0

Heat Transfer

1990;():V004T09A001. doi:10.1115/90-GT-008.
FREE TO VIEW

Results are presented of an experimental investigation into the influence of mainstream acceleration on the heat transfer coefficient downstream of injection through a row of 35° holes in a flat plate. A mass transfer analogue technique was used, with two uniform acceleration parameters, K (=ν(du/dx)/u2), of 1.9 × 10−6 and 5.0 × 10−6 in addition to the zero acceleration base-line case.

Two injectants, air and carbon dioxide, were employed to give coolant to mainstream density ratios of 1.0 and 1.52 respectively. The blowing rate varied from 0.5 to 2.0.

The heat transfer coefficient beneath the film reduced progressively as the acceleration increased, with maximum reductions from the zero acceleration datum case of about 27%. In the presence of acceleration, the heat transfer coefficient at a given blowing rate was dependent on the density ratio, an increase in the density ratio leading to a decrease in the heat transfer coefficient. An empirical correlation of the data over most of the range of densities and blowing rates of the experiments has been developed.

Commentary by Dr. Valentin Fuster
1990;():V004T09A002. doi:10.1115/90-GT-009.
FREE TO VIEW

The influence of high mainstream turbulence on leading edge film effectiveness and heat transfer coefficient was studied. High mainstream turbulence was produced by a passive grid and a jet grid. Experiments were performed using a blunt body with a semi-cylinder leading edge with a flat afterbody. The mainstream Reynolds number based on leading edge diameter was about 100,000. Spanwise and streamwise distributions of film effectiveness and heat transfer in the leading edge and on the flat sidewall were obtained for three blowing ratios, through rows of holes located at ±15° and ±40° from stagnation. The holes in each row were spaced three hole-diameters apart and were angled 30° and 90° to the surface in the spanwise and streamwise directions respectively. The results indicate that the film effectiveness decreases with increasing blowing ratio, but the reverse is true for the heat transfer coefficient. The leading edge film effectiveness for low blowing ratio (B = 0.4) is significantly reduced by high mainstream turbulence (Tu = 9.67% and 12.9%). The mainstream turbulence effect is diminished in the leading edge for higher blowing ratios (B = 0.8 and 1.2) but still exists on the flat sidewall region. Also, the leading edge heat transfer coefficient for blowing ratio of 0.8 increases with increasing mainstream turbulence; but the effect for other blowing ratios (B = 0.4 and 1.2) is not so systematic as for B = 0.8. Surface heat load is significantly reduced with leading edge film cooling.

Commentary by Dr. Valentin Fuster
1990;():V004T09A003. doi:10.1115/90-GT-010.
FREE TO VIEW

The effects of injection rate and strength of curvature on film cooling performance of gas injected through a row of holes on a convex surface is studied. Comparisons are made to film cooling of concave and flat surfaces. Three different relative strengths of curvature (ratio of radius of curvature to radius of injection hole), two density ratios (0.95 and 2.0), and a wide range of blowing rates (0.3 to 2.7) are considered. A foreign gas injection technique (mass transfer analogy) is used. The strength of curvature was controlled by varying the injection hole diameter. At low blowing rates, film cooling is more effective on the convex surface than on a flat or a concave surface. The cross stream pressure gradient present in curved flows tends to push the jet into the convex wall. As the injection rate is increased, normal and tangential jet momentum promote lift-off from the convex surface, thereby lowering performance. In contrast, previous studies show that on a concave surface, tangential jet momentum, flow instabilities, and blockage improve performance on a concave surface as blowing rate is increased.

Topics: Film cooling
Commentary by Dr. Valentin Fuster
1990;():V004T09A005. doi:10.1115/90-GT-024.
FREE TO VIEW

This work is concerned with fully-developed constant-density turbulent flow through rectangular straight ducts rotating in an orthogonal mode. Ducts of both square and 2:1 aspect ratio cross-sections have been examined. For the square duct, predictions have been performed for Reynolds numbers of 33,500 and 97,000 and for the 2:1 aspect ratio duct the computations were carried out for a Reynolds number of 33,500. Values of the inverse Rossby number (Ro = ΩD/Wb) ranged from 0.005 to 0.2. Except in the immediate vicinity of the wall, the standard high-Reynolds-number version of the k-ε model is used to account for the effects of turbulence. Across the near-wall sublayer the damping of turbulence is modelled through a low-Reynolds-number one-equation model.

Low rotational speeds cause the formation of a pair of symmetric streamwise vortices. At higher rotational speeds, flow instabilities on the pressure side lead to transition to a more complex four-vortex structure. The transition point depends on both the cross-sectional geometry and the flow Reynolds number. Moreover, over a range of Rossby number, either two- or four-vortex solutions are possible depending upon initial conditions.

The rotation leads to significant differences between the values of friction factor and Nusselt number on the suction and pressure surfaces of the duct. The degree of heat transfer augmentation on the pressure side is found to depend on the Reynolds number as well as on Rossby number. In contrast, heat-transfer attenuation on the suction side is only Rossby-number dependent.

Commentary by Dr. Valentin Fuster
1990;():V004T09A006. doi:10.1115/90-GT-025.
FREE TO VIEW

Results are presented from an experimental study designed to obtain detailed radial heat transfer coefficient distributions applicable to the cooling of disk-cavity regions of gas turbines. An experimental apparatus has been designed to obtain local heat transfer data on both the rotating and stationary surfaces of a parallel geometry disk-cavity system. The method employed utilizes thin thermochromic liquid crystal coatings together with video system data acquisition and computer-assisted image analysis to extract heat transfer information. The color display of the liquid crystal coatings is detected through the analysis of standard video chromanance signals. The experimental technique used is an aerodynamically steady but thermally transient one which provides consistent disk-cavity thermal boundary conditions while yet being inexpensive and highly versatile. A single circular jet is used to introduce fluid from the stator into the disk-cavity by impingement normal to the rotor surface. The present study investigates hub injection of coolant over a wide range of parameters including disk rotational Reynolds numbers of 2 to 5 · 105, rotor/stator spacing-to-disk radius ratios of .025 to .15, and jet mass flow rates between .10 and .40 times the turbulent pumped flow rate of a free disk. The results are presented as radial distributions of local Nusselt numbers. Rotor heat transfer exhibits regions of impingement and rotational domination with a transition region between, while stator heat transfer shows flow reattachment and convection regions with evidence of an inner recirculation zone. The local effects of rotation, spacing, and mass flow rate are all displayed. The significant magnitude of stator heat transfer in many cases indicates the importance of proper stator modeling to rotor and disk-cavity heat transfer results.

Commentary by Dr. Valentin Fuster
1990;():V004T09A007. doi:10.1115/90-GT-026.
FREE TO VIEW

Detailed radial distributions of rotor heat transfer coefficients are presented for three basic disk-cavity geometries applicable to gas turbines. The experimental apparatus has been designed to obtain local heat transfer data on a number of easily interchangeable rotor surfaces. The method employs thin thermochromic liquid crystal coatings upon the rotor surfaces together with video system data acquisition and computer-assisted image analysis to detect surface color display and to extract heat transfer information. A thermally transient, aerodynamically steady technique is used which attains consistent thermal boundary conditions over the entire disk-cavity. Cooling air is introduced into the disk-cavity via a single circular jet mounted perpendicularly into the stator at one of three radial locations; 0.4, 0.6 or 0.8 times the rotor radius. Rotor heat transfer coefficients have been obtained over a range of parameters including disk rotational Reynolds numbers of 2 to 5 · 105, rotor/stator hub spacing-to-disk radius ratios of .025 to .15, and jet mass flow rates between .10 and .40 times the turbulent pumped flow rate of a free disk. The rotor surfaces include a parallel rotor-stator system, a rotor with 5 percent diverging taper, and a similarly tapered rotor with a rim sealing lip at its extreme radius. Results are presented showing the effects of the parallel rotor, which indicate strong variations in local Nusselt numbers for all but rotational speed. These results are compared to associated hub injection data of Part I of this study, demonstrating that overall rotor heat transfer is optimized by either hub injection or radial location injection of coolant dependent upon the configuration. Results with the use of the tapered rotor show significant local Nusselt number radial variation changes over those of the parallel rotor, while the addition of a rim sealing lip appears to increase the level of the radial distribution.

Commentary by Dr. Valentin Fuster
1990;():V004T09A008. doi:10.1115/90-GT-041.
FREE TO VIEW

Detailed heat transfer and aerodynamic measurements have been made on an annular cascade of highly loaded nozzle guide vanes. The tests were carried out in an Isentropic Light Piston test facility at engine representative Reynolds number, Mach number and gas-to-wall temperature ratio.

The aerodynamics indicate that the vane has a weak shock at 65–70% axial chord (mid span) with a peak Mach number of 1.14. The influence of Reynolds number and Mach number on the Nusselt number distributions on the vane and endwall surfaces are shown to be significant. Computational techniques are used for the interpretation of test data.

