Modern Gas Turbine Systems || Turbines for industrial gas turbine systems

  • Published on
    04-Apr-2017

  • View
    222

  • Download
    0

Transcript

  • Published by Woodhead Publishing Limited, 2013 188 6 Turbines for industrial gas turbine systems M. SCHNIEDER and T. SOMMER , Alstom (Schweiz) AG, Switzerland DOI : 10.1533/9780857096067.2.188 Abstract : This chapter reviews materials, design and performance issues relating to turbines in power stations. Key performance issues include aerodynamics and cooling as well as durability and damage mechanisms. The chapter the looks at key components and their performance requirements. Key words: gas turbines, aerodynamics, interfaces, damage mechanisms. 6.1 Introduction In terms of total engine output, the vast majority of the new orders for gas turbines for power generation in recent years have been for gas turbines with a power output of 120 MW or more. In fact, power generation order surveys (McNeely, 2007; Anon, 2008; Gailloreto, 2009; Haight, 2010) show that this market had contributed about two-thirds or more of total power output of gas turbine orders since 2005. In contrast, gas turbines with a power output below 60 MW have contributed at most up to a quarter of the orders, and often signifi cantly less. The focus of the following chapter will, therefore, be on this segment of the market, and the authors will attempt to give an over- view of the key features of the turbines used in this type of engine, and to point out the main differences from aircraft engines. For a historical perspective, it is interesting to look at the development of the key turbine boundary conditions since the introduction of the fi rst power generation gas turbine in 1939 (Van der Linden, 1988). The gas tur- bine pressure ratio, and with it the turbine pressure ratio, of single-shaft gas turbines with a conventional gas turbine cycle, has grown fi ve-fold in 70 years of power gas turbine history, and Fig. 6.1 shows that this growth
  • Turbines for industrial gas turbine systems 189 Published by Woodhead Publishing Limited, 2013 was nearly linear with time. While the reheat cycle engines, introduced in the 1990s with the Alstom GT24 and GT26 engines, were omitted from this picture, it should be pointed out that, as far as the low pressure turbine is concerned, these engines fi t into the general picture represented by Fig. 6.1 . The hot gas temperature at the turbine inlet has more than doubled, again following a linear trend ( Fig. 6.2 ), and the turbine exhaust mass fl ow has grown ten-fold. The development of the turbine exhaust temperature has 4 6 8 10 12 14 16 18 20 1930 1940 1950 1960 1970 1980 1990 2000 2010 Year G T pr es su re ra tio ( –) 6.1 GT pressure ratio over year. 400 600 800 1000 1200 1400 1600 1930 1940 1950 1960 1970 1980 1990 2000 2010 Year T ur bi ne in le t t em pe ra tu re ( °C ) 6.2 Turbine inlet temperature over year.
  • 190 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 been somewhat less linear ( Fig. 6.3 ) with only a low increase of exhaust temperature throughout the 1950s and 1960s, followed by two large jumps in the late 1970s and in the 1990s. These jumps are related to the develop- ment of combined-cycle power plants, where an optimisation of the plant effi ciency results in higher exhaust temperatures. Live steam temperatures around 500°C were used when combined-cycle plants fi rst became commer- cially attractive in the 1980s (although the idea of combined-cycle plants had already been discussed much earlier, see e.g. Meyer, 1939; Seippel and Bereuter, 1960). With the advent of the F-class gas turbines in the 1990s, live steam temperatures up to nearly 600°C were achieved, leading to highly effi cient combined-cycle plants. This remarkable evolution of the turbine boundary conditions was made possible by improvements in materials, manufacturing, cooling technology and aerodynamics. Advances in these areas were driven by improvements in the underlying methods, such as fi nite element methods and computational fl uid dynamics, as well as in experimental techniques, such as laboratory testing methods, that were a prerequisite for the development of advanced cooling features. The development steps in power gas turbines are closely linked to aero- industry turbines. Although the fi rst power gas turbines were introduced nearly simultaneously with the fi rst jet engines, the aircraft engine indus- try had a leading role in gas turbine development for a long time, with some 5–10 years lag between the technology introduction in commercial 250 300 350 400 450 500 550 600 650 1930 1940 1950 1960 1970 1980 1990 2000 2010 E xh au st t em pe ra tu re ( °C ) Year 6.3 Turbine exhaust temperature over year.
  • Turbines for industrial gas turbine systems 191 Published by Woodhead Publishing Limited, 2013 aircraft engines and power gas turbines. One example of this lag is the introduction of cooled turbine blades. Since the commercial introduction of the F-class engines, the difference in technology has been signifi cantly reduced and there are now some areas where some of the technology is led today by the heavy-duty gas turbine industry. Again, an example in the fi eld of cooling comes to mind: due to the stricter emissions requirements, power gas turbines generally have more uniform turbine inlet tempera- ture traverses than aircraft engines, and they use more advanced cooling technologies for vane and blade platforms. It is expected that in the future new regulations will lead to similar constraints in aircraft engines, so that low emissions technology will also be required for aircraft engines. Engine requirements and mission profi les of power generation gas tur- bines are quite different from those of aircraft engines, leading to different drivers for the turbine development. Engine weight, for example, is a crucial factor in aircraft engines. In heavy-duty power gas turbines, on the other hand, weight in itself is not a key driver and may only be considered in the sense that cost is infl uenced by the amount of material used. Other major differences that infl uence turbine design are the differences in lifetime and inspection intervals, as well as in the proportion of time spent at full tem- perature between overhaul intervals. Finally, the overall engine design and assembly process is also a key factor affecting the turbine. While multi-shaft arrangements are frequent in the so-called aero-deriv- ative gas turbines used in power generation engines up to about 100 MW, gas turbines with higher power output are exclusively single-shaft arrange- ments, and the turbines are of the axial fl ow type. With few exceptions of gas turbines with just above 100 MW output, where gear boxes are used, the rotational speed is synchronous to the electrical grid and the engines run at either 3000 or 3600 rpm, depending on the market the gas turbine is designed for. The generator acts to fi x the rotational speed, and only small deviations occur, due to relatively small grid frequency variations about the nominal value. This is, of course, very different from aircraft engines where the speed varies over a wide range with load. Integration between turbine parts and of the turbine into the engine is clearly a key topic, both for a discussion of turbine design in general and for illustrating the differences between aircraft engines and heavy-duty power gas turbines. The detailed discussion in the remainder of the chapter will, therefore, start with a chapter on integration and interfaces, before going through the individual engineering disciplines involved in turbine design. The focus will then go back into a more integrative view, in an outline of typical turbine parts and operational characteristics. The chapter will close with a discussion of current research trends expected to make their way into future power generation gas turbines.
  • 192 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 6.2 Interfaces and integration 6.2.1 Secondary air systems (SAF) Secondary air systems (SAF) provide the cooling air from the compressor to the combustor and turbine component (see Fig. 6.4 ). Besides the cooling air supply, SAFs ensure an outfl ow into the hot gas path under all operating conditions, that is, positive pressure margins. For failsafe design the pressure must be throttled and suffi cient pressure margin kept, which is a constraint on the performance of the engine. To improve engine performance, the modern SAF system may draw from a number of technology features: increased number of supply systems into the combustor and turbine • component to better match the different pressure requirements; use of recooling for improvement of combined-cycle performance, • reduced cooling air temperature where required at structural or rotor parts, control of whole engine model clearance, and operating pinch point clearance control; use of active control devices for the orifi ces, to adjust the required pres-• sure and mass fl ow requirements over load and ambient conditions; increased number of variable guide vanes in the compressor to ensure • load/pressure redistribution adjustment in relation to the compressor end pressure; advanced sealing technology. • 6.2.2 Combustor A key topic for turbine cooling is the hottest area of the turbine, that is, the turbine inlet/combustor outlet. Emission requirements have resulted in pre- mix combustion technology with low turbine inlet turbulent intensity and fl at temperature profi le to be used. To meet CO emission requirements, high tem- perature levels can be kept to the low part load regime (Eppler et al ., 2008). 6.4 Example of modern sAF turbine systems (left: SGT5–8000H, right: GT26).
