processed data, and model predictions in the context of an operational environment. The latter capability is the focus of this paper. Although numerous techniques have emerged from the discipline of artificial intelligence for automated reasoning (e.g., rule-based expert systems, blackboard systems, case-based reasoning, neural networks, etc.), none of these techniques are able to satisfy all of the requirements for reasoning about condition-based maintenance. This paper provides an assessment of automated reasoning techniques for CBM and identifies a particular problem for CBM, namely, the ability to reason with negative information (viz., data which by it’s absence is indicative of mechanical status and health). A general architecture is introduced for CBM automated reasoning, which hierarchically combines implicit and explicit reasoning techniques. Initial experiments with fuzzy logic are also described.

Commentary by Dr. Valentin Fuster
1996;():V005T15A008. doi:10.1115/96-GT-101.
FREE TO VIEW

Multiple pressure data acquisition has typically consisted of hundreds of pipes connected via a test-bed interface to mechanical multiplexers and/or to groups of electronically scanned sensors. This paper describes a new method of high speed digital pressure scanning with intelligent pressure modules and its consequent impact on turbine testing methods and working practices. Also discussed is the remote calibration of modules and the concept of pre-rig or build shop connecting pressure tappings direct to the test measuring equipment, thus reducing commissioning, calibration and down time on the test-bed.

Commentary by Dr. Valentin Fuster
1996;():V005T15A009. doi:10.1115/96-GT-102.
FREE TO VIEW

The exploitation of different unsteady quantity measurements for identifying mechanical alterations on a radial compressor with a vaned diffuser is examined in this paper. Measurements of sound emission, casing vibration and unsteady inner wall pressure are performed. The mechanical alterations considered have been chosen in order to reproduce or simulate faults in the compressor. They include the insertion of an inlet obstruction, an obstruction in a diffuser passage, variation of impeller tip clearance, and impeller fouling. Processing these measurement data leads to the derivation of fault signatures which can be utilized for identifying them. The suitability of measuring each of the above physical quantities is discussed with respect to their sensitivity to particular faults. The dependence of the fault signatures on operating point is also examined. It’s demonstrated that minor faults which do not affect compressor operation and are not detectable by performance monitoring, can possibly be detected by the proposed methodology.

Topics: Compressors
Commentary by Dr. Valentin Fuster
1996;():V005T15A010. doi:10.1115/96-GT-103.
FREE TO VIEW

A method for analysing the circumferential temperature pattern measured at the turbine exit of Gas Turbines is presented. The purpose of the analysis is to connect the form of the temperature pattern to the condition of the hot section, and more particularly, to burner malfunctions. The behaviour of temperature patterns for different conditions of operation is discussed, on the basis of available experimental data. It is shown that with appropriate formulation, even small variation of the temperature pattern can be detected, and the particular malfunctioning burners can be identified. The proposed method exhibits a very high sensitivity to even small changes in burner operating condition, giving thus the possibility for improving corresponding techniques currently in use.

Commentary by Dr. Valentin Fuster
1996;():V005T15A011. doi:10.1115/96-GT-104.
FREE TO VIEW

A simplified scheme for scheduling multivariable controllers for robust performance over a wide range of plant operating points is presented. The approach consists of scheduling only the output matrix of a dynamic controller, thus significantly reducing the number of parameters to be scheduled. Given a robust controller at a nominal design point, designed such that it gives a stable closed-loop system at various off-design operating points, the parameters of the controller output matrix are optimized such that the closed-loop response at the off-design points closely matches the design point closed-loop response. The optimization problem formulation for the synthesis of controller scheduling gains is discussed. Results are presented for controller scheduling for a turbofan engine for a conceptual Short Take-Off and Vertical Landing aircraft. The simplified controller scheduling is shown to provide satisfactory response for engine models corresponding to significant gross thrust variations from the nominal design point.

Commentary by Dr. Valentin Fuster
1996;():V005T15A012. doi:10.1115/96-GT-105.
FREE TO VIEW

Throughout the 20th century, the evolution of control systems has proceeded from extremely simplistic systems to those with large complicated architectures and strong centralized control. Recently, the advantages of more distributed control architectures have been increasingly recognized, including even closing the loops at the actuation locations. This trend toward more distributed control systems has been mirrored in the new field of so-called “smart” structures in which the enhanced performance, multi-functionality and efficiency of biological constructs (arising from their ability to adapt and the distributed hierarchical nature of their organization) is emulated in new advanced structural designs. Considerable work has been carried out to develop fiber optic technology for smart structures applications. Recent developments in fiber optic communications and in fiber optic sensing indicate that fiber optic technology will be very important to the successful development of many of the smart structures of the future. It has long been recognized that one of the key functionalities required for a structure to be termed smart is an ability to sense, i.e. if the results of an action cannot be observed, closed loop control is impossible. For advanced control purposes, the use of long gauge length fiber optic sensors configured as matched filters for the preprocessing of large amounts of data appears to offer considerable potential benefit. In this paper, a formal definition of a smart structure is provided and an overview of the technologies involved in the smart structures area is given. Potential beneficial use of these technologies in advanced control systems is then discussed with specific emphasis placed on the appropriate application of fiber optic technology.