Commentary by Dr. Valentin Fuster
1990;():V004T09A009. doi:10.1115/90-GT-042.
FREE TO VIEW

Comparisons with experimental heat transfer and surface pressures were made for seven turbine vane and blade geometries using a quasi-three-dimensional thin-layer Navier-Stokes analysis. Comparisons are made for cases with both separated and unseparated flow over a range of Reynolds numbers and freestream turbulence intensities. The analysis used a modified Baldwin-Lomax turbulent eddy viscosity model. Modifications were made to account for the effects of: 1) freestream turbulence on both transition and leading edge heat transfer; 2) strong favorable pressure gradients on re-laminarization; and 3) variable turbulent Prandtl number on heat transfer. In addition, the effect on heat transfer of the near-wall model of Deissler is compared with the Van Driest model.

Commentary by Dr. Valentin Fuster
1990;():V004T09A011. doi:10.1115/90-GT-044.
FREE TO VIEW

Flowfield measurements with relevance to gas turbine film cooling were carried out behind a second row of holes located 40 hole diameters downstream of a first row and staggered with respect to it. The effects of the thicker boundary layer and that of injection from the first row were examined. Experiments were carried out for jet-to-mainstream density ratios of 1 and 2 and results compared to those obtained by Pietrzyk et al. (1989a,b) behind the first row. Results show that the dominant structures identified in the flowfield downstream of the first row were also present behind the second row. Due to the thicker boundary layer, though, the mean flowfield and the turbulence and shear stress fields were altered. One of the significant results was that the thicker boundary layer enables the jets to penetrate deeper into the mainstream.

Commentary by Dr. Valentin Fuster
1990;():V004T09A012. doi:10.1115/90-GT-045.
FREE TO VIEW

Results are presented which illustrate the effects of single embedded longitudinal vortices on heat transfer and injectant downstream of a single film-cooling hole in a turbulent boundary layer. Attention is focussed on the changes resulting as circulation magnitudes of the vortices are varied from 0.0 to 0.15 m**2/s. Mean temperature results are presented which show how injectant is distorted and redistributed by vortices, along with heat transfer measurements and mean velocity surveys. Injection hole diameter is 0.952 cm to give a ratio of vortex core diameter to hole diameter of about 1.5–1.6. The freestream velocity is maintained at 10 m/s, and the blowing ratio is approximately 0.5. The film-cooling hole is oriented 30 degrees with respect to the test surface. Stanton numbers are measured on a constant heat flux surface with a non-dimensional temperature parameter of about 1.5. Two different situations are studied: one where the injection hole is beneath the vortex downwash, and one where the injection hole is beneath the vortex upwash. For both cases, vortex centers pass well within 2.9 vortex core diameters of the centerline of the injection hole.

To quantify the influences of the vortices on the injectant and local heat transfer, the parameter S is used, defined as the ratio of vortex circulation to injection hole diameter times mean injection velocity. When S is greater than 1.0–1.5, injectant is swept into the vortex upwash and above the vortex core by secondary flows, and Stanton number data show evidence of injectant beneath the vortex core and downwash near the wall for x/d only up to 33.6. For larger x/d, local Stanton numbers are augmented by the vortices by as much as 23 percent relative to film-cooled boundary layers with no vortices. When S is less than 1.0–1.5, some injectant remains near the wall beneath the vortex core and downwash where it continues to provide some thermal protection. In some cases, the protection provided by film cooling is augmented because of vortex secondary flows which cause extra injectant to accumulate near vortex upwash regions.

Commentary by Dr. Valentin Fuster
1990;():V004T09A013. doi:10.1115/90-GT-046.
FREE TO VIEW

This study investigates the influence of incidence on convective heat transfer to highly curved surfaces of a film cooled turbine rotor blade. A computational study of free stream inviscid aerodynamics without cooling at various incidences is followed by well documented measured heat transfer data sets. The heat transfer experiments are discussed for cases with and without film cooling, performed under realistic gas turbine flow conditions in the short duration heat transfer facility of the von Karman Institute for Fluid Dynamics. The precise location of the stagnation point and the iso-Mach number contours in the passage for each incidence (−10°, 0°, 10°, +10°) are presented for a nominal exit Mach number of 0.94. The free stream mass flow rate was kept constant for each experiment at different incidence levels. Three rows of compound angled discrete cooling holes are located near the leading edge in a shower-head configuration. Two rows of staggered discrete cooling holes are located on the suction side and a single row of cooling holes is located on the pressure side. The short duration measurements of quantitative wall heat fluxes on nearly isothermal blade surfaces both in the presence and absence of coolant ejection are presented. The study indicated that the change of the position of the stagnation point strongly altered the aerodynamic behaviour and convective heat transfer to the blade in approximately the first 30 % of both the pressure side and the suction side in the presence and absence of film cooling. The immediate vicinity of the stagnation point was not significantly affected by changing incidence without cooling. Transitional behaviour both on the suction surface and on the pressure surface was significantly influenced by the changes in approching flow direction. Flow separation associated with incidence variations was also observed. Extremely low levels of convective heat transfer coefficients were experienced near the regions where small separation bubbles are located.

Commentary by Dr. Valentin Fuster
1990;():V004T09A014. doi:10.1115/90-GT-053.
FREE TO VIEW

Experimental data on a rough surface for both transitionally rough and fully rough turbulent flow regimes are presented for Stanton number distribution, skin friction coefficient distribution and turbulence intensity profiles. The rough surface is composed of 1.27 mm diameter hemispheres spaced in a staggered array four base diameters apart on an otherwise smooth wall. Special emphasis is placed on the characteristics of heat transfer in the transitionally rough flows. Stanton number data are reported for zero pressure gradient incompressible turbulent boundary layer air flow for nominal freestream velocities of 6, 12, 28, 43, 58 and 67 m/s, which give x-Reynolds numbers up to 10,000,000. These data are compared with previously published rough surface data, and the classification of a boundary layer flow into transitionally rough and fully rough regimes is explored. Moreover, a new heat transfer model for use in the previously published discrete element prediction approach is presented. Computations using the discrete element model are presented and compared with data obtained from two different rough surfaces. The discrete element predictions for both surfaces are found to be in substantial agreement with the data.

Commentary by Dr. Valentin Fuster
1990;():V004T09A015. doi:10.1115/90-GT-061.
FREE TO VIEW

Experimental measurements of the overall cooling effectiveness for full coverage discrete hole effusion cooling are presented for a wide range of practical geometries and for a density ratio between the coolant and combustion gases of 2.5. The influence of the number of holes per unit surface area was investigated at two fixed total hole areas or design pressure losses of 3% and 0.1%, at a relatively low coolant flow rate per unit surface area. Hole configurations suitable for both combustor and turbine blade cooling were investigated with hole sizes from 1.4 to 0.6mm at 3% design pressure loss and 1.3 to 3.3mm at 0.1% design pressure loss. The diameter change at a fixed pressure loss was for a constant total hole area with more holes as the size was reduced. This was shown to increase the cooling effectiveness through improved film cooling. Enlarging the hole size for a fixed number of holes and hence reducing the pressure loss for a fixed coolant mass flow was also shown to improve the cooling effectiveness through better film cooling. Major reductions in current combustor wall cooling flows were demonstrated for some full coverage effusion geometries.

Commentary by Dr. Valentin Fuster
1990;():V004T09A016. doi:10.1115/90-GT-062.
FREE TO VIEW

The increased severity of the thermal environment of high pressure turbine blades and vanes requires accurate calculations for the successful design of these parts. In this paper, the prediction of the temperature field in the near-cooling-hole region on a film cooled turbine vane is presented. The surface distribution of the heat transfer coefficient and the film cooling effectiveness on the vane in presence of one or several film cooling injections is obtained from boundary layer calculations and via experimental correlations. Cooling jet coalescence is taken into account as well as the main parameters governing this physical phenomenon. The internal boundary conditions result from available correlations. The study was conducted on two different configurations : a flat plate including an injection through two rows of holes and a turbine vane including three injections through two rows of holes on the suction side. Thermal computations using a three-dimensional finite element code yield strong temperature distortions and high temperature gradients around the injection zones. The study also indicates that the three-dimensional temperature field just downstream of the injections becomes two-dimensional when jet coalescence takes place. The influence of one or several obstructed injection holes on the temperature field is studied; important effects are observed when the main flow temperature is high.

Topics: Temperature , Turbines
Commentary by Dr. Valentin Fuster
1990;():V004T09A017. doi:10.1115/90-GT-078.
FREE TO VIEW

Detailed heat transfer measurements have been made around a film-cooled transonic gas turbine rotor blade in the Oxford Isentropic Light Piston Tunnel. Film cooling behaviour for four film cooling configurations has been analysed for a range of blowing rates both without and with simulated nozzle guide vane (NGV) shock wave and wake passing. The superposition model of film cooling has been employed in analysis of time-mean heat transfer data, while time resolved unsteady heat transfer measurements have been analysed to determine interaction between film-cooling and unsteady shock wave and wake passing. It is found that there is a significant change of film-cooling behaviour on the suction surface when simulated NGV unsteady effects are introduced.