  • Turbines for industrial gas turbine systems 193 Published by Woodhead Publishing Limited, 2013 Two principle combustor concepts have evolved over time, the can annular combustor (e.g. General Electric 7FA, Siemens-Westinghouse SGT6–5000) and the ring annular combustor (e.g. Alstom GT26, Siemens SGT5–4000). Reheat technology (Alstom GT26) or staged combustion offers additional emission reduction benefi ts. Common to all modern GT system interfaces is that, for premix combustion with fl at temperature profi les, the turbine end- walls (platforms, blade tips, blade shrouds, heat shields, etc.) need particular attention and differ signifi cantly from aircraft engine turbines. The combustor provides the hot gas temperature interface into the tur- bine. Modern gas turbine systems have certain characteristics that signifi - cantly impact the combustor turbine interface: Premix combustion, reheat and staged combustion in relation to aero-• engine combustor technology have small turbulence levels. Ring annular combustors, compared to can annular combustors, also feature reduced leakage introduction into the turbine interface and therefore more homogeneous fl ow conditions. Emission requirements in terms of NO • x (peak temperature driven) and CO (part load and cold ambient driven) drive vs small temperature dis- tortion in radial and circumferential direction . As a consequence, end- walls are thermally highly loaded. Hot streak migration is present to a much smaller extent. Combustor pressure drop (compressor exit to turbine inlet fl ow path) is • required to achieve suffi cient cooling pressure over the operating range, mainly for the fi rst stage vane, but has a performance penalty. Low pres- sure drop design technology is required for the front stage turbine parts. Bow wave and trailing edge wave effects of the front stage vane have to be minimised to reduce the SAF purge fl ow requirements. Combustor and turbine components form large structural parts with sig-• nifi cant thermal expansion over the operating range setting geometrical constraints to the interface. Premix combustion often requires staging to avoid thermo-acoustic pul-• sations within the combustor. Staging increases the circumferential tem- perature distortion. 6.3 Aerodynamics Perhaps the most important factor determining general turbine design is how the overall expansion in the turbine is distributed into individual tur- bine stages. The fi rst consideration is the choice of the number of stages. The impact of this choice on the aerodynamic design and the turbine effi - ciency can be understood using the Smith chart (1963), where turbine stage effi ciency is presented as a function of the fl ow coeffi cient C Uax / and the
  • 194 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 loading coeffi cient ΔH U/ 2 ( Fig. 6.5 ). The highest stage effi ciency is achieved for small values of fl ow and loading coeffi cients. Due to the fi xed rotational speed in power generation gas turbines, the blade speed U is only a function of the radius. The only other way to infl u- ence the stage loading for a given overall enthalpy drop is by the choice of the number of stages. A higher number of stages will result in higher stage effi ciencies but, due to the lower stage enthalpy drop, there will be more stages with high gas temperatures. The choice of the stage loading thus becomes a trade between effi ciency, blade cooling, and cost, as well as mechanical and structural design considerations. Note that a reduction in the number of stages does not necessarily lead to a proportional reduc- tion in parts count in the turbine, since an increase in stage loading tends to increase the blade count per stage. Historically, this trade led to vastly different choices in mean stage load- ing made by different companies. For example, there is a factor of about two between the estimated mean turbine stage loading in the fi rst power genera- tion gas turbines from Brown Boveri and General Electric (Van der Linden, 1988; Somerscales and Henricks, 1984). There continued to be a large spread up into the 1970s and 1980s, but values for stage loading appear to have converged to values typically between 1.3 and 1.5. This is probably driven by a convergence of engine application design towards gas turbines for high effi ciency combined-cycle power plants. In view of the fi ve-fold increase in turbine pressure ratio discussed in the introduction, the turbine mean stage loading has remained remarkably 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 94% 93% 92% 91% 90% 89% 88%Δ H / HH U 2 UU CaxCC /U 6.5 Smith chart.
  • Turbines for industrial gas turbine systems 195 Published by Woodhead Publishing Limited, 2013 constant in time, particularly since most companies have reduced the num- ber of stages rather than increasing them. This was made possible by a strong increase in mean blade speed, that is, mean radius ( Fig. 6.6 ). This is testament to advances in other areas, such as design, materials, durability and manufacturing of rotors. The determination of the distribution of the turbine enthalpy drop to the individual stages entails a trade with much the same considerations as the choice of the mean stage loading. On the one hand, aerodynamic effi ciency is higher when more work is done in the rear stages, both because their stage loading is lower due to the higher mean radius and less secondary fl ow and tip leakage losses. On the other hand, more work done on the front stages results in a higher temperature reduction early on in the turbine. The aero- dynamic effi ciency impact and the cooling and leakage air consumption can be can be stacked up against each other in terms of their impact on the gas turbine cycle effi ciency in a quantitative assessment. Risk aspects, such as the relative risk of having higher or lower gas temperatures in a second stage, are more diffi cult to quantify, but still play an important role in tur- bine design. Typical Smith chart locations of the fi rst and last turbine stages in mod- ern power generation gas turbines are shown in Fig. 6.7 . Turbines in these engines are all designed with a mean radius that increases from inlet to exit, such that the blade speed is higher in the last stage. Furthermore, the higher temperature in the fi rst stage leads to a higher enthalpy drop in the fi rst stage than in the last one. The fi rst stage loading is close to the numbers reported for HP turbine stages in aircraft engines, typically in the range Δ ≈UH / 2 1.5 to 2.5 (Cohen et al ., 1987), depending on the application. 100 150 200 250 300 350 400 450 500 550 1930 1940 1950 1960 1970 1980 1990 2000 2010 Year M ea n bl ad e sp ee d (m /s ) 6.6 Mean blade speed over year.
  • 196 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 The shape of the velocity triangles for a turbine stage is defi ned by the choices of the fl ow coeffi cient, stage loading and stage reaction R h h h hT = 2 3 h 1 3h The stage reaction in power generation as well as aircraft engines is usually chosen to be near RT = 0 5.5 This results in the best stage effi ciency, since RT = 0 5 implies equal exit Mach numbers in the vane and the blade. Since the entropy generated in a turbine stage is related to the square of the exit Mach numbers Δ − sΔΔ Mp v B is B/ (=cp ), ,B is χ ζ ζM +v is V +,V 1 2 2Mζ+ the optimum effi ciency is achieved for equal exit Mach numbers in the two blade rows, at least when the loss coeffi cients are equal in both rows. For fi rst stages, however, the stage reaction is often chosen to be as low as RT = 0 3 , since a lower reaction leads to a lower relative total temperature at the blade inlet. The velocity triangles for a typical fi rst stage with a stage loading and fl ow coeffi cient from Fig. 6.7 for two choices of stage reaction, RT = 0 5 and RT = 0 3 , respectively, are shown together with representative blade shapes in Fig. 6.8 . In constructing the diagrams, the usual assumptions of constant blade speed and axial velocity as well as equal stage inlet and exit velocities were made. Strictly, the latter assumption is violated in a fi rst stage where there is a strong acceleration of the fl ow from the combustor interface (or vane inlet) to the stage exit. Nevertheless, the accuracy of the estimates is 2.4 2.6 2.8 ΔH / HH U 2 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.4 0.5 0.6 Load typic al las t stage 0.7 0.8 93% 92% 91% 90%0 89%% 88% 0.9 1.0 1.1 CaxCC /U Typical fi rst stage 94% 6.7 Typical Smith chart location, fi rst stage, last stage.
  • Turbines for industrial gas turbine systems 197 Published by Woodhead Publishing Limited, 2013 still good enough for our purpose. The velocity triangle and typical blade shape for a rear stage are shown in Fig. 6.9 . Clearly, the resulting velocity triangles are quite different, with much higher fl ow turning in a fi rst stage than in a last one. There are very little published data on the profi le aerodynamics in power gas turbines. In the fi eld of aircraft engines, signifi cant advances have been made in the last 20 years. Moustapha et al . (2003), for example report an increase in blade mean Zweifel coeffi cient, from about 0.85 around 1980 to values of 1.15 in the case of uncooled blades and 1.05 for cooled blades by –2.0 –1.5 –1.0 –0.5 0.0 0.5 1.0 1.5 2.0 Cax C t, W t –2.0 –1.5 –1.0 –0.5 0.0 0.5 1.0 1.5 2.0 –1 0 1 2 3 –1 –0.5 0 1 2 3 Cax C t, W t –0.5 3.5 0.5 3.5 6.8 RrR = 0 3 velocity triangles, RrR = 0 5 and RrR = 0 3.