Commentary by Dr. Valentin Fuster
1996;():V005T15A013. doi:10.1115/96-GT-106.
FREE TO VIEW

Several organizations, including Westinghouse, CREE, and ATM, as well as researchers in Japan and Europe, are working to develop SiC power devices for reliable, high power and high temperature environments in military, industrial, utility, and automotive applications. Other organizations, such as NASA Lewis and several universities, are also doing important basic work on basic SiC technology development. It has been recognized for two decades that the superior properties of SiC lead to range of devices with higher power, greater temperature tolerance, and significantly more radiation hardness than silicon or GaAs. This combination of superior thermal and electrical properties results in SiC devices that can operate at up to ten times the power density of Si devices for a given volume.

Recent research has focused on the development of vertical metal oxide semiconductor field effect transistor (VMOSFET) power device technology, and complementary high speed, temperature-tolerant rectifier-diodes for power applications. We are also evaluating applications for field control thyristors (FCT) and MOS turn-off thyristors (MTO). The technical issues to be resolved for these devices are also common to other power device structures. The present paper reviews the relative benefits of various power devices structures, with emphasis on how the special properties of SiC enhance the desirability of specific device configurations as compared to the Si-based versions of these devices. Progress in SiC material quality and recent power device research will be reviewed, and the potential for SiC-based devices to operate at much higher temperatures than Si-based devices, or with enhanced reliability at higher temperatures will be stressed. We have already demonstrated 1000V breakdown, current densities of 1 kA/cm2, and measurements up to 400°C in small diodes. The extension of this work will enable the implementation of highly distributed aircraft power control systems, as well as actuator and signal conditioning electronics for next generation engine sensors, by permitting electronic circuits, sensors and smart actuators to be mounted on or at the engine.

Commentary by Dr. Valentin Fuster
1996;():V005T15A014. doi:10.1115/96-GT-107.
FREE TO VIEW

Technology assessments during the 1980’s projected the development of advanced military fighter aircraft which would require propulsion systems which could accommodate multi-mission capability with super maneuverability. These propulsion systems would be required to provide significantly improved thrust to weight, reduced thrust specific fuel consumption, and up and away thrust vectoring capabilities. Digital electronic control systems with significantly expanded capabilities would be required to handle these multi-function control actuation systems, integration with flight control systems, and to provide fail-operational capability.

This paper will discuss the challenges that were presented to propulsion system control designers, the innovation of technology to address these challenges, and the transition of that technology to production readiness. Technology advancements will be discussed in the area of digital electronic control capability and packaging, advanced fuel management systems, high pressure fuel hydraulic actuation systems for multi-function nozzles, integrated flight propulsion controls, and higher order language software development tools. Each of these areas provided unique opportunities where technology development programs and flight prototyping carried concepts to reality.

Commentary by Dr. Valentin Fuster
1996;():V005T15A015. doi:10.1115/96-GT-108.
FREE TO VIEW

The drive power requirements for several representative actuator applications are discussed, and the possibilities of meeting those requirements with conventional silicon electronics, and also with higher temperature capability electronics are presented. For all the usual drive circuits, both normally-off power transistors and conventional diodes are essential. Silicon devices can meet these requirements, if the electronics can be maintained near room temperature. If not, then devices based on other materials systems, such as silicon carbide, are needed. Basic material properties suggest that these devices can be capable of controlling large power levels, much larger than for silicon. However, these devices are in the very early stage of development, and have been demonstrated at only low power levels. Finally, the development of efficient blue light emitting diodes in the gallium nitride materials system is discussed. This resulting investment in this system may enhance the development of higher power, high temperature nitride-based devices, offering an exciting alternate to silicon carbide.

Commentary by Dr. Valentin Fuster
1996;():V005T15A016. doi:10.1115/96-GT-146.
FREE TO VIEW

Conventional 3-hole wedge probes fail to measure the correct static pressure when operating in close proximity to a wall or boundary through which the probe is inserted. The free stream pressure near the outer wall of a turbomachine may be over indicated by upto 20% dynamic head.