Commentary by Dr. Valentin Fuster
1990;():V004T09A018. doi:10.1115/90-GT-092.
FREE TO VIEW

The response of a boundary layer in the stagnation region of a two-dimensional body to fluctuations in the freestream is examined. The analysis is restricted to laminar incompressible flow. The assumed form of the velocity distribution at the edge of the boundary layer represents both a pulsation of the incoming flow, and an oscillation of the stagnation point streamline. Both features are essential in accurately representing the effect which freestream spatial and temporal nonuniformities have upon the unsteady boundary layer. Finally, a simple model is proposed which relates the characteristic parameters in a turbulent wake to the unsteady boundary-layer edge velocity. Numerical results are presented for both an arbitrary two-dimensional geometry and a circular cylinder.

Commentary by Dr. Valentin Fuster
1990;():V004T09A019. doi:10.1115/90-GT-094.
FREE TO VIEW

The mean velocity field in the boundary layer and the streamwise and spanwise distributions of heat transfer coefficient have been measured on a concave wall in the presence of naturally-generated Görtler vortices, with and without a preceding flat wall. In near-zero pressure gradient, enhancement of the streamwise-averaged heat transfer above flat surface levels was associated with the attainment of a Görtler number of around ten, as found in previous experiments in a different flow facility with higher wall curvature, but occurred before the onset of severe distortion in the velocity profiles. Velocity gradient parameters K of 0·20 × 10−6 and 0·75 × 10−6 resulted in a more regular vortex structure, with spanwise averaged heat transfer reaching two to three times predicted levels. At K = 1·8 × 10−6, vortex amplification was suppressed to such an extent that no significant heat transfer enhancement took place. Comparison of measured Stanton numbers with those derived from skin friction factors (obtained from velocity profiles) suggested that the heat transfer enhancement is not simply a result of fuller velocity profiles in vortex downwash regions.

Commentary by Dr. Valentin Fuster
1990;():V004T09A020. doi:10.1115/90-GT-095.
FREE TO VIEW

After compressor discharge air has initially been used to cool the heat shields of the hot gas inlet casing, it can subsequently be employed for film cooling of the first-stage vane shrouds. Since the flow field near these shrouds is three-dimensional, the film cooling effectiveness cannot be predicted correctly by common two-dimensional codes. The secondary flow transports the film from the pressure side to the suction side where it can even climb up the airfoil to cool its trailing section.

Such film cooling effectiveness was first investigated experimentally in a linear vane cascade at atmospheric pressure. The temperatures and static pressure levels at the adiabatic shrouds, as well as the temperature measurements within the vane cascade, are reported for different cooling film blowing rates.

In addition, the secondary flow was analysed numerically using a partially-parabolic computer code for 3D viscous flows. It involves mutual interaction of the boundary layer with the mainstream. The secondary flow can also be modelled with this algorithm, which requires less numerical effort than solving the fully 3D elliptic flow equations. The numerical results of the experiment and numerical predictions are compared. In addition, the application of these results to a high-temperature gas turbine is presented.

Commentary by Dr. Valentin Fuster
1990;():V004T09A021. doi:10.1115/90-GT-100.
FREE TO VIEW

A decade ago several important fundamental heat transfer phenomena were identified which were considered basic to the ability to predict heat transfer loads in aircraft gas turbines. The progress in addressing these fundamentals over the past ten years is assessed in this paper. Much research effort has been devoted to their study in university, industry and government laboratories and significant progress has been achieved. Advances in computer technology have enabled the modeling of complex three-D fluid flow in gas turbines so necessary for heat transfer calculations. Advances in instrumentation plus improved data acquisition have brought about more reliable data sets. While much has advanced in the ‘80’s, much challenging research remains to be done. Several of these areas are suggested in the paper.

Commentary by Dr. Valentin Fuster
1990;():V004T09A022. doi:10.1115/90-GT-120.
FREE TO VIEW

An investigation was conducted to study the unsteady effects of a moving wake as it passes over a laminar boundary layer in a stagnation region. Arguments are presented which show that in this region, the wake-induced unsteadiness may be treated, for the most part, as an inviscid, unsteady freestream which impresses itself on the boundary layer flow. As a result, the boundary layer equations remain valid and, for relatively small oscillations, a solution to the equations may be obtained using standard perturbation techniques. A related experiment is then described and the results are examined in light of this analytical approach.

Commentary by Dr. Valentin Fuster
1990;():V004T09A023. doi:10.1115/90-GT-121.
FREE TO VIEW

The flow phenomena of wakes shed by upstream blade rows is a well–known problem in turbomachinery which influences blade forces, vibrations, losses and heat transfer. With respect to the heat load to turbine blades this problem becomes even more complex because of the interaction between wake, potential flow and the boundary layer along the surface of the airfoil.

Experimentally evaluated mean heat-transfer coefficients obtained under different unsteady initial conditions are reported. The heat–transfer measurements have been carried out in the cascade test facility at the ITS in Karlsruhe using a rotating bar wake generator placed upstream of the cascade to simulate the wake passing process. The variation of the wake parameters includes different wake passing frequencies, cascade inlet Reynolds numbers and wake inclination angles. In addition, the relevant parameters of the unsteady wake have been measured by means of a fixed hot–wire anemometer using the ensemble average technique. The results are compared to those from the literature for the wake of a cylinder in cross–flow. They also serve as experimental base for parallel theoretical analyses.

Commentary by Dr. Valentin Fuster
1990;():V004T09A024. doi:10.1115/90-GT-137.
FREE TO VIEW

A theory for transition from laminar to turbulent flow with an unsteady, periodic passing of turbulent wakes in the free stream has recently been presented by the authors. The theory considers a time-averaged transitional flow caused by the formation and propagation of turbulent strips along the surface. To apply the theory, however, both the origin and a quantity related to the production rate of these turbulent strips must be known. In this paper, after a brief review of the theory, a dimensional analysis of the problem is presented and data from experiments re-examined in light of the result. From this, an expression for the time-averaged intermittency is obtained which may be used to calculate the time-averaged distributions of various boundary layer quantities for wake-induced transitional flow.

Topics: Turbulence , Wakes , Strips
Commentary by Dr. Valentin Fuster
1990;():V004T09A025. doi:10.1115/90-GT-138.
FREE TO VIEW

The influence of rotation on local heat transfer in a rectangular-sectioned duct has been experimentally studied for the case where the ductrotates about an axis orthogonal to its own central axis. The coolant used was air with the flow direction in the radially outwards direction. This rotating flow geometry is encountered in the internal cooling of gas turbine rotor blades.

Commentary by Dr. Valentin Fuster
1990;():V004T09A026. doi:10.1115/90-GT-174.
FREE TO VIEW

A numerical simulation of temperature distortion at inlet to a rotating turbine rotor has been performed. A 3D Navier-Stokes code the 2D boundary layer code - STAN5, have been used.

The results show that the hot gas is transported to the pressure surface of the blade and that hot gas also migrates to the blade pressure side tip. At locations greater than 50–60% axial chord the hot gas enters the tip-gap and emerges over the suction side. These computations are in agreement with previous experimental results.

The secondary flows within the turbine rotor are enhanced by the introduction of inlet radial temperature distortion and this is in accord with previous analytical work. It is shown that the heat flux near the tip region on the pressure side of the blade can be increased by up to 76% due to the re-distribution of the inlet temperature distortion.

Commentary by Dr. Valentin Fuster
1990;():V004T09A027. doi:10.1115/90-GT-175.
FREE TO VIEW

Heat transfer measurements made on the rotor blade of a full stage model turbine operating at engine representative conditions are presented. The measurement technique of mounting thin film heat transfer gauges on enamel coated turbine blades enables the heat transfer rate to be measured across a frequency range of d.c. to 100kHz. The output is amplified using electronic circuits housed inside the shaft before transmission through a slipring and digital signal processing routines are used to calculate the heat transfer rate. A calibration experiment in which the gauge is pulsed with a laser beam is described in detail.

The results are compared with data from a previous two-dimensional simulation of wake-passing flow on the midheight section of the same blade. The mean heat transfer rate recorded in the two experiments shows reasonable agreement. Fluctuations in the unsteady heat transfer signal at NGV passing frequency are seen at the same locations in data from both experiments, however the magnitude of the fluctuations seen on the rotor are much smaller than those recorded in the two-dimensional simulation. Frequency spectra and correlation analysis of heat transfer traces recorded on the rotor are also presented.

Commentary by Dr. Valentin Fuster
1990;():V004T09A028. doi:10.1115/90-GT-219.
FREE TO VIEW

The local heat transfer from a plane rotating disc enclosed in a casing has been studied experimentally. The disc of 800 mm diameter can be run up to 2000 min−1 at axial distances between disc and casing varied between 5 and 55 mm. Centrifugal or alternatively centripetal flow of cooling air at rates up to ṁ = 1 kg/s can be applied, both with or without an inlet swirl.

With the disc rotating in a closed casing (ṁ = 0 kg/s) the influence of the characteristic dimensionless groups on the local heat transfer has been investigated. At a fixed radius, a variation of the local Reynolds Number by either speed or density results in corresponding changes of the heat transfer. However, with a variation of the radius different heat transfer-Re relations are found. In fact, the temperature distribution in the gas caused by the heat flux results in an additional influence of free convection, to be expressed by a Grashof Number. This is confirmed by a comparison of the experimental results with calculations based on Reynolds Analogy and measured friction coefficients. The discrepancies found can be explained only, if in addition to the limitations of the analogy, the influence of free convection is taken into account. Additional results of ongoing experiments concerning the influence of the geometry of the cavity between disc and casing, of the coolant flow rate and of the swirl are presented.