  • 198 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 1990. Without giving numbers for the actual lift coeffi cients, Haselbach et al . (2001) report the design and testing of a redesigned low pressure turbine where ultra-high lift blading was applied with an 11% increase in lift coef- fi cient over the previous ‘high lift’ blades. The profi les feature high suction side diffusion with a ratio of peak to trailing edge Mach number in excess of 1.3. Cascade tests of a high lift profi le, with a Zweifel coeffi cient just below 1.3 and a similar level of diffusion, are reported by Mich à lek et al . (2010). Much of the work in aircraft engine low pressure turbines is related to bet- ter understanding of boundary layer transition, including unsteady effects such as wake passing. Nevertheless, one would expect that, with the high Reynolds numbers present in power gas turbines, at least some sort of ‘high lift’ profi le could be used with high suction side diffusion, even though due to higher airfoil blockage for cooled blades, the achievable lift levels may be somewhat lower than for the thin profi les in aircraft engine low pressure turbines. Due to the lack of data, a direct comparison of the Zweifel lift coeffi cient for power gas turbines is not feasible. On the other hand, a reasonably good estimate of the ratio of axial chord to blade pitch l sasl / , or axial solidity, is possible. Dyschlevskij, in his 1965 PhD thesis, derived a correlation for the optimum blade spacing accounting for the relative blade thickness t lmatt x / as well as the blade turning (Abianz, 1979). s l k t l opt = ⎛ ⎝⎜ ⎛⎛ ⎝⎝ ⎞ ⎠⎟ ⎞⎞ ⎠⎠ −⎛ ⎝ ⎛⎛⎛⎛ ⎝⎝ ⎛⎛⎛⎛ ⎞ ⎠ ⎞⎞⎞⎞ ⎠⎠ ⎞⎞⎞⎞180 1 3 2 3 1 3 β β+2 32 3+ β22 β33 cos cos / matt x Dyschlevski’s correlation was translated into the corresponding value for axial solidity using the relations between true chord, axial chord and inlet –1.55 –1.0 –0.5 0.0 0.5 1.0 1.5 –1 0 1 2 3 Cax C t, W t –0.5 1.5 3.5 6.9 Last stage blade velocity triangles.
  • Turbines for industrial gas turbine systems 199 Published by Woodhead Publishing Limited, 2013 and exit angles from Cohen et al . (1987). Isolines of optimum axial solidity as a function of inlet and exit angle for a relative thickness of t lmatt x / .l 2. are shown in Fig. 6.10 . Referring back to the typical range of fi rst stage blade inlet and exit angles discussed earlier, the optimum axial solidity is roughly l saxl / .s 1 3. for t lmatt x / .l 2. . For a lower blade thickness, less solidity is required, for example, l saxl / .s 1 1. 5 for t lmatt x / .l 0 1. . For fi rst stage vanes, the optimum solidity would be below l sasl / .s 7. . The axial solidity for the different blade rows of current F-, G- and H-class turbines is shown in Fig. 6.11 . The observed values are close to the optimum values given by the correlation from Dyschlevski, dating back to 1965, indi- cating that even modern turbines are using quite conservative values of lift coeffi cients. This is confi rmed by the few publications, such as Bahls (2002) discussing profi le lift distributions where the suction side diffusion is char- acterised by a ratio of peak to exit Mach number of below 1.25. It is clear that one driver calling for high lift blading in aircraft engines, that is the impact on engine weight, plays no role in the design of a power gas turbine. Thus, it appears that the turbine designers came to the conclusion that the drawbacks of the higher lift in terms of higher secondary fl ow losses and the impact on the tip leakage (Moustapha et al ., 2003) outweigh the benefi t of a lower blade count. The axial solidity in the rear stages is also higher than required from Fig. 6.10 . Here the blade chord is driven by mechanical and frequency requirements more than by aerodynamics. A detailed discussion of the different stacking options for three-dimen- sional (3D) aerodynamic design can be found in Denton and Xu (1999). Haselbach et al . (2001) show an application of 3D stacking to an aircraft 0 10 20 30 40 50 60 70 80 Exit angle (°) In le t a ng le ( °) 1.5 1.4 1.3 1.2 1.1 0.81.0 0.7 0 0. 0.9 0. 0 40 50 60 7000 80 6.10 Optimum axial solidity with typical angles for fi rst stage vane and blade and last stage blade.
  • 200 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 engine low pressure turbine, where compound lean is applied to the vanes. For power gas turbines, only two applications of signifi cant compound lean have been reported (Aoki, 2000), where two generations of 3D stacking were applied to the vanes in the 501F and 501G engines from Mitsubishi, respectively, and their counterparts from the Siemens SGT6–5000 series. The exhaust diffuser design has a signifi cant impact on turbine perfor- mance. With increasing turbine exit Mach number, the total pressure loss in the diffuser for any given recovery factor cp increases as shown in Fig. 6.12 , thus reducing the turbine pressure ratio. In terms of overall gas turbine performance impact a 1% increase in total pressure loss is equivalent to 0.2–0.3% turbine effi ciency reduction. Early on in gas turbine development, the exhaust mass fl ow was usually small relative to the available area, thus leading to turbine exit Mach numbers well below 0.3 ( Fig. 6.13 ). By the mid- to late 1960s, the exit Mach numbers had risen above 0.3, and diffuser design started to take more importance (Seippel, 1966). For the lower Mach num- bers, a rather low area ratio in the exhaust diffuser was acceptable. This then led to simple diffuser designs resembling a sudden expansion. The reason is clear when looking at the pressure recovery plotted against the inlet-to-exit area ratio in Fig. 6.14 . The fi gure gives both the ideal recovery and the actual recovery for a sudden expansion. For ratios of inlet-to-exit area of 0.7 and above, the difference in pressure recovery becomes quite small and there is no real benefi t in going to more elaborate designs. However, for ratios in the region of 0.2, as frequently used in modern gas turbines, the differ- ence becomes very large. In terms of aerodynamics, the layouts of current exhaust systems of gas turbines from different companies are quite similar. In a fi rst section, an annular diffuser with a small area increase is located. Here a number of struts are located that contain a support structure for 0.6 0.8 1.0 1.2 1.4 1.6 1.8 V1 B1 V2 B2 V3 B3 V4 B4 Vane/blade row A xi al so lid ity 6.11 Axial solidity of vanes and blades in F-, G- and H-class turbines.
  • Turbines for industrial gas turbine systems 201 Published by Woodhead Publishing Limited, 2013 the bearing, arranged radially or leaned, and often profi led to reduce the swirl exiting the turbine. Following this so-called exhaust gas housing, an annular diffuser is located, typically making up about half the area ratio. At the end, there are again supporting struts, and a sudden expansion is located immediately downstream, before an equalising section fi nally either takes the fl ow to the HRSG or the fl ow is turned into the stack. The sudden expansion is sometimes made less severe by a cone at the inner diameter, 0 1 2 3 4 5 6 7 8 9 10 11 0.3 0.5 0.7 0.9 Turbine exit Mach number T ot al to s ta tic p re ss ur e lo ss ( % ) cp=0.90 cp=0.75 cp=0.70 cp=0.85 cp=0.80 6.12 Diffuser pressure loss as function of turbine exit Mach number. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 1930 1940 1950 1960 1970 1980 1990 2000 2010 Year E xi t M ac h nu m be r 6.13 Turbine exit Mach number over time (year).