This paper reports a series of experiments aimed at quantifying this so-called ‘wall proximity effect’. It is shown from a factorial experiment that probe wedge angle, stem design and free-stream Mach number all have a significant influence. The yaw angle sensitivity of wedge probes is also found to depend on the proximity of the probe to the wall of introduction. Flow visualisation studies on large scale probe models are described, and a qualitative model of the probe local flow structures is developed. This model is used to explain the near wall characteristics of the actual size wedge probes. In Part 2 of this paper, the experimental data is used to validate CFD calculations of the flow field around a wedge probe. A simple analytical model of the probe/flow interaction is developed from the CFD solutions.

Topics: Probes , Wedges
Commentary by Dr. Valentin Fuster
1996;():V005T15A017. doi:10.1115/96-GT-147.
FREE TO VIEW

An experimental study of wedge probe wall proximity effects is described in Part 1 of this paper. Actual size and large scale model probes were tested to understand the mechanisms responsible for this effect, by which free stream pressure near the outer wall of a turbomachine may be over indicated by upto 20% dynamic head.

CFD calculations of the flow over two-dimensional wedge shapes and a three-dimensional wedge probe were made in support of the experiments, and are reported in this paper. Key flow structures in the probe wake were identified which control the pressures indicated by the probe in a given environment. It is shown that probe aerodynamic characteristics will change if the wake flow structures are modified, for example by traversing close to the wall, or by calibrating the probe in an open jet rather than in a closed section wind tunnel. A simple analytical model of the probe local flows was derived from the CFD results. It is shown by comparison with experiment that this model captures the dominant flow features.

Topics: Modeling , Probes , Wedges
Commentary by Dr. Valentin Fuster
1996;():V005T15A018. doi:10.1115/96-GT-148.
FREE TO VIEW

Since their introduction, main engine fuel pumping systems for aircraft gas turbine engines have remained relatively unchanged. The main engine fuel pump has been an engine accessory gearbox driven, positive displacement pump (except for the Concorde), until recently when centrifugal pumps were introduced on Pratt-Whitney and General Electric military engines. This paper describes some of the issues which must be addressed as pumping system technology moves into the 21st century and gives a description of two programs which address these issues.

Commentary by Dr. Valentin Fuster
1996;():V005T15A019. doi:10.1115/96-GT-242.
FREE TO VIEW

A performance simulation program has been used to simulate deteriorated performance of a new augmented turbofan engine developed for fighter aircraft and generate a fault pattern library.

This fault pattern library forms a basis for both understanding degradation trends of engine usage and in developing an engine health monitoring system. An efficient health monitoring method is proposed to identify the engine faults along with measurement uncertainties and faulty instruments, and to reduce false alarms. A pattern matching method is used to discriminate the engine faults by matching the measurement patterns throughout the fault pattern library.

The comparison of this approach to conventional gas path analysis has demonstrated that this approach has comparable ability to monitor engine gas path performance degradation, and provides some capability to handle measurement uncertainties and faults. It also provides a good base for future capability in conjunction with other engine inspection and/or monitoring methods.

Commentary by Dr. Valentin Fuster
1996;():V005T15A020. doi:10.1115/96-GT-274.
FREE TO VIEW

It is shown that emissions are becoming significant performance parameters. Dry Low NOx combustion is favourably evaluated in terms of fuel consumption when compared to NOx abatement with water injection or SCR.

The conversion of two industrial gas turbines to Dry Low NOx (DLN) Combustion summarised.

Four years experience of commissioning and operation of a number of DLN gas turbines is described.

Emission levels around 25 ppmv NOx on gas has been achieved. Operation on natural gas is generally very reliable, but some initial problems were encountered.

Emission measurements and especially continuos emission monitoring have proved to be unreliable and caused a lot of trouble during the commissioning of some plants.

Commentary by Dr. Valentin Fuster
1996;():V005T15A021. doi:10.1115/96-GT-288.
FREE TO VIEW

Possible changes to F404 turbofan engine test procedures have been investigated to reduce cost and time. The current procedures specify either fixed waiting times or stable conditions before acquiring data for performance checks. These stabilization times can account for several thousand litres of fuel. A literature-supported, experimental study is described to quantify stabilized operation. A set of derived stabilization indices are compared to measured data. Possible reductions in single tests ranged from 30 to 200 seconds. Variations are presented resulting from ambient conditions, and the extent of throttle movement Fuel savings of 300 to 700 litres are estimated depending on the maintenance test and are projected to about $100K on an annual, fleet-wide basis. Heat transfer modelling and validation issues are also discussed.