Commentary by Dr. Valentin Fuster
1990;():V004T09A029. doi:10.1115/90-GT-254.
FREE TO VIEW

Experiments have been conducted to study the turbulent heat transfer and friction for fully developed flow of air in a square channel in which two opposite walls are roughened with 90° full ribs, parallel and crossed full ribs with angles-of-attack (α) of 60° and 45°, 90° discrete ribs, and parallel and crossed discrete ribs with = 60°, 45°, and 30°. The discrete ribs are staggered in alternate rows of three and two ribs. Results are obtained for a rib height-to-channel hydraulic diameter ratio of 0.0625, a rib pitch-to-height ratio of 10, and Reynolds numbers between 10,000 and 80,000. Parallel angled discrete ribs are superior to 90° discrete ribs and parallel angled full ribs, and are recommended for internal cooling passages in gas turbine airfoils. For α = 60° and 45°, parallel discrete ribs have higher ribbed wall heat transfer, lower smooth wall heat transfer, and lower channel pressure drop than parallel full ribs. Parallel 60° discrete ribs have the highest ribbed wall heat transfer and parallel 30° discrete ribs cause the lowest pressure drop. The heat transfer and pressure drops in crossed angled full and discrete rib cases are all lower than those in the corresponding 90° and parallel angled rib cases. Crossed arrays of angled ribs have poor thermal performance and are not recommended.

Commentary by Dr. Valentin Fuster
1990;():V004T09A030. doi:10.1115/90-GT-255.
FREE TO VIEW

An experimental investigation of the effects of nozzle operating conditions on the development of nozzle-exit boundary layers of highly heated air free jets is reported in this paper. The total pressure measurements in the nozzle-exit boundary layer were obtained at a range of jet Mach numbers from 0.1 to 0.97 and jet total temperatures up to 900 K. The analysis of results shows that the nozzle-exit laminar boundary-layer development depends only on the nozzle-exit Reynolds number. For the nozzle-exit turbulent boundary layer, however, it appears that the effects of the jet total temperature on the boundary-layer integral characteristics are independent from the effect of the nozzle-exit Reynolds number. This surprizing finding has not yet been reported. Further, laminar boundary-layer profiles were compared with the Pohlhausen solution for a flat-wall converging channel and an acceptable agreement was found only for low Reynolds numbers. For turbulent boundary layers, the dependence of the shape factor on relative Mach numbers at a distance of one momentum thickness from the nozzle wall resembles Spence’s prediction. Finally, the calculated total pressure loss coefficient was found to depend on the nozzle-exit Reynolds number for the laminar nozzle-exit boundary layer, while for the turbulent exit boundary layer this coefficient appears to be constant.

Commentary by Dr. Valentin Fuster
1990;():V004T09A031. doi:10.1115/90-GT-256.
FREE TO VIEW

The performance characteristics of a heat flux microsensor have been measured and analyzed. This is a new heat flux gage system that is made using microfabrication techniques. The gages are small, have high frequency response, can measure very high heat flux, and output a voltage directly proportional to the heat flux. Each gage consists of a thin thermal resistance layer sandwiched between many thermocouple pairs forming a differential thermopile. Because the gage is made directly on the measurement surface and the total thickness is less than 2µm, the presence of the gage contributes negligible flow and thermal disruption. The active surface area of the gage is 3 mm by 4 mm, with the leads attached outside this area to relay the surface heat flux and temperature signals.

Gages were made and tested on glass and silicon substrates. The steady and unsteady response was measured experimentally and compared with analytical predictions. The analysis was performed using a one-dimensional, transient, finite-difference model of the six layers comprising the gage plus the substrate. Steady-state calibrations were done on a convection heat transfer apparatus and the transient response was measured to step changes of the imposed radiative flux. As an example of the potential capabilities, the time-resolved heat flux was measured at a stagnation point with imposed freestream turbulence. A hot-film probe placed outside the boundary layer was used to provide a simultaneous signal showing the corresponding turbulent velocity fluctuations.

Commentary by Dr. Valentin Fuster
1990;():V004T09A032. doi:10.1115/90-GT-285.
FREE TO VIEW

Transient cooling techniques were developed for the direct simultaneous measurement of full coverage multi-jet impingement heat transfer coefficient and cooling effectiveness under high temperature combustion convective wall heating conditions. A step change in the impingement coolant flow rate was made and the change in temperature recorded as a function of time, from which the heat transfer coefficient was calculated. A comparison between steady state and transient techniques was also made on a conventional low temperature electrically heated test facility and good agreement was found. There was reasonable agreement between the high temperature heat transfer coefficients and the low temperature results, using similar transient cooling techniques. There was little influence of the coolant to wall temperature ratio on the impingement heat transfer coefficient or the cooling effectiveness obtained from the high temperature test rig. The transient technique was used to study the influence of crossflow in the impingement gap on both the cooling effectiveness and the heat transfer coefficient for a range of low pressure loss impingement walls with an X/D of 1.9–11.

Commentary by Dr. Valentin Fuster
1990;():V004T09A033. doi:10.1115/90-GT-286.
FREE TO VIEW

The present study reports on experimental information regarding the turbulent free convection flow and thermal field structure, along an isothermally heated vertical flat plate with ribs. The hight to distance ratio of the ribs was equal to 0.1.

The obtained experimental data consist of visualization of the flow, as well as detailed mean temperature and turbulent intensity distribution measurements covering the ribbed part of the heated plate. Heat transfer rates and their enhancement, due to the presence of the ribs, have been estimated from the obtained temperature profiles and by using Newton’s cooling law. Comparisons with corresponding measurements in forced convection reported in the literature, are also made.

The described experiments constitute a first approach towards a more detailed investigation regarding the significant free convection currents existing under certain conditions in cooled channels. However, they also reveal some physical aspects of the heat transfer enhancement processes in ribbed surfaces which are believed to be applicable in more general situations and could lead to establishing some design criteria for ribbed surfaces.

Commentary by Dr. Valentin Fuster
1990;():V004T09A034. doi:10.1115/90-GT-300.
FREE TO VIEW

A new method has been developed for measuring local heat transfer coefficients at rough surfaces. The technique was applied to an idealised section of a large scaled model of a turbine blade cooling passage to assess the effect of surface irregularities which result from the blade manufacturing process. The experimental method is described in full and the results are presented for tests on an isolated pin-fin situated in fully developed channel flow. The effect of the thermal conductivity of the roughness elements is discussed.

Commentary by Dr. Valentin Fuster
1990;():V004T09A035. doi:10.1115/90-GT-301.
FREE TO VIEW

Detailed measurements of surface static pressures and heat transfer rates on the aerofoil and hub end wall of an annular nozzle guide vane (in the absence of a downstream rotor) are presented. Heat transfer rates have been measured using thin film gauges in an annular cascade in the Pyestock Isentropic Light Piston Cascade. Test Mach numbers, Reynolds numbers and cascade geometry are fully representative of engine conditions. The results of 3-D calculations of surface Mach number and 2-D calculations of aerofoil heat transfer are presented and compared with the measurements.

A new method of calculating end wall heat transfer using the axisymmetric analogue for three-dimensional boundary layers is described in detail. The method uses a 3-D Euler solver to calculate the inviscid surface streamlines along which heat transfer coefficients are calculated. The metric coefficient which describes the lateral convergence or divergence of the streamlines is used to include three-dimensional effects in the calculation. The calculated heat transfer rates compare well with the measured values. Reference is made to surface flow visualization in the interpretation of the results.

Commentary by Dr. Valentin Fuster
1990;():V004T09A036. doi:10.1115/90-GT-329.
FREE TO VIEW

A numerical method is presented for the determination of heat transfer rates in a cylindrical cooling duct within turbine blades which rotate about an axis orthogonal to its own axis of symmetry.

The equations of motion and energy are solved in conjunction with the k-ε model of turbulence using the finite element method. The predicted results are compared with experimental data and it is clearly demonstrated that conduction in the solid boundary must be taken into account if satisfactory agreement is to be achieved.

Excluding these effects can lead to an over-estimation of the maximum wall temperature by approximately 50%.

Commentary by Dr. Valentin Fuster
1990;():V004T09A037. doi:10.1115/90-GT-330.
FREE TO VIEW

Heat transfer and leakage loss measurements were obtained for compressible flows in typical straight-through labyrinth seals with high rotational speeds. The experiments are an extension of our earlier measurements in a stationary test facility. In order to ensure direct comparisons to the original experiments, the principal dimensions of the test facility and gas dynamic parameters of the hot gas were kept similar. The new study encompasses a wide range of Taylor numbers, Reynolds numbers and clearances between the rotating annular fins and the stationary shroud. Heat transfer coefficients are determined for the stator as well as for the rotor. Temperature measurements along the cooled rotor were performed utilizing a high accuracy telemetric system. Continuous clearance control was achieved by employing specially designed gauges. Detailed pressure and temperatur measurements in the axial as well as in the circumferential direction were performed. Heat transfer coefficients and loss parameters are presented and compared with those obtained under steady state conditions.