  • 202 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 while at other times the inner diameter at the end of the bearing tunnel ends abruptly (Brummel, 2002). Examples of a diffuser from 1966 and a modern one are shown in Fig. 6.15 . Modern diffusers typically have an area ratio of about 0.2 from turbine exit to the equalising section, and reach a pressure recovery in the region of 0.75. 6.4 Cooling Major development drivers and achievements for GT development have been power output, GT effi ciency and reduction in emissions. These perfor- mance achievements have been mainly realised through improvements in the detailed component design (Figs 6.1–6.3). For the turbine component, 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 Inlet area/exit area (–) C p (– ) 6.14 Diffuser pressure recovery vs are ratio; ideal diffuser and sudden expansion. 6.15 Typical diffuser arrangement 1966 and modern GT.
  • Turbines for industrial gas turbine systems 203 Published by Woodhead Publishing Limited, 2013 the evolution beyond the performance and aerodynamic evolution can be attributed to: Increase in hot gas temperature capability through the introduction of • new materials, coatings and cooling technology (Figs 6.16–6.18). This allows the spread in between the driving gas temperature that the part sees and the metal temperature that is achieved through active cooling to be increased. The constraint for the metal temperature is imposed by lifetime requirements to be achieved. The allowable metal temperature has been raised through the introduc-• tion of new materials, the so-called super alloys. Introduction of ceramic coatings that function as thermal barrier coat-• ings, as well as oxidation protection coatings to protect the mechanical load carrying base metal super alloys. Manufacturing technology that allows complex geometrical features to • be realised mainly through investment castings, thermal sprayed coat- ings and three-dimensional machining operations. Simple convective cooling schemes have been introduced during the mid-1960s to the 1970s (Figs 6.16 and 6.18). The manufacturing technol- ogy used was mainly based on forging and machining. Aero-engine tech- nology available by then allowed an accelerated introduction of casting 2600 2400 2200 2000 1800 1600 1400Tu rb in e en tr TT y te m pe ra tu re ( °K ) 1200 1000 1950 1960 1970 1980 YearYY 1990 2010 Introduction of blade cooling Sophisticated cooling systems Film impingement convection Convection Allowable metal temperature Uncooled turbines Transpiration and others New cooling concept Pr oje cte d tre nd ne w m at er ial Simple cooling 6.16 Aero-engine cooling technology chart.
  • 204 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 technology. Advanced convective schemes using means of impingement cooling for static parts followed in the late 1970s and the beginning of the 1980s. Improvements in secondary air systems, such as pre-swirl nozzle systems and advanced sealing technology, allowed having a positive BFM for the front stage static and rotating parts. Subsequently, fi lm cooling technology was introduced, fi rst on the front stage vanes leading edges and later on the rotating blades. Advancements in manufacturing and materials technology led to the introduction of investment castings (ceramic cores), fi rst for directionally 4000 3800 3600 1920ºC 2204ºC D es ire d tr en d (I hp te t) 1648ºC Composite ceramics Carbon–carbon Metal matrix composites Ceramics Fibre–reinforced superalloys Ods superalloys1093ºC Conventionally cast Eutectics Single crystals Directionally solidified superalloys 3400 3200 3000 2800 2600 2400 2200 1371ºC Thermal-barrier coatings 2000 1800 1600 1400 M at er ia l s ur fa ce te m pe ra tu re , ( °F ) 1950 1960 1970 1980 YearYY 1990 2000 2010 2020 6.17 Aero-engine evolution of material surface temperature over time.
  • Turbines for industrial gas turbine systems 205 Published by Woodhead Publishing Limited, 2013 solidifi ed materials and later for single crystal alloys. Investment casting opened the opportunity of using multi-pass cooling systems. Initially, area enhancing features were used, and towards the end of the 1970s angled rib turbulators appeared in the open literature (Han, 1978) to improve the heat transfer in between the bulk cooling fl ow and the near wall fl ows. Angled rib turbulators introduce longitudinal vorticity into the cooling fl ow path, and thereby enhance the fl ow exchange. Film cooling technology, introduced in the beginning of the 1980s, was cylindrical in shape fi rst. With the F-class engines, fan shaped cooling hole technology through laser or EDM hole drilling was introduced. Due to unavoidable mismatch in pressure between the cooling and the hot gas fl ows, the large momentum of the fi lms led to fi lm lift off from the metal surface and loss of fi lm cooling protection. Fan shaped cooling holes reduce the momen- tum and avoid fi lm lift off, and thereby improve fi lm cooling effectiveness. Thermal barrier coatings, which had fi rst been tested at NASA in 1978 (Miller, 1997) began to be commercially available for the heavy-duty gas turbine industry in the beginning of the 1990s. Detailed data are difficult to obtain and difficult to generalise, since turbine design parameters, such as reaction vs impulse turbine scheme selection or materials used, may influence the specific requirements when the technology elements are needed. Estimates are given subse- quently. Turbines with inlet temperatures of up to 850°C can remain uncooled. At around 875°C vane cooling and, above 900°C, first stage blade cooling is required. Temperature levels above 1100°C require first stage vanes to be cooled with showerhead film cooling and the second stage to be convectively cooled depending on the first stage 0.9 C oo lin g ef fe ct iv en es s 0.8 0.7 1970s 1960s 1980s 1990s Cooling efficiency 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.6 0.5 0.4 0.3 0.2 0.1 0 0 1.0 2.0 Cooling flow (m*) 3.0 6.18 Aero-engine cooling effectiveness over heat load parameter.
  • 206 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 work extraction. The first F-class engine (GE 7FA turbine), which was shipped in 1988, was stated to be rated at 1250°C. Many cooling tech- nology features introduced in the 1990s were first published with the NASA US E 3 -aero-engine programme (Lakshminarayana, 1996). Cooling technology is commonly classifi ed using a set of parameters ( Fig. 6.18 ), which are subsequently explained. Φ = ( ) ( ) − − Cooling effectiveness Cooling effectiveness is defi ned as a normalised metal temperature. The metal temperature is normalised with the cooling air feed temperature and the driving hot gas temperature. At a value of one, the metal temperature is at the cooling air temperature, at a value of zero the metal temperature is at the hot gas temperature. The parameter defi nes the cooling requirements, that is, highly cooled versus weakly cooled. ηcη = ( )c cT Tc c ( )m cT Tm c , ,c , Cooling efficiency The cooling effi ciency refers to that of the internal heat exchanger. The average cooling air temperature at which the cooling fl uid leaves the cooled part is normalised with the cooling air feed (cooling inlet temperature to the part) and the average metal temperature. At a value of one, the cooling air leaves the cooling system at the metal temperature, that is, maximum heat pick-up, maximum potential used. At a value of zero the cooling fl uid leaves the cooling system at the cooling air inlet temperature, that is, no heat pick- up. It defi nes the cooling technology level achieved, that is, well cooled or badly cooled. m c m A p c c* = ⋅HG HGHTC Heat load parameter (Mass flow function))) The heat load parameter (sometimes called the mass fl ow function) sets the heat transported by cooling medium in relation to the heat input into blade. ηfilmη = ( )film HG ( )HG film H− H− Film cooling effectiveness The fi lm cooling effectiveness is defi ned as a normalised fi lm temperature. The fi lm temperature is normalised with the cooling air feed temperature and the driving hot gas temperature. At a value of one the fi lm temperature is at the cooling air temperature, at a value of zero the fi lm temperature is at the
  • Turbines for industrial gas turbine systems 207 Published by Woodhead Publishing Limited, 2013 hot gas temperature. The parameter defi nes the effectiveness of the cooling fi lm, that is, non-mixed-out fi lm vs mixed-out fi lm. Above certain temperature levels, fi lm cooling is required for reducing the hot gas temperature to the fi lm temperature. Without fi lm cooling the driving hot gas temperature is limited to what is achievable by convective cooling. With thermal barrier coatings the base metal interface temperature requirements remain similar, but the cool- ing mass fl ows can be reduced. The four parameters mentioned above are interlinked. The cooling effec- tiveness is an engine-level constraint determined by the chosen cycle and part material. The higher the cooling effectiveness requirements are, the higher the cooling effi ciency that can be achieved. A large driving temper- ature difference allows more heat to be taken out. At a certain limit, fi lm cooling is required, since convective cooling is limited. Thinner walls signifi - cantly improve the convective cooling effi ciency achievable. In practice for some designs a local cooling effectiveness to be achieved drives the total cooling fl ow amount, that is, the whole blade needs to be fl owed to achieve the cooling requirements at a local geometry. On the other hand, bringing cooling fl uid directly to such a location may result in no heat pick-up of the cooling fl uid and consequently a low cooling effi ciency. Using advanced cooling technology measures, this can be overcome, for example, near wall cooling, wall cooling, and transpiration cooling. Figure 6.16 shows the evolution of blade cooling systems for aero-engines. Figure 6.18 shows a corresponding cooling effectiveness chart. Typical cool- ing schemes of a heavy-duty gas turbine from the beginning of the 1980s are shown in Fig. 6.19 . 6.5 Durability and damage mechanisms The main damage mechanisms for turbine blades and vanes are creep, high cycle and low cycle fatigue, as well as oxidation. Creep occurs when a component is exposed to high loads at high temperature for a long time, resulting in a time-dependent deformation (Webster and Ainsworth, 1994). Under the stresses prevalent in turbines, creep can be split into three phases: primary, secondary and tertiary creep (Dowling, 1999; Webster and Ainsworth, 1999). In the primary creep phase, the creep rate decreases in time. In the secondary phase, the strain rate is approximately constant and fi rst voids start to form. In the tertiary strain phase, the strain rate increases progressively in time and creep cracks start to develop, eventually lead- ing to rupture failure. Creep damage is accelerated with higher mechanical loads and with higher metal temperature ( Fig. 6.20 ). For a safe design it is generally considered that creep should remain within the secondary phase throughout the lifetime of a blade. Creep can lead to failure before rupture is reached. Excessive creep deformation can, for example, lead to a clash
  • 208 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 Third stage vane Second stage blade and vane First stage blade and vane 6.19 Beginning of 1980s Alstom GT13E: Typical fi ve stage turbine layout out and corresponding cooling systems. II Secondary I Primary III Tertiary ε I II I σ,σσ Τ ε t t II σ = const T = constT ε = constε 6.20 Creep strain relation.