Topics: Engines , Testing
Commentary by Dr. Valentin Fuster
1996;():V005T15A022. doi:10.1115/96-GT-289.
FREE TO VIEW

Rejecting, or reducing, the effect of external disturbances on process parameters is an important problem in control design. In this paper we apply multivariable control techniques to reduce the effect of input disturbances, such as variations in the line frequency, on key internal parameters of an industrial gas turbine. The parameter we are most interested in is the combustion reference temperature, an estimated variable that is used by the controller to schedule division of fuel to various fuel nozzles and determine switching points between combustion modes. The dynamic response of this parameter correlates well with the dynamic response of fuel air ratio inside the combustor. Therefore, an important step in improving combustor performance is better regulation of the combustion reference temperature. We show that the use of a multivariable controller in place of the existing decentralized controller makes the disturbance rejection problem much easier to solve. As the gas turbine is inherently a multivariable system — i.e. the inputs, fuel and air, are coupled to the outputs, power and exhaust temperature — this result is not entirely surprising. We use a frequency domain, control design technique known as the Edmund’s method. The linear models are obtained using system identification techniques. We present results from a field test of the controller (implemented on a GE Frame 7E turbine) in the form of data comparing the response of the multivariable controller with that of the existing (decentralized) controller. These results clearly show that by using a multivariable controller the effects of the external disturbances can be reduced by a factor of 3 when compared with the existing design.

Commentary by Dr. Valentin Fuster
1996;():V005T15A023. doi:10.1115/96-GT-316.
FREE TO VIEW

We show that a neural network can be successfully used in place of an actual model to estimate key unmeasured parameters in a gas turbine. As an example we study the combustion reference temperature, a parameter that is currently estimated via a nonlinear model inside the controller and is used in a number of critical mode-setting functions within the controller such as calculating the fuel-split between various manifolds. We show that a feedforward neural network using simple back propagation learning can be built for estimating combustion reference temperature. The neural network matches the accuracy of the current estimate; and it is more robust to errors in its internal parameters. This is advantageous from the point of view of implementation since a number of errors creep in when running the algorithm on a digital controller, and an estimator that is not robust with respect to such errors can degrade the performance of the whole system.

Commentary by Dr. Valentin Fuster
1996;():V005T15A024. doi:10.1115/96-GT-343.
FREE TO VIEW

The application of wavelet analysis to diagnosing faults in Gas Turbines is examined in the present paper. Applying the Wavelet Transform to time signals obtained from sensors placed on an engine, gives information which is in correspondence to their Fourier Transform. Diagnostic techniques based on Fourier analysis of signals can therefore be transposed to the Wavelet analysis. In the paper the basic properties of wavelets, in relation to the nature of turbomachinery signals, are discussed. The possibilities for extracting diagnostic information by means of wavelets are examined, by studying the applicability to existing data from vibration, unsteady pressure and acoustic measurements. Advantages offered, with respect to existing methods based on harmonic analysis, are discussed as well as particular requirements related to practical application.

Commentary by Dr. Valentin Fuster
1996;():V005T15A025. doi:10.1115/96-GT-349.
FREE TO VIEW

It is an established fact that the efficiency of a gas turbine engine has an inverse relationship with the clearance between the rotor blades and the casing (Tip Clearance, or TC). TC is an essential measurement during the testing of development engines. While commercial TC measurement systems are available, their applicability to an engine is dictated by engine size, geometry, physical accessibility and temperature distribution around the measurement region.

This paper describes the development of a TC measurement system, based on the capacitive measurement principle, which was undertaken to satisfy the application requirements of a specific class of gas turbine engines. The requirements included a relatively long and flexible cable to route the electrical signals out of the engine. The TC measurement system was successfully used during engine testing and valuable data were obtained.

Commentary by Dr. Valentin Fuster
1996;():V005T15A026. doi:10.1115/96-GT-350.
FREE TO VIEW

A fast-response probe for the measurement of total temperature in unsteady or fluctuating compressible flows is discussed. The operation of the device is based on the measurement of transient heat flux at different probe operating temperatures. The heat flux gauges used in the present investigation were thin film platinum resistance thermometers painted onto the stagnation region of hemispherical fused quartz probes. The quartz probes had a diameter of approximately 3 mm. Uncertainty estimates indicate that temperature measurements with an accuracy of better than ± 3 K are possible. As a demonstration of the accuracy and utility of the device, total temperature and convective heat transfer coefficient measurements were obtained in a number of supersonic nozzle tests, and in a compressible turbulent jet experiment.

Commentary by Dr. Valentin Fuster
1996;():V005T15A027. doi:10.1115/96-GT-445.
FREE TO VIEW

An adaptive control system for a gas turbine engine which diagnoses conditions of axial compressor faults is proposed and analyzed. Nonlinear models of the gas turbine, neural networks and genetic algorithms are used in this research. The adaptive control system minimizes the reduction in gas turbine performance deriving from non-destructive faults. In the absence of faults, the system automatically performs the optimization of the engine between overhauls. Improvements concern