Commentary by Dr. Valentin Fuster
1990;():V004T09A038. doi:10.1115/90-GT-331.
FREE TO VIEW

Experiments were conducted to determine the effects of buoyancy and Coriolis forces on heat transfer in turbine blade internal coolant passages. The experiments were conducted with a large scale, multi–pass, smooth–wall heat transfer model with both radially inward and outward flow. An analysis of the governing flow equations showed that four parameters influence the heat transfer in rotating passages (coolant–to–wall temperature ratio, Rossby number, Reynolds number and radius–to–passage hydraulic diameter ratio). These four parameters were varied over ranges which are typical of advanced gas turbine engine operating conditions. It was found that both Coriolis and buoyancy effects must be considered in turbine blade cooling designs and that the effect of rotation on the heat transfer coefficients was markedly different depending on the flow direction. Local heat transfer coefficients were found to decrease by as much as 60 percent and increase by 250 percent from no rotation levels. Comparisons with a pioneering stationary vertical tube buoyancy experiment showed reasonably good agreement. Correlation of the data is achieved employing dimensionless parameters derived from the governing flow equations.

Topics: Heat transfer
Commentary by Dr. Valentin Fuster
1990;():V004T09A039. doi:10.1115/90-GT-332.
FREE TO VIEW

Turbine blade cooling is imperative in advanced aircraft engines. The extremely hot gases that operate within the turbine section require turbine blades to be cooled by a complex cooling circuit. This cooling arrangement increases engine efficiency and ensures blade materials a longer creep life. One principle aspect of the circuit involves serpentine internal cooling passes throughout the core of the blade. Roughening the inside surfaces of these cooling passages with turbulence promoters provides enhanced heat transfer rates from the surface.

The purpose of this investigation was to study the effects of rotation, aspect ratio, and turbulator roughness on heat transfer in these rib-roughened passage. The investigation was performed in an orthogonally rotating setup to simulate the actual rotation of the cooling passages. Single pass channels, roughened on two opposite walls, with turbulators positioned at a 45° angle to the flow, in a criss-cross arrangement, were studied throughout this experiment. The ribs were arranged such that their pitch-to-height ratio remained at a constant value of 10. An aspect ratio of unity was investigated under three different rib blockage ratios (turbulator height/channel hydraulic diameter) of 0.1333, 0.25, and 0.3333. A channel with an aspect ratio of 2 was also investigated for a blockage ratio of 0.25. Air was flown radially outward over a Reynolds number range of 15000 to 50000. Rotation number was varied from 0 to 0.3. Stationary and rotating cases of identical geometries were compared. Results indicate that rotational effects are more pronounced in turbulated passages of high aspect and low blockage ratios for which a steady increase in heat transfer coefficient is observed on the trailing side as rotation number increases while the heat transfer coefficient on the leading side shows a steady decrease with rotation number. However, the all-smooth-wall classical pattern of heat transfer coefficient variation on the leading and trailing sides is not followed for smaller aspect ratios and high blockage ratios when the relative artificial roughness is high.

Topics: Heat transfer
Commentary by Dr. Valentin Fuster
1990;():V004T09A040. doi:10.1115/90-GT-333.
FREE TO VIEW

Unshrouded blades of axial turbine stages move in close proximity to the stationary outer seal, or shroud, of the turbine housing. The pressure difference between the concave and convex sides of the blade drives a leakage flow through the gap between the moving blade tip and adjacent wall. This clearance leakage flow and accompanying heat-transfer are of interest because of long obvious effects on aerodynamic performance and structural durability, but understanding of its nature and influences has been elusive. Previous studies indicate that the leakage through the gap is mainly a pressure-driven flow whose magnitude is related strongly to the airfoil pressure loading distribution and only weakly, if at all, to the relative motion between blade tip and shroud. A simple flow and heat-transfer model incorporating these features can be used to estimate both tip and shroud heat transfer provided that reasonable estimates of the clearance gap size and clearance leakage flow can be made. The present work uses a numerical computation of the leakage flow to link the model to a specific turbine geometry and operating point for which a unique set of measured local tip and shroud heat fluxes are available. The resulting comparisons between the model estimates and measured heat-transfer are good. The model should thus prove useful in the understanding and interpretation of future measurements, and should additionally prove useful for providing early design estimates of the levels of tip and shroud heat transfer that need to be compensated for by active turbine cooling.

Commentary by Dr. Valentin Fuster
1990;():V004T09A041. doi:10.1115/90-GT-334.
FREE TO VIEW

Film-cooling in the presence of mainstream pressure gradients typical of gas turbines has been studied experimentally on a flat plate This paper describes, measurements of the spanwise averaged effectiveness and heat transfer coefficient for an inclined slot and a single row of holes in the presence of favourable, zero and adverse pressure gradients. Acceleration parameters of K = 2.62×10−6 and - 0.22 × 10−6 were achieved at the point of injection where the freestream unit Reynolds number was held constant at Re/m = 2.7 × 107. The flow was accelerated to high Mach number and results are analysed using a superposition model of film-cooling which included the effects of viscous energy dissipation. The experimental results show the effects of pressure gradient differ between the geometries and a discussion of these results is included. The unblown turbulent boundary layer with pressure gradient were also studied. Experiments were performed using the Isentropic Light Piston Tunnel, a transient facility which enables conditions representative of those in the engine to be attained.

Commentary by Dr. Valentin Fuster

Electric Power

1990;():V004T10A001. doi:10.1115/90-GT-112.
FREE TO VIEW

The Combined Cycle Plant (CCP) offers the best solutions to curb air pollution and the greenhouse effect, and it represents today the most effective heat engine ever created. At Graz University of Technology work is conducted in close cooperation with industry to further enhancement of CC systems with regard to raising output and efficiency. Feasibility studies for intake air climatization, overload and partload control, introduction of aeroderivate gas turbines in conjunction with high temperature steam cycles, proposals for cooling and the use of hydrogen as fuel are presented.

Commentary by Dr. Valentin Fuster
1990;():V004T10A002. doi:10.1115/90-GT-284.
FREE TO VIEW

Parts of the latest design, incorporating improvements developed over 19 years of operating experience, can be utilized on the Model 7001B gas turbines.

This paper describes the initial application of the 7001E technology to an entire turbine retrofit. Component design improvements, interfaces, and limitations are described as well as expected reliability improvements.

Emphasis will be given to Field Service Engineering experience in installing this first change, since labor can be a significant part of the conversion.

Topics: Reliability , Turbines
Commentary by Dr. Valentin Fuster
1990;():V004T10A003. doi:10.1115/90-GT-335.
FREE TO VIEW

This paper presents a comparison of measures to improve the efficiency of combined gas and steam turbine cycles. A typical modern dual pressure combined cycle has been chosen as a reference. Several alternative arrangements to improve the efficiency are considered. These comprise the dual pressure reheat cycle, the triple pressure cycle, the triple pressure reheat cycle, the dual pressure supercritical reheat cycle and the triple pressure supercritical reheat cycle. The effect of supplementary firing is also considered for some cases. The different alternatives are compared with respect to efficiency, required heat transfer area and stack temperature. A full exergy analysis is given to explain the performance differences for the cycle alternatives. The exergy balance shows a detailed breakdown of all system losses for the HRSG, steam turbine, condenser and piping.

Topics: Combined cycles
Commentary by Dr. Valentin Fuster
1990;():V004T10A004. doi:10.1115/90-GT-345.
FREE TO VIEW

The new 200 MW-class MS 9001F gas turbines allow combined cycle plants to reach even higher output levels and greater efficiency ratings. Size factor and higher firing temperatures, with a 3-pressure level steam reheat cycle, offer plant efficiencies in excess of 53 %. Heat recovery steam generators have been designed to accommodate catalytic reduction elements limiting flue gas NOx emissions to as low as 10 ppm VD (15 % O2). A range of steam turbine models covers the different possible configurations.

Various arrangements based on the 350 or 650 MW power generation modules can be optimally configured to the requirements of each site.

Commentary by Dr. Valentin Fuster
1990;():V004T10A005. doi:10.1115/90-GT-351.
FREE TO VIEW

This paper describes a highly flexible computer method for thermodynamic power cycle calculations (PCC). With this method the user can model any cycle scheme by selecting components from a library and connecting them in an appropriate way. The flexibility is not restricted by any predefined cycle schemes.

A power cycle is mathematically represented by a system of algebraic equations. The structure of mathematical cycle models as well as different approaches to set up and solve the resulting equations with computer programs are discussed in the first section.

The second section describes the developed method. The mass and energy balance equations are set up and solved with an semi-parallel algorithm. As input only the cycle’s topology and component parameters must be entered. Information about the calculation sequence and the convergence method can be omitted completely. The example of two simple steam cycles demonstrates the applied technique.

The method requires only a few — if any at all — iterations. Calculation time and storage requirements can be kept low enough to calculate even very complex cycles on personal computers. At the end of the paper input data and results for a complex cycle scheme as it may occur in reality are given to demonstrate the performance finally.