  • Turbines for industrial gas turbine systems 209 Published by Woodhead Publishing Limited, 2013 with neighbouring parts, to the loss of shroud coupling or the failure of a sealing feature to perform its duty. In fatigue, cracks can form under repeated application of a stress that is well below the tensile strength. High cycle fatigue is generally charac- terised by low strain within the elastic range, but with a high number of cycles. High cycle fatigue in turbines is typically caused by vibrations with vanes or blades either running in resonance to some outside excitation or by self-excited vibrations in the case of fl utter. Low cycle fatigue, on the other hand occurs after relatively few cycles covering a high strain range leading to elastic–plastic material behaviour. In thermal fatigue, the stresses are generated by local temperature gradients, which vary between idle, part load, base load and engine off conditions. The strain range covered through a transient cycle may be larger than that between the two end states of the cycle, depending on the ratio of the time scale of the change to the char- acteristic thermal time scale of a part. For example, a rapid increase in hot gas temperature will initially lead to a high temperature gradient between the blade surface and the remainder of the wall and, therefore high thermal stresses. In time thermal conduction will even out the temperature gradient so that the thermal stresses in the end state may be lower than during the transient (Moustapha et al ., 2003). Oxidation occurs when metal is exposed to high temperatures and a reac- tion between the metal surface and the remaining oxygen in the hot gas path takes place. This reaction leads to a reduction of wall thickness. Coatings are usually applied to reduce oxidation. Even though these coatings have higher oxidation temperatures than the base metal, coating becomes depleted over time, at which point base metal oxidation will start ( Fig. 6.21 ). Typical areas where oxidation is a major concern are the tips of unshrouded blades, where at best excessive amounts of cooling air may be required to suffi ciently reduce the metal temperature. At worst, certain areas cannot be cooled suf- fi ciently at all. With the exception of high cycle fatigue, all these damage mechanisms are strongly dependent on the metal temperatures at the critical locations. Table 6.1 shows the sensitivities to metal temperature from Dailey (2000). The life of a blade is doubled or halved with a change of 15°C and 30°C, depending on the failure mechanism, highlighting the importance of accu- rate temperature predictions. An understanding of the importance of stresses due to centrifugal loads can be obtained by looking at the parameter AN 2 , which correlates to the hub stress of blades for a given taper ratio. In aircraft engine high-pressure turbines, AN 2 is typically in the range 5000 90002 ≤2< AN (Moustapha et al ., 2003) whereas in modern power gas turbines usually 2000 45002 ≤2< AN . In last stage blades of power gas turbines, on the other hand, values of AN 2 15 000≈ and higher are reached, whereas the typical range for aircraft engine low pressure turbines is
  • 210 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 11000 140002 ≤2< AN . At the same time, the last stage turbine blades in power gas turbines are running at much higher temperature levels since the exhaust temperatures ( Fig. 6.3 ) are 150–200°C higher than in aircraft engines. As a consequence, the stresses in aircraft engine high-pressure turbine blades are typically about evenly split between centrifugal stresses and ther- mal stresses. In power gas turbines, on the other hand, thermal loads account for about 90% of the stresses in a fi rst stage blade. Due to the lower cen- trifugal loads, blade creep is usually not an issue in such turbine blades and, due to the fi xed rotation speed, high cycle fatigue can be avoided by tuning the blade eigenfrequencies to be out of resonance with known sources of excitation, although blades can briefl y run through resonance at start-up and shut-down. Thus, the main concerns are oxidation and thermal fatigue. 25 000 20 000 15 000 10 000Li fe ti m e (h ) 5000 0 0 25 50 125 150 175 200 225 Coating thickness (μm) 250 275 300 325 350 375 400 1100ºC 1075ºC 1050ºC 1025ºC 1000ºC 975ºC 950ºC925ºC900ºCSV 20 Overlay 75 100 6.21 Oxidation of overlay coating. Table 6.1 Failure mechanism and sensitivity to metal temperature at critical locations Mechanism Metal temperature change to double/halve life Creep 15°C Corrosion (oxidation and sulphidation) 20°C Low cycle fatigue 30°C High cycle fatigue (vibration) Not primarily temperature driven
  • Turbines for industrial gas turbine systems 211 Published by Woodhead Publishing Limited, 2013 In the rear stages, on the other hand, creep becomes the relevant damage mechanism in turbine blades for power generation. Power gas turbines differ from aircraft engines not only in the major damage mechanisms, but also in their lifetime requirements and operating concepts. In order to keep within emission regulations at part load (CO), combined-cycle part load effi ciency at higher part load and in order to keep temperatures suffi ciently high for the water steam cycle, power gas turbines operate with the turbine inlet temperature at or near the base load values throughout much of the operating range (Bauer and Rofka, 2002; Eppler et al ., 2008). Thus, there is roughly an order of magnitude difference in terms of time spent at peak temperature between aircraft engines and power gas turbines (Bunker, 2004). With the increased fl exibility required by the cur- rent market, gas turbines will run anywhere from daily start–stop cycles to long periods of continuous base load, where continuous operation for half to three-quarters of a year is not uncommon. As noted above, high cycle fatigue due to resonance can usually be avoided by appropriate frequency tuning. However, with the high mass fl ow in modern power gas turbines, last stage blades become very long, with airfoil heights up to about 0.8 m. With increasing airfoil height, the eigenfrequency of the blade will drop. Once frequencies are below a cer- tain threshold, self-excited aeroelastic vibrations may occur, leading to high cycle fatigue damage. Typically a blade design is deemed to be safe if the Strouhal number Str U( )fLff / for the fi rst bending mode is above Str ≈ 0 3 to Str ≈ 0 35 , depending on the source. Higher values above Str ≈ 1 are required for the fi rst torsional mode. Since it is diffi cult to achieve suf- fi ciently high frequencies to reach these Strouhal numbers, most engine manufacturers have chosen to use shroud coupling on their rear stages. Shroud coupling will increase the frequency and, with suffi cient shroud stiffness, reduce the torsional component of the eigenmode. Siemens, as the only exception, at least for engines of the E- and F-class, have used hollow blades with a long axial chord, raising both the frequency due to stiffness at the inner diameter and increasing the Strouhal number with a higher tip chord. Typical Campbell diagrams for a free standing and a coupled blade are shown in Fig. 6.22 . Here, blade vibrations were recorded via strain gauge measurements during start-up and over-speed in a spin pit test. In the free standing blade, there is a single eigenmode per family of modes for which the frequency increases with rotational speed due to cen- trifugal load stiffening. Thus the frequency recorded in this test is increas- ing with speed. For the coupled blade, each modal family has a different eigenmode and frequency for each possible nodal diameter or inter-blade phase angle. When running through a resonance, the nodal diameter num- ber and the number of excitations per revolution coincide. Thus, the higher nodal diameters reach resonance early in the run-up. Since higher nodal
  • 212 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 diameters have higher resonance frequencies, there is a general trend of lower eigenfrequencies with higher speed. 6.6 Typical parts and interfaces In the following section a brief summary is given on cooling features and turbine parts commonly used in today’s modern gas turbine systems, and some key areas of concern. On each of the features, a large number of pub- lications exist (Han et al ., 2000): The leading edge of an airfoil is of particular interest, since optimised • aerodynamic design requires a small radius, which results in a high heat load in the leading edge laminar stagnation region due to its thin bound- ary layer. For many applications the leading edge is a critical area of concern, with showerhead fi lm cooling applied on the front stages to reduce the heat load. Today’s manufacturable wall thicknesses limit the range where pure con-• vective cooling can be applied, and fi lm cooling technology is a common feature to reduce the heat load. For fi lm cooling a secondary cooling fl uid is introduced into the main stream gas and kept attached to the exposed walls to protect the walls. Slot fi lm cooling does not meet mechanical requirements, such that discrete hole fi lm cooling is required. Many stud- ies have been conducted to avoid the disturbance of the thin protecting fi lm and achieve protection signifi cantly downstream from the place of injection ( s / d > 50). The intention is to have as little mixing as possible between the cooling and the mainstream fl ow. The injection locally dis- turbs the thermal boundary layer and therefore results in a heat trans- fer coeffi cient (HTC) enhancement, which has to be balanced with the temperature effect benefi t of the fi lm. Film cooling holes need to have a positive outfl ow, including margin (often called back fl ow margin – BFM), 6.22 Typical Campbell diagram for uncoupled and coupled blades.
  • Turbines for industrial gas turbine systems 213 Published by Woodhead Publishing Limited, 2013 under all operating and transient conditions (e.g. vane wake) and there- fore impose constraints on the layout of a cooling system. An area of par- ticular importance is the leading edge area where the fi lm cooling holes cannot be angled into the fl ow direction of the mainstream gas (typical values 30°) and need to be pointed against the mainstream. Generally, the cooling system leading edge region is reliant on the fi lm cooling hole internal convective cooling effect. This effect is of less benefi t if the fi lm effect is increased. The leading edge hole length can be increased to raise the convective effect by introducing an angle away from the GT machine axis, often called compound fi lm cooling. For turbine internal convective cooling, the following methods are • commonly used, or a combination thereof: multi-pass cooling, jet impinge- ment cooling, rib-turbulated cooling and pin-fi n turbulated cooling. Many more features have been examined as for example vortex generators, dimples and others. A summary can be found in Han et al . (2000). Multi-pass cooling is a method that was introduced when the manufactur-• ing method of investment casting offered a technological benefi t in terms of cooling effi ciency increase compared to the formerly used forg- ing method. Forging mostly resulted in single pass cooling channels, and connecting two with a connecting bend and turning the fl ow towards its original fl ow direction allowed the reuse of the cooling fl uid and is well suited to form a geometry useable for a ceramic core. Still, the hole blade cross-section has to be fl owed, putting constraints on the achievable internal Ma number at a given amount of cooling mass fl ow and there- fore the achievable HTC. Increasing the number of multi-pass channels can increase the achievable HTC levels (Dailey, 2000) and thereby the cooling effi ciency. Walls that are not connected to the hot gas side (often called webs) remain relatively cold. Increasing the number of passages results in increased channel aspect ratios and results in low bulk fl ow exchange with the hot near wall fl ow, and thus limits the achievable cool- ing effi ciency. Jet impingement cooling ( Fig. 6.23 ) with cast-in holes, or with holes • formed by an insert, allows the hot near wall fl ow region to be cooled by an internal jet fi lm (Goldstein, 2000 ). Thus, not the whole cross-section needs to be fl owed, and the achievable internal HTC-levels/cooling effi - ciency tends to be higher. The use of inserts is limited for rotating parts. For static parts, inserts are commonly used but put a constraint on the allowable 3D aerodynamic external shape, since inserts need to able to fi t in. Recent developments have therefore been undertaken to cast-in the jet impingement cooling holes not only into the split plane of the core (required to separate the core die tool after injection) but also in other areas. After the use of the cooling fl uid for jet impingement cool- ing, the fl uid is often reused as fi lm cooling fl uid subsequently.
  • 214 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 Rib-turbulated cooling ( Fig. 6.23 ) was fi rst used as roughness elements • by placing them normal to the fl ow direction or as area enhancement ribs to increase the internal heat exchanger area onto the hot side walls (pressure or suction side). Angling the ribs towards the fl ow results in vortex pairs and enables a bulk fl ow heat exchange with the close wall region (Han et al ., 1978). Today’s angled rib turbulators are a common technology used in many turbine parts, mainly on rotating components but also on static parts. From the leading edge towards the trailing edge of an airfoil the thick-• ness of the profi le signifi cantly reduces. For mechanical reasons it is often required to combine cooling features with geometrical features that are able to carry load. Pin-fi n turbulated cooling is a method ( Fig. 6.23 ) that offers heat transfer enhancement and can carry mechanical load. Half- pins can also be used to set the impingement to wall stand-off distance if jet impingement cooling is used, and therefore the feature is commonly applied in turbine cooling. Typical cooling schemes for front stage parts can be seen in Figs 6.24–6.28 and Plate XII (see colour section between pages 346 and 347). Figures 6.25 and 6.26 show the front stage vane and blade of an aero-engine (Rolls-Royce Trent 800). In Plate XII and Fig. 6.27 a front stage vane and blade cooling system of a heavy-duty gas turbine and in Fig. 6.28 the cooling system of a front stage heat shield of an unshrouded blade are shown. Hot gas at THTT , PHPP Film PMPP a x Jet orifice plate z Target plate Coolant at TOTT , POPP Staggered Jet orifice In-line a d/ dd 3 d/2dd d √33 Symbol Die 6.23 Internal cooling feature. Top left and middle left: rib-turbulated cooling, bottom left: pin-fi n cooling. Right: jet impingement cooling.
  • Turbines for industrial gas turbine systems 215 Published by Woodhead Publishing Limited, 2013 6.24 Aero-engine (Rolls-Royce) evolution of front stage rotating blade cooling schemes. Crossflow ribs LP ng airCooli HP 6.25 Aero-engine (Rolls-Royce) front stage rotating blade cooling scheme.
  • 216 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 Front stage vanes commonly use impingement cooling in combination with fi lm cooling. Thermal barrier coating is sprayed to the external surface. At the leading edge shower head fi lm cooling is used. The trailing edge is thin, and pin-fi n turbulators are implemented. Some engines use vane twins to reduce costs. Not all surfaces may be available for thermal barrier coating (TBC) or fi lm cooling hole drilling then. Two principle confi gurations are shown for the front stage blade ( Fig. 6.27 ): 1. A large multi-pass (1–3-3–1) type blade with split plane leading edge impingement and in some cases trailing edge impingement, which evolved 6.26 Aero-engine (Rolls-Royce) front stage vane cooling scheme. TE Cooling air LE 404 48440 20 42 4242 42 42 52 30 24 36 22 50 34 44 48 32 525225 24 6.27 Left: Siemens fi rst stage unshrouded blade, right: Alstom front stage blade.