Commentary by Dr. Valentin Fuster
1990;():V004T10A006. doi:10.1115/90-GT-362.
FREE TO VIEW

The 501F is a 150 MW-class 60 Hz engine jointly developed by Westinghouse Electric Corporation and Mitsubishi Heavy Industries, Ltd. This paper describes the full load shop test program for the prototype engine, as carried out in Takasago, Japan. The shop test included a full range of operating conditions, from startup through full load at the 1260°C (2300°F) design turbine inlet temperature. The engine was prepared with more than 1500 instrumentation points to monitor flow path characteristics, metal temperatures, displacements, pressures, cooling circuit characteristics, strains, sound pressure levels, and exhaust emissions. The results of this shop test indicate the new 501F engine design and development effort to be highly successful. The engine exceeds power and overall efficiency expectations, thus verifying the new concepts and design improvements.

Topics: Combustion , Turbines
Commentary by Dr. Valentin Fuster
1990;():V004T10A007. doi:10.1115/90-GT-363.
FREE TO VIEW

The fully loaded factory test of the CW251B12 45 MW class industrial gas turbine is described in this paper. This gas turbine is the latest uprating of the W251 series of engines. The main objectives of the factory test were the verification of the performance and the mechanical integrity of the new engine model. A brief description of the main features of the engine, the application of the first unit, the test facility, and the engine instrumentation used in the test is included. Details of the engine performance test results, telemetry test data results, and the hot end component metal temperature measurements are provided.

Topics: Gas turbines
Commentary by Dr. Valentin Fuster
1990;():V004T10A008. doi:10.1115/90-GT-365.
FREE TO VIEW

This paper describes existing, developing, and needed methods for detection, identification, and diagnosis of problems in combustion turbines. The use of combustion turbines for electrical power generation is growing, and advanced models of large industrial turbines are now starting to enter service. In view of the harsh operating conditions and severe service to which these new turbines will be exposed, this paper evaluates sensors and signal analysis methods to detect and diagnose the problems which may surface in operation. Generic problems which have been observed in combustion turbine installations in the recent past are identified, and methods for detecting these problems, quantifying them, and isolating their causes are analyzed.

Commentary by Dr. Valentin Fuster
1990;():V004T10A009. doi:10.1115/90-GT-366.
FREE TO VIEW

In this paper, a general analysis of combined gas-steam cycles for power plants firing with both hydrocarbons and coal derived gas is reported.

The purpose of this paper is to study the influence on power plants performance of different kind of fuels and to evaluate the most significant parameters of both gas and combined cycle.

Results are presented for plant overall efficiency and net specific work, steam to gas mass flow ratio, dimensionless gas turbine specific speed and diameter, CO2 emissions etc., as functions of gas cycle pressure ratio and of the combustion temperature.

Furthermore, for an existing power plant with a 120 MW gas turbine, the authors try to establish in which measure the combined cycle characteristic parameters, the gas turbine operating conditions, and the heat recovery steam generator efficiency, are modified by using synthetic fuels of different composition and calorific value.

The influence is also analyzed either of bottoming steam cycle saturation pressure or — in a dual pressure steam cycle — of dimensionless fraction of steam mass flow in high pressure stream.

The acquired results seem to constitute useful information on the criteria for the optimal design of a new integrated coal gasification combined cycle (IGCC) power plant.

Commentary by Dr. Valentin Fuster
1990;():V004T10A010. doi:10.1115/90-GT-367.
FREE TO VIEW

The environmental impact of unfired combined-cycle blocks of the GUD® type is compared with that of equivalent reheat steam boiler/turbine units. The outstandingly high efficiency of GUD blocks not only conserves primary-energy resources, but also commensurately reduces undesirable emissions and unavoidable heat rejection to the surroundings.

In addition to conventional gas or oil-fired GUD blocks, integrated coal-gasification combined-cycle (ICG-GUD) blocks are investigated from an ecological point of view so as to cover the whole range of available fossil fuels. For each fuel and corresponding type of GUD power plant the most appropriate conventional steam-generating unit of most modern design is selected for comparison purposes. In each case the relative environmental impact is stated in the form of quantified emissions, effluents and waste heat, as well as of useful byproducts and disposable solid wastes.

GUD blocks possess the advantage that they allow primary measures to be taken to minimize the production of NOx and SOx, whereas both have to be removed from the flue gases of conventional steam stations by less effective and desirable, albeit more expensive secondary techniques, e.g. flue-gas desulfurization and DENOX systems. In particular, the comparison of CO2 release reveals a significantly lower contribution by GUD blocks to the greenhouse effect than by other fossil-fired power plants.

Commentary by Dr. Valentin Fuster
1990;():V004T10A011. doi:10.1115/90-GT-370.
FREE TO VIEW

An analysis on off–design performances of fully–fired gas turbine / boiler cogeneration is presented. The complication can be eliminated by using non–dimensional parameters and dimensional analysis in off–design investigation. The suitable dimensionless parameters on boiler side have been derived, which can be easily applied to the correction of cogeneration test on site at the first stage of plant commissioning process if the parameter variations with ambient condition are not supplied by manufacturer. According to the method developed, the off–design performances of a demonstration fully–fired combined cycle cogeneration in China have been analyzed. Two operating choices are investigated thoroughly. From the results obtained, some considerations of part–load performances of such type power system are given.

Commentary by Dr. Valentin Fuster
1990;():V004T10A012. doi:10.1115/90-GT-371.
FREE TO VIEW

The Chugoku Electric Power Co., Ltd. is currently constructing a 700MW combined cycle power plant at Yanai, Yamaguchi Prefecture, based on new design concepts.

This power plant consists of six single-shaft combined cycle units with construction in two phases.

Operation of the first phase, which includes three units with an output of 350 MW, will begin in November, 1990; operation of the second phase, which also includes three units with an output of 350 MW, will begin in December, 1992.

The notable features of this power plant are: High efficiency; short starting and stopping time; suitability for daily start/stop and large load variations; thorough NOx countermeasures; easy-to-use man-machine interface control systems; and short construction period based on improvement of construction methods.

The NOx countermeasures for the gas turbines are based on a newly-developed dry low NOx combustor which is designed so that the NOx emission level is 75 ppm or less for the entire load.

A functionally and hierarchically distributed digital control system has been used for the plant control system, so that the reliability, maintenability, and controllability will be improved.

In particular, an improvement was made to the man-machine interface. A CRT operation system and a large 110-inch screen are used together with an operation support system and a power plant office automation system.

Equipment installation is proceeding smoothly and field testing of the first units will begin in April, 1990.

Because this power plant is located in the Inland Sea National Park, thorough consideration has been taken in its appearance, such as color coordination, and landscaping of the power station area.

Commentary by Dr. Valentin Fuster
1990;():V004T10A013. doi:10.1115/90-GT-375.
FREE TO VIEW

This paper summarizes a body of experience with external fires, such as those involving lube oil, liquid fuel, or fuel gas, in gas turbine–generator installations. It then describes one such fire in a sprinklered building, indicating the lack of effectiveness of the sprinklers in controlling the fire and preventing damage to the building. Conclusions are then drawn as to the optimum methods of protecting against such fires.

A concept of using excess–flow check valves in fuel lines, and throughout the lubrication system, for maximum protection from fires and other effects of a fuel or lubrication line break or separation, is described.

Topics: Fire , Gas turbines
Commentary by Dr. Valentin Fuster
1990;():V004T10A014. doi:10.1115/90-GT-378.
FREE TO VIEW

Following a thorough market analysis, the MS 9001F heavy duty gas turbine has been designed using aerodynamic scaling based on the 60 Hz MS 7001F.

Effort put into the design has been shared by the engineering departments of ALSTHOM and GE.

This paper discusses the market surveys for large heavy duty gas turbines as well as the basis of design for the MS 9001F, which has been derived from the MS 7001F. Specifically discussed are the role of scaling, the design characteristics of the MS 7001F and the MS 9001F, the results of 7001F prototype testing, the test plan for the MS 9001F, plant lay out possibilities and ratings.

The MS 9001F gas turbine uses advanced aircraft engine technology in its design, with a rating based on a firing temperature of 1260°C (2300°F), which is 156°C (280°F) higher and with compressor inlet flow 50% greater than its predecessor, the MS 9001E.

Topics: Gas turbines
Commentary by Dr. Valentin Fuster
1990;():V004T10A015. doi:10.1115/90-GT-379.
FREE TO VIEW

In this paper we investigate the possible penetration of combined-cycle plants in the Belgian electricity generation system after the Belgian Government has not considered as appropriate the construction of a next nuclear plant at the present time. First the characteristic features of the Belgian production capabilities are given. The share of gas turbines, turbojets and already existing combined-cycle plants and their operation modes are emphasized. Then, alternative options to nuclear energy are presented, i.e. repowering of existing plants and construction of new combined-cycle plants. The potentialities of gas turbines and CC plants as well as their future in Belgium are investigated.

Finally we discuss the equipment plan for the Belgian generation system proposed by the management committee of the electrical plant operators.

From the results of our research about repowering, gas turbines and new CC plants, we derive recommendations for the future production means in Belgium.

Commentary by Dr. Valentin Fuster
1990;():V004T10A016. doi:10.1115/90-GT-392.
FREE TO VIEW

A database consisting of 18 heavy-duty and aeroderivative gas turbine engine models, fired on natural gas, is evaluated for NOx exhaust emissions with and without water and steam injection. CO exhaust emissions are also considered. Engine baseload power outputs range from 2.9 to 83.5 MW, compressor pressure ratios are from 7.2 to 30.0, and turbine inlet temperatures are from 1150 to 1515K. The engine models are from the late 1970s to the current period, and all use diffusion flame combustors.

Baseload, uncontrolled NOx exhaust emissions, corrected to 15% O2 dry conditions, vary from 67 to 240 ppmv. CO exhaust emissions vary from 7 to 96 ppmv. Except for three low-NOx aero-derivative engines, the uncontrolled NOx exhaust emissions scale with engine pressure ratio and fuel-air ratio. A correlation formula is developed, and discussed relative to formulas in the literature. NOx control by water injection shows a fairly wide band; at a water-to-fuel mass ratio of 0.8, the NOx reduction varies from 58 to 82 percent. Engines with the highest uncontrolled NOx show the largest percentage reduction by water injection. On the other hand, NOx control response with steam injection exhibits less variation across the engine models. The relation of CO to NOx levels and and the response of CO to water and steam injection are examined, though quantitative correlations are not made.

Commentary by Dr. Valentin Fuster

Industrial and Cogeneration

1990;():V004T11A001. doi:10.1115/90-GT-071.
FREE TO VIEW

This paper presents the results of emissions testing and combustion system dynamics testing of a “Quiet Combustor” equipped MS7001E gas turbine at the Midway Sunset Cogeneration Company in Fellows, California. Water injection is used to control NOx emissions to 25 ppmvd without selective catalytic reduction. Test results include NOx, CO, unburned hydrocarbons, VOC, and formaldehyde emissions levels, and combustor dynamic pressure levels. Combustion system hardware mechanical performance is described following the initial combustion system inspection.

Commentary by Dr. Valentin Fuster
1990;():V004T11A002. doi:10.1115/90-GT-072.
FREE TO VIEW

A new 25MW class gas turbine H-25 has been developed to meet both 50 Hz and 60 Hz power generation needs with high reliability and efficient performance utilizing advanced high temperature gas turbine technologies. Because of its superior performance, over 32% in simple cycle efficiency, the H-25 gas turbine enables significant reduction of fuel consumption compared with previous 25MH class turbines. The H-25 also achieves higher thermal efficiency in a combined cycle power plant and a cogeneration plant with a waste heat recovery boiler. The first unit entered commercial operation at the end of November 1988 after full load factory test instrumented with more than 800 pieces of instrumentation.

This paper describes the design features and the latest technological refinements in each component of the H-25 gas turbine. Also presented are the component test programs and the results of full load factory test and field test on the first unit.

Topics: Gas turbines
Commentary by Dr. Valentin Fuster
1990;():V004T11A003. doi:10.1115/90-GT-103.
FREE TO VIEW

The baseline principles and deposit formation mechanism in the axial and centrifugal compressors of the power plants of various applications are discussed in the paper. The results of experimental investigation of fouling effect on the compressor and plant performances as well as the dismantling technique of deposit cleaning are shown.

Commentary by Dr. Valentin Fuster
1990;():V004T11A004. doi:10.1115/90-GT-104.
FREE TO VIEW

This paper describes the major machinery used in a 108 megawatt combined cycle Cogeneration Plant at the Union Carbide’s Seadrift petrochemical complex. Philosophy of design, erection, start-up, and the first year of operation is discussed. Specific machinery problems and their remedies are described.

Commentary by Dr. Valentin Fuster
1990;():V004T11A005. doi:10.1115/90-GT-105.
FREE TO VIEW

The NH3 slip pollutant generated by a recently completed cogeneration plant equipped with an SCR system was accurately monitored by a calculation method without the expense of an NH3 analyzer. The method of calculating the NH3 slip developed by the authors is based on measured variables and material balances. Discussed in further detail is the cogeneration plant, SCR characteristics, source of NH3 slip, variable measurements required for the NH3 slip calculation, equation development, accuracy of results, and NH3 analysis methods.

Commentary by Dr. Valentin Fuster
1990;():V004T11A007. doi:10.1115/90-GT-107.
FREE TO VIEW

On-line compressor cleaning of heavy duty gas turbines in industrial cogeneration service has proven to be a cost effective method of reducing the rate of performance loss associated with compressor fouling. By selecting an injection system, and by varying the dosage, frequency, and cleaning products used, an optimum cost effective method can be achieved. This paper evaluates the online compressor cleaning system at two 300 Mw cogeneration plants (8 gas turbines) that have been using this system since 1986.

Commentary by Dr. Valentin Fuster
1990;():V004T11A008. doi:10.1115/90-GT-139.
FREE TO VIEW

Within the past few years, gas turbines have been integrated in several new world-class ethylene production plants, for the first time using the exhaust as a source of preheated combustible oxygen for the cracking furnaces. The economic inducements and technical impact of such integration on the process are discussed. The general ethylene cracking and recovery process is described, and the various ways of integrating gas turbines are compared, culminating in the current leading designs. Means of providing ambient air backup to protect furnace operation from gas turbine trips are discussed. Furnace group sizing and oxygen demand for the major feedstocks, including naphtha and ethane/propane, are compared with the current range of oxygen and power available from single and dual gas turbines on the world market. Methods of partial integration, where gas turbine integration of the entire ethylene plant would produce more power than can be economically utilized or consume more premium fuel than available, are discussed. Fuel savings relative to ambient air operation are parametrized with percent exhaust oxygen and exhaust temperature. Aeroderivative and industrial gas turbine types are compared. Comparative economics with another means of gas turbine cogeneration, that of auxiliary boiler replacement with a combined cycle in a central utility, are presented.

Commentary by Dr. Valentin Fuster
1990;():V004T11A009. doi:10.1115/90-GT-176.
FREE TO VIEW

Each combustion turbine installation has peculiar to it site-related characteristics that, if considered carefully, will define the type of inlet system that best fits the operational requirements. Historically, and unless owner specified, the OEM/Packager supplies his standard, all-purpose air filter configuration as an integral part of the package. The standard may or may not be the best selection for filtering combustion inlet air at the intended specific site location. This technical paper will match inlet air filtration systems to several typical ambient conditions, using laboratory tests and actual site experience. This paper can also be used as a technically current selection guide for those considering various inlet air filtration requirements for new and retrofit systems.

Commentary by Dr. Valentin Fuster
1990;():V004T11A010. doi:10.1115/90-GT-177.
FREE TO VIEW

The need for adequate inlet air treatment for combustion gasturbines has been recognized for years. The benefit of high efficiency air filtration, mainly in high ambient dust environments does not require any further discussion as the shortfalls of the multi-stage static systems are known. The size of the conventional automatic pulse jet self cleaning airfilters is a serious handicap when considering upgrading/retrofitting outdated static filter systems or for any application in very limited spaces.

The new filter concept is described. It is an automatic Pulse Jet high efficiency self cleaning filter in which the major limiting factor of the previous design has been turned into an upgrading feature; the incoming air velocity direction basically eliminates the well-known pulsed dust re-entrainment. Other key design characteristics allow for capitalizing upon the reduced foot print area and justify the technical and economical replacement of the out-of-date filters with no or minor changes to the support structure. The system can in certain cases be incorporated within the old filterhouse itself.

Additional interesting features are discussed, supported by laboratory test results and field data from units operating in various environmental conditions.

Topics: Filtration
Commentary by Dr. Valentin Fuster
1990;():V004T11A011. doi:10.1115/90-GT-178.
FREE TO VIEW

‘Single-Blade’ Diverters are now a standard component of gas turbine combined cycle and co-generation plants. The paper explains the key features which have to be considered in selecting a Diverter, such as the sealing system, flap assembly construction, drive components and actuation systems. Further the latest hydraulic operating systems are described. With these it is now possible to use Diverters to finely regulate the heat input to the recovery system as well as to switch over rapidly to simple cycle operation or auxiliary firing. Finally the paper shows how it is possible with a two module Diverter to avoid using separate Isolators (dampers) even for the largest gas turbines currently available.

Topics: Blades
Commentary by Dr. Valentin Fuster
1990;():V004T11A012. doi:10.1115/90-GT-179.
FREE TO VIEW

Direct heat cogeneration using gas turbines presents an unusually attractive means of supplying two end products when both are required simultaneously. Most cogeneration applications involve steam as the thermal end product. While steam has many industrial and commercial uses, the conversion of thermal output to usable steam involves losses in the heat recovery steam generator (HRSG). When the heat rejected by a gas turbine can be used directly, system losses can be minimized. One such application exists in the kaolin industry.

Kaolin is a white alumina-silicate clay used in many products, including paper, medicines, plastics and paints. One step in producing kaolin requires large spray dryers, the heat for which can be supplemented by cogeneration.