  • Turbines for industrial gas turbine systems 217 Published by Woodhead Publishing Limited, 2013 from the US aero-engine technology. Such schemes are commonly used by General Electric (7FA) and Siemens (SGT5–4000, SGT6–5000). 2. A common multi-pass (1–3) type blade commonly used by Alstom (GT26), with advanced fi lm cooling on the leading edge and the blade tip, which evolved from the European aero-engine technology (Rolls- Royce). All front stage blades use fi lm cooling with thermal barrier coating. Some blades are made out of single crystal, others use directionally solidifi ed materials. Front stage heavy-duty gas turbine blades have no shrouds, due to the size and the loading of the blades. On multi-pass blades, core with- drawal of the ceramic core out of the core die tool imposes constraints on the cooling geometry feasible. Modern blades often use tip squealer arrangements with TBC coating to reduce the heat load and overtip leakage (Bunker, 2004). Highly loaded tips may require cold cooling air supplied directly to the tip and tip fl ag features can be used. A typical front stage heat shield can be seen in Fig. 6.28 . The cooling tech- nology used is very similar to the front stage vane. Film cooling is limited due to the disturbance of the passing blade. TBCs applied need to be rub resistant to allow aerodynamic clearance optimisation. In many cases high temperature abradable systems are used on the blade and heat shield to form a sealing system. 112 108 204 170 146 1321152 172 1741 171 173173 166 1671 162 211 210 187 1441 134 186 154 148 1751 178176 205 184 185 190 192 189 191 130 1881 137 193 194 140 156 136 150 168 1691 164 158 1141 138 1799177177 181 180 182 133 183 6.28 Right General Electric front stage heat shield (shroud) of an unshrouded blade.
  • 218 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 6.7 Future trends Future trends in turbine component design can only be estimated. Some engine parameters can be envisioned as follows. Gas turbine fl exibility in operation is an important driver. Turbine extended operation is required. Examples are grid frequency stabilisation and fast start-up capability to balance alternative energy generation, for example, due to fl uctuations in wind. Emission requirements for NO x , CO, CO 2 cap- ture achievability, for example, through fl u gas recirculation, fuel fl exibility for coal gasifi cation or hydrogen fuels, require fl exibility on the combustor interface in terms of temperature operation range and gas composition. Performance improvements in terms of power output and combined-cycle effi ciency will continue. The turbine hot gas temperature will increase beyond today’s level, and improved cooling schemes to further reduce cooling air consumption in spite of the hot gas temperature increase will be required. Published research, technologies used in aircraft engines, and patent applications may offer an insight into the next 20 year range. Often, relevant technology features are hidden and not published in patents if they are not directly accessible. From the many available examples, some advanced tech- nology features have been taken and will be discussed below. Examples of advanced technology features: Three-dimensional design features. • Loaded endwall designs. • Advanced convective cooling schemes for large components. • Advances in tip clearance reduction. • Advances in fi lm cooling. • Advances in material technology. • Advances in coating technology. • Advances in manufacturing technology. • 6.7.1 Aerodynamics A large amount of research has been carried out in the fi eld of non-axisym- metric endwall contouring (e.g. Rose, 1994; Rose et al ., 2001; Hartland and Smith, 2002; Paisner et al ., 2007). The fi rst applications were aimed at reduc- ing the transverse pressure amplitude at the trailing edge of a vane platform (Rose, 1994). Later it was recognised that there is a potential for reducing secondary fl ow losses by the application of non-axisymmetric endwall con- touring. This feature has been applied in commercial civil aircraft engines (Rose et al ., 2001), but not to power generation gas turbines. One would expect that eventually contoured endwall technology will be transferred from aircraft engines to power generation turbines ( Fig. 6.29 ).
  • Turbines for industrial gas turbine systems 219 Published by Woodhead Publishing Limited, 2013 The other main area of aerodynamic research that might in the future be applied to power gas turbines is the work on tip leakage reduction for unshrouded blades. Much of the work looks into the application of a so-called winglet or mini-shroud (Booth, 1981, 1985; Staubach et al ., 1996, Sjolander, 1997; Harvey and Ramsden, 2001; Harvey et al ., 2006). Such concepts have been applied in demonstrator engines and Harvey et al . (2006) report that the performance of their winglet design matched the one of a shrouded blade with two fi ns. This technology is still in the research stage even for aircraft engines ( Fig. 6.30 ). Further discussion can be found in Denton and Xu (1999), Moustapha et al . (2003) and Harvey (2004). 0.6 0.8 1.0 1.2 Non-dimensional stage work 1% stage efficiency 2a. 100% speed Profiled end wall turbine Datum turbine 6.29 Stage effi ciency improvement due to profi led end walls (Rose et al ., 2001). 1.009 (Datum bladed + Ref) 0.772 Winglet Winglet, Deep groove Deep groove, Ribs Double winglet 0.753 0.662 0.616 0.812 0.776 0.776 Groove (7% Tmax) Winglet, Deep groove, Ribs Deep groove (14% Tmax) Thin profile (57% Tmax) 6.30 Different shapes of blade tip winglets.
  • 220 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 Blade tip clearance control is to some extent in direct competition with advanced tip designs. In the fi rst stages of power generation turbines, typical relative tip clearances were 3% and more of the blade height. However, several manufacturers have recently introduced clearance control measures by either controlling the radial position of the casing or by axial movement of the rotor relative to the stator (Fukuizumi et al ., 2003). Other means of reducing the radial clearance is the application of an abradable coating to the stator heat shield (sometimes called shroud). If the steady state tip clearance can be reduced to suffi ciently low levels, the application of a winglet may no longer be attractive. 6.7.2 Cooling and manufacturing A signifi cant turbine part improvement relies on the advancement in manufacturing technology. Today the manufacturing processes, mostly cast- ing, limit possible advancements in cooling technology. Some insights on the pathways into the future can be drawn from the published patents and patent applications: More complex cores through: • Build cores out of several more detailed cores. – Dissolvable cores to manufacture the ceramic that are used in the – casting process. Build parts with: • Joining at the cold less-loaded side removing the constraints for the – hot fi llets that join the airfoil and endwalls. Shelves. – Hybrid parts with combinations of ceramic and standard super-alloy – materials to have, for example, leading edge inserts. Rapid prototype manufacturing technology for part manufacture. • Layered coatings which include transpiration cooling features. • Minimisation of cooling features and reduction in wall thickness. • TBC technology spraying to allow increased temperature deltas across • the ceramic layer moving towards a pure ceramic part. 6.8 References Anon, (2008), ‘32nd Power generation order survey’ , Diesel & Gas Turbine Worldwide , October, p. 32. Aoki S, (2000), ‘Trend and key technologies for gas turbine combined cycle power generation in a globally competitive market and environmental regulations’, Proceedings of the 2000 International Joint Power Generation Conference , July 23–26, Miami Beach, Florida.
  • Turbines for industrial gas turbine systems 221 Published by Woodhead Publishing Limited, 2013 Bals H, (2002), ‘Axialturbinen’, in Lechner C and Seume J, eds, Station ä re Gasturbinen , Berlin , Springer, 357–382. Bauer A and Rofka S, (2002), ‘Station ä res betriebsverhalten’, in Lechner C and Seume J, eds., Station ä re Gasturbinen , Berlin , Springer, 939–959. Booth T C, (1981), ‘Rotor-tip leakage part 1 – basic methodology’ , ASME Turbo Expo 1981 , March 8–12, Houston, Texas, 81-GT-71. Booth T C, (1985), ‘Importance of tip clearance fl ows in turbine design’ , VKI Lecture Series 1985–05: Tip Clearance Effects in Axial Turbomachines , Von Karman Institute for Fluid Dynamics , Belgium. Bunker R S, (2004), ‘Blade tip heat transfer and cooling techniques’ , VKI Lectures Series 2004–02: Turbine Blade Tip Design and Tip Clearance Treatment , Von Karman Institute for Fluid Dynamics , Belgium. Cohen H, Rogers G F C and Saravenemuttoo H I H, (1987), Gas Turbine Theory , 3rd edn, Harlow , Longman Scientifi c & Technical. Dailey G, (2000), ‘Design and calculation issues’ , VKI Lectures Series 2000–03: Aerothermal Performance of Internal Cooling Systems in Turbomachines , Von Karman Institute for Fluid Dynamics , Belgium. Denton J and Xu L, (1999), ‘The exploitation of three-dimensional fl ow in turbo- machinery design’ , in Denton J, ed., Developments in Turbomachinery Design , Professional engineering Publishing Ltd , Suffolk. Dowling N E, (1999), Mechanical Behaviour of Materials , Upper Saddle River, New Jersey, Prentice-Hall Inc. Dyschlevskij V, (1965), in Abianz V K, ed., Theory of Aviation Gas Turbines , Chapter VI. Losses in Turbine Flow Path, Moscow , Mashinostroenie. Eppler V, Lindvall K and Marx P, (2008), ‘GT26 with sequential combustion – combining operational fl exibility and clean, reliable power’ , VDI Fachtagung Station ä re Gasturbinen , 26 Nov, Leverkusen, Germany. Fukuizumi Y, Masada J, Kallianpur V and Iwasaki Y, ‘Application of “H gas turbine” design technology to increase thermal effi ciency and output capability of the Mitsubishi M701G2gas turbine’, ASME Turbo Expo 2003, June 16–19, Atlanta, Georgia, USA, GT2003-38956. Gailloreto S, (2009), ‘33rd Power generation order survey’ , Diesel & Gas Turbine Worldwide , October, p. 32. Goldstein R J, (1971), ‘Film cooling’ , in Irvine, T F and Hartnett, J P, eds, Advances in Heat Transfer , Issue 7, Elsevier. Gritsch M, Schulz A and Wittig S, (1997), ‘Adiabatic wall effectiveness measure- ments of fi lm cooling holes with expanded exits’ , ASME Turbo Expo 1997 , June 2–5, Orlando Florida, USA , 97-GT-164. Hartland J and Gregory-Smith D, (2002), ‘A design method for the profi ling of end walls in turbines’ , ASME Turbo Expo 2002 , June 3–6, Amsterdam, The Netherlands , 2002-GT-30433. Han J C, Glicksman L R and Rosenow W M, (1978), ‘An investigation of heat trans- fer and friction for rib roughened surfaces’ , Journal of Heat & Mass Transfer , 21 , 1143–1156. Han J-C, Dutta S and Ekkad S V, (2000), Gas Turbine Heat Transfer and Cooling Technology , London , Taylor & Francis. Harvey N W and Ramsden K, (2001), ‘A computational study of a novel turbine rotor partial shroud’ , ASME J of Turbomachinery , 123 , 534–543.
  • 222 Modern gas turbine systems Published by Woodhead Publishing Limited, 2013 Harvey N W, (2004), ‘Aerothermal implications of shroudless and shrouded blades’ , VKI Lectures Series 2004–02: Turbine Blade Tip Design and Tip Clearance Treatment , Von Karman Institute for Fluid Dynamics , Belgium. Harvey N W, Newman D A, Haselbach F and Willer L, (2006), ‘An investigation into a novel turbine rotor winglet. Part 1: Design and model rig test results’ , ASME Turbo Expo 2006 , May 8–11, Barcelona, Spain , 2006-GT-90456. Haselbach F, Schiffer H-P, Horsmann M and Dressen S, (2001), ‘The application of ultra high lift blading in the BR715 LP turbine’ , ASME Turbo Expo 2001 , June 4–7, New Orleans, Louisiana , 2001-GT-0436. Haight B, (2010), ‘2010 Power generation order survey’ , Diesel & Gas Turbine Worldwide , May, p. 26. Horlock J H, (1966), Axial Flow Turbines , London , Butterworths , 127. Japikse D and Baines N C, (1997), Introduction to Turbomachinery , Oxford , Oxford University Press . Kr ü ckels J, Arzel T, Kingston T R and Schnieder M, (2007), ‘Turbine blade thermal design process enhancements for increased fi ring temperatures and reduced coolant fl ow’ , ASME Turbo Expo 2007 , May 14–17, Montreal, Canada , 2007-GT- 27457. Lakshminarayana B, (1996), Fluid Dynamics and Heat Transfer of Turbomachinery , New York , John Wiley & Sons. McNeely M, (2007), ‘31st Power generation order survey’ , Diesel & Gas Turbine Worldwide , October, p. 30. Meyer A, (1939), ‘The combustion gas turbine: Its history, development and pros- pects’, The Brown Boveri Review , 26 (6), 127–140. Mich à lek J, Monaldi M and Arts T, (2010), ‘Aerodynamic performance of a very high lift low pressure turbine airfoil (T106C) at low Reynolds and high Mach num- ber with effect of free stream turbulence intensity’ , ASME Turbo Expo 2010 , June 14–18, Glasgow, UK , 2010-GT-22884. Miller R A, (1997), ‘Thermal barrier coatings for aircraft engines: History and direc- tions , Journal of Thermal Spray Technology , 6 (1), 36–42. Moustapha S H, Okapuu U and Williamson R G, (1987), ‘Infl uence of rotor blade aerodynamic loading on the performance of a highly loaded turbine stage’ , Transactions ASME Journal of Turbomachinery , 109 , 155–162. Moustapha H, Zelesky M F, Baines N C and Japikse D, (2003), Axial and Radial Turbines , Concepts NREC , White River Junction. Patel K V, (1980), ‘Research on a high work axial gas generator turbine’ , SAE 800618. Praisner T J, Allen-Bradley E, Grover E A, Knezevici D C and Sjolander S A, (2007), ‘Application of non-axisymmetric endwall contouring to conventional and high-lift turbine airfoils’ , ASME Turbo Expo 2007 , May 14–17, Montreal, Canada , 2007-GT-27579. Rose M G, (1994), ‘Non-axisymmetric endwall profi ling in the HP NGV’s of an axial fl ow gas turbine’ , ASME Turbo Expo 1994 , June 13–16, The Hague, The Netherlands , 94-GT-249. Rose M G, Harvey N W, Seaman P, Newman D A and McManus D, (2001), ‘Improving the effi ciency of the Trent 500 HP turbine using non-axisymmetric end-walls: Part II – experimental validation’ , ASME Turbo Expo 2001 , June 4–7, New Orleans, Louisiana , 2001-GT-0505.
  • Turbines for industrial gas turbine systems 223 Published by Woodhead Publishing Limited, 2013 Seippel C, (1966), ‘The evolution of compressor and turbine bladings in gas turbine design’ , ASME Turbo Expo 1996 , March 13–17, Z ü rich, Switzerland , 66-GT-106. Seippel C and Bereuter R, (1960), ‘The theory of combined steam and gas turbine installations’ , The Brown Boveri Review , 47 , 783–799. Sjolander S A, (1997), ‘Overview of tip-clearance effects in axial turbines ’, VKI Lecture Series 1997–01: Secondary and Tip Clearance Flows in Axial Turbines , Von Karman Institute for Fluid Dynamics , Belgium. Smith S F, (1963), ‘A simple correlation of turbine effi ciency’ , Journal of the Royal Aeronautical Society , 69 , 467–470. Somerscales E F C and Hendrickson R L, (1984), ‘3500 kW gas turbine at the Schenectady plant of the General Electric Company’ , ASME National Historic Landmark Dedication Brochure. http://www.asme.org/about-asme/history/ landmarks Staubach J B, Sharma O P and Stetson G M, (1996), ‘Reduction of tip clearance losses through 3-D airfoil designs’ , ASME Turbo Asia Conference: November 5–7, Jakarta, Indonesia, Paper No. 96-TA-13. Sunden , B and Faghri, M, eds, (2001), Heat Transfer in Gas Turbines , Southhampton , WIT Press. Van der Linden S, (1988), ‘The World’s First Industrial Gas Turbine Set at Neuch â tel (1939)’ , ASME International Historic Landmark Commemorative Brochure. http://www.asme.org/about-asme/history/landmarks Webster G A and Ainsworth R A, (1994), High Temperature Component Life Assessment , London , Chapman & Hall. Wilson D G and Korakianitis T, (1998), The Design of High-effi ciency Turbomachinery and Gas Turbines , 2nd edn, Upper saddle River, New Jersey, USA, Prentice- Hall Inc . Zweifel O, (1945), ‘The axial spacing of turbomachine blading, especially with large angular defl ection’ , Brown Bovery Review , 32 (12), 436–444.
  • © Woodhead Publishing Limited, 2013 Plate XII (Chapter 6) GT26 LPT front stage vane.