This study outlines conditions necessary to achieve economic viability with cogeneration and suggests ways to maximize economic benefits. Risks associated with cogeneration ventures are also discussed.

Commentary by Dr. Valentin Fuster
1990;():V004T11A013. doi:10.1115/90-GT-180.
FREE TO VIEW

In this paper a new process is described to improve the efficiency of power generation in waste incineration plants. It is proposed to superheat the steam in a gas turbine/waste heat boiler combination rather than in the incinerator boiler.

Although this is not discussed in this paper the same process may be applied in nuclear power stations where the superheating of steam is also a problem.

A global risk-analysis of the addition of equipment is made in this paper. On basis of a case-study the financial advantages are demonstrated.

Commentary by Dr. Valentin Fuster
1990;():V004T11A014. doi:10.1115/90-GT-181.
FREE TO VIEW

Fluidized-bed catalytic cracking is one of the most commonly used processes in petroleum refining to convert heavy high-boiling point components of crude oil into gasoline and distillate components.

An energy conservation measure for such a process, utilizes an axial flow compressor which furnishes air for combustion in the regenerator where the coke deposits are burned off the catalyst and also drives the catalyst through the system.

The flue gases from the regenerator are expanded in an expansion turbine which drives the compressor, whereas the excess energy is used to drive an electric generator. The exhaust gases are utilized further in a heat recovery boiler to produce process steam.

A parametric study involving variation of air pressure and expander inlet temperature using specially devised computer program, was used to analyse the performance of the proposed system. Furthermore, the system is economically evaluated and compared with the conventional cracking system using a gas turbine engine. The proposed system offers leading performance and economic advantages in comparison with the conventional one.

Commentary by Dr. Valentin Fuster
1990;():V004T11A015. doi:10.1115/90-GT-182.
FREE TO VIEW

A new, 13MW class, heavy duty gas turbine, the “MF-111” was developed for use as a prime mover for cogeneration, combined cycle and repowering applications. The use of such equipment in refineries presents special challenges as regards the combustion of nonstandard fuels, tolerance of industrial environments, and accomodation of site-specific design requirements. Such circumstances add substantially to the tasks of proving and adjusting the design of a new gas turbine, meeting stringent emissions requirements and introducing to the world of industrial gas turbines the benefits of F-class (1250°C burner outlet temperature) levels of thermodynamic performance.

This paper describes how these challenges have successfully been met during the three calendar years and ten machine-years of MF-111 refinery-application experience accumulated to-late.

Topics: Gas turbines
Commentary by Dr. Valentin Fuster
1990;():V004T11A016. doi:10.1115/90-GT-183.
FREE TO VIEW

With emission standards becoming more and more stringent, reducing Lube Oil Reservoir Vent (LOV) mist emissions is receiving increased attention. This paper explores the nature of this oil mist, explains emission standards, describes some of the equipment that has been used in an attempt to control these emissions, and details the use of fiber beds in this application. Fiber beds have been more successful than other approaches at controlling LOV mist because of their ability to exceed emission standards with low pressure drop, minimal maintenance requirements, and guaranteed collection efficiency.

Commentary by Dr. Valentin Fuster
1990;():V004T11A017. doi:10.1115/90-GT-250.
FREE TO VIEW

Gas Turbine Power Augmentation in a Cogeneration Plant using inlet air chilling is investigated. Options include absorption chillers, mechanical (electric driven) chillers, thermal energy storage. Motive energy for the chillers is steam from the gas turbine exhaust or electrical energy for mechanical chillers. Chilled water distribution in the inlet air system is described. Overall economics of the power augmentation benefits is investigated.

Topics: Gas turbines
Commentary by Dr. Valentin Fuster
1990;():V004T11A018. doi:10.1115/90-GT-283.
FREE TO VIEW

Cogeneration has proven to be an ideal application for gas turbines, one which subsidiaries of Texaco Producing Inc. and Mission Energy Company have developed in three successful projects to date. The 38 MW Mid-Set project is the latest, having been commissioned for commercial operation in April, 1989.

Several of the Mid-Set design features were developed from past operating experience, while other design features were employed to meet requirements specific to the project. Strict pollution control required the use of water injection, selective catalytic reduction, continuous emissions monitoring and minimal production of waste water. Other design considerations were plant availability, operability and efficiency for the continuous, base load operation of a turbine-generator. The design features and initial operating history of the Mid-Set cogeneration plant are described in this paper.

Commentary by Dr. Valentin Fuster
1990;():V004T11A019. doi:10.1115/90-GT-309.
FREE TO VIEW

This paper examines the utilisation of aero gas turbines fitted to aircraft which are close to the end of their useful lives. When the aircraft are scrapped the engines can be removed, modified and employed for land or sea applications. The engine chosen as a possible candidate for ‘recycling’ is a two spool bypass engine.

A performance analysis has been carried out, which indicates that this scheme can yield good levels of output and efficiency.

Preliminary examinations indicate that there are economic advantages in converting these engines for other uses. Two possible conversions are examined: one for a pure industrial engine, and one as the gas side of a combined cycle power plant. The results obtained from this feasibility analysis appear attractive; the anticipated cost of purchasing and conversion is predicted to be significantly lower than purchasing new equipment.

Commentary by Dr. Valentin Fuster
1990;():V004T11A020. doi:10.1115/90-GT-325.
FREE TO VIEW

The technological developments in the steam boiler field, induced by changes in the use of different fins and the need to save energy, has its consequential effects on the heat transfer techniques.

As the extended surface tube is one of the important components of the steam boilers and particularly of waste heat recovery boilers (WHRB) behind turbines, we want to draw the attention to recent developments in the field of finned tubes, opening new possibilities for the optimization of construction costs of WHRB on one hand and reducing the fouling problem by a new geometry of the helically wound fins on the other hand.

Commentary by Dr. Valentin Fuster
1990;():V004T11A021. doi:10.1115/90-GT-368.
FREE TO VIEW

This paper discusses the design, installation, operation and performance of a direct mixing evaporative cooling system as an alternate to conventional wetted rigid media systems used for gas turbine power augmentation. The economic return on investment, the ease of installation, and the potential problems as they apply to a typical gas turbine based cogeneration facility are discussed and analyzed. Especially of interest is the low capital cost of this system, approximately one fourth that of a conventional system.

Commentary by Dr. Valentin Fuster
1990;():V004T11A022. doi:10.1115/90-GT-369.
FREE TO VIEW

A kind of developed exhaust-heated cycle arrangement comprising two or three reheaters and three intercoolers as well as an extra primary air blower (atmospheric combustion) is proposed. As far as classical exhaust-heated cycle is concerned, its thermal efficiency is quite low (15–20%) and therefore, a conception of a developed system provided with additional heat exchangers and combustion chambers has been worked out. The proposed installation operates on natural gas, which is becoming more and more popular now. A numerical example of calculation of a plant of 5250 kW effective power is presented. The “free-shaft” principle has been adopted. All the compressors are driven by a set of two turbines and the other two turbines are used for driving the electric generator. (Extra power that could be obtained from natural gas expander was not taken into account). In general it may be stated that the modified exhaust-heated cycle arrangement may compete with closed cycle turbine arrangements (e.g. Escher Wyss type). Moreover it may — to some extent — be superior to semi-closed cycle arrangements, since in the system proposed all the turbines are driven by compressed air whereas in the semi-closed cycle two turbines are driven by combustion gases.

Commentary by Dr. Valentin Fuster
1990;():V004T11A023. doi:10.1115/90-GT-372.
FREE TO VIEW

Steam injection has been employed in gas turbines for over twenty years. Initially the emphasis was on injection for small amounts of power augmentation and NOx reduction in the turbine exhaust gas. More recently it has been used for massive power increases (more than 50% on some gas turbines) and efficiency improvements (more than 20%). Equipment selection, operation and economics are essential ingredients in producing the high-purity steam required in a steam-injected gas turbine cycle.

The most common means of producing steam for the steam-injection cycle is by means of a waste heat boiler operating in the turbine exhaust gas stream. Steam generated in this boiler may then be injected into the compressor discharge, combustor or turbine sections of the gas turbine to improve performance.

Manufacturers require extremely high purity steam for injection into their gas turbines; less than 30 parts per billion (PPB) of some contaminants is not an unusual requirement. If this steam quality is not obtained, serious damage can occur, particularly in the turbine hot section. To meet these stringent steam quality requirements without excessive amounts of boiler blowdown, it is necessary to provide highly demineralized makeup water to the boiler, i.e. less than 1 PPM TDS (Total Dissolved Solids). Low silica concentrations are particularly important since silica can vaporize at higher boiler pressures, pass through the moisture separators and deposit on turbine components. The selection of equipment required to produce high quality makeup water from various grades of raw water is critical to the successful operation of the steam injection plant. Because the steam cannot be recovered effectively, it is necessary to install a large water treatment system to provide the quantities of makeup required for steam injection. Equally critical to the cycle is the type of drum moisture separation used in achieving manufacturers’ recommended steam quality.

Just as the steam injection cycle has a dramatic impact on the economics of a gas turbine power plant, so too do the operation and selection of steam purification equipment influence the overall economics of the steam injection cycle.

Commentary by Dr. Valentin Fuster

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In