0

ASME Conference Presenter Attendance Policy and Archival Proceedings

2017;():V05BT00A001. doi:10.1115/GT2017-NS5B.
FREE TO VIEW

This online compilation of papers from the ASME Turbo Expo 2017: Turbomachinery Technical Conference and Exposition (GT2017) represents the archival version of the Conference Proceedings. According to ASME’s conference presenter attendance policy, if a paper is not presented at the Conference by an author of the paper, the paper will not be published in the official archival Proceedings, which are registered with the Library of Congress and are submitted for abstracting and indexing. The paper also will not be published in The ASME Digital Collection and may not be cited as a published paper.

Commentary by Dr. Valentin Fuster

Internal Air Systems and Seals (With Turbomachinery)

2017;():V05BT15A001. doi:10.1115/GT2017-63001.

Conventional gas turbine secondary air system in early stage typically uses a fixed throttle unit between supply side which is compressor bleeding point and demand side which is turbine blade. The cooling air mass flow strongly depends on the extraction pressure characteristics of compressor. Optimal amount of cooling air is supplied only in design point in this way. The cooling air mass flow would be either too much or too less in off design condition.

Recently, heavy duty gas turbine manufacturers introduced an active control method for secondary air system. The main strategy is to adjust the cooling air valve set point as a function of gas turbine load percentage in order to adjust cooling air pressure ratio and cooling air mass flow as well. With this active control strategy, cooling mass flow is separated from compressor extraction pressure characteristics, and it can provide a better way to deal with combustion contaminant issues. But it is still a problem that there is no dependence relationship between cooling air valve set point and operating ambient temperature in that strategy. That is to say, the cooling air pressure ratio is constant while varying ambient temperature at base load.

In order to quantitatively analyze this phenomena, a 1-dimensional integrated gas turbine thermodynamic analysis method is first applied to obtain the extraction pressure characteristics of compressor for all bleeding points. In the meantime, the optimal cooling air mass flow for turbine blades in different operating conditions is evaluated by a 0-dimensional heat transfer assessment method. A 1-dimensional fluid network analysis method is then employed to calculate the cooling air mass flow variation characteristics for 2 typical throttle configurations between compressor bleeding points and turbine blades, the first one is setting a fixed throttle unit, and the second one is setting constant cooling air pressure ratio by a cooling air control valve. Quantitative calculation results show that the cooling air supply will not always meet the optimal requirements at different ambient temperature conditions with neither of the 2 configurations.

This paper further optimized the active control strategy. With the optimized strategy, cooling air supply not only no longer depends on extraction characteristics of compressor, but also could be actively adjusted according to the optimal requirements of turbine blades at different ambient temperature conditions. Performance evaluation results show that the optimized active control strategy could enhance the overall efficiency without exceeding maximum allowable metal temperature of turbine blades.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A002. doi:10.1115/GT2017-63060.

The cavities between the rotating compressor discs in aeroengines are open, and there is an axial throughflow of cooling air in the annular space between the centre of the discs and the central rotating compressor shaft. Buoyancy-induced flow occurs inside these open rotating cavities, with an exchange of heat and momentum between the axial throughflow and the air inside the cavity. However, even where there is no opening at the centre of the compressor discs — as is the case in some industrial gas turbines — buoyancy-induced flow can still occur inside the closed rotating cavities. The closed cavity also provides a limiting case for an open cavity when the axial clearance between the cobs — the bulbous hubs at the centre of compressor discs — is reduced to zero.

Bohn and his co-workers at the University of Aachen have studied three different closed-cavity geometries, and they have published experimental data for the case where the outer cylindrical surface is heated and the inner surface is cooled. In this paper, a buoyancy model is developed in which it is assumed that the heat transfer from the cylindrical surfaces is analogous to laminar free convection from horizontal plates, with the gravitational acceleration replaced by the centripetal acceleration. The resulting equations, which have been solved analytically, show how the Nusselt numbers depend on both the geometry of the cavity and its rotational speed. The theoretical solutions show that compressibility effects in the core attenuate the Nusselt numbers, and there is a critical Reynolds number at which the Nusselt number will be a maximum. For the three cavities tested, the predicted Nusselt numbers are in generally good agreement with the measured values of Bohn et al. over a large range of Raleigh numbers up to values approaching 1012. The fact that the flow remains laminar even at these high Rayleigh numbers is attributed to the Coriolis accelerations suppressing turbulence in the cavity, which is consistent with recently-published results for open rotating cavities.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A003. doi:10.1115/GT2017-63091.

The industry bristle material of choice for brush seals has been the cobalt-based alloy Haynes®25 (also known as L605) for over 30 years. Haynes®25 has excellent oxidation resistance and wear properties in brush seal applications up to temperature of 620°C [1148°F]. Above this temperature creep resistance becomes undesirable for brush seal bristles and has lead to alternative sealing solutions to be implemented in these turbine locations. Nickel-based alloys have been explored as an alternative for Haynes®25 but have been shown to gall badly and wear quickly in comparison.

As increases in turbine performance have resulted in an increase in operating temperatures, it has lead to a need to find a bristle material that performs at temperatures above the limitation of Haynes®25. Initial experimental data has been obtained for a new cobalt-based alloy that shows potential for use as a bristle material at temperatures above 620°C [1148°F]. Further experimental results also indicate that the material appears to have better wear characteristics than Haynes®25 and may prove to be a feasible alternative in some cases.

This paper outlines a material selection process for brush seals, along with development of the alloy for use within brush seals and details of the comparative testing carried out at Cross Manufacturing Company.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A004. doi:10.1115/GT2017-63163.

The objective of this study is to develop a Computational Fluid Dynamics (CFD) based methodology for analyzing and predicting leakage of worn or rub-intended labyrinth seals during operation. The simulations include intended tooth axial offset and numerical modeling of the flow field. The purpose is to predict total leakage through the seal when an axial tooth offset is provided after the intended/unintended rub.

Results indicate that as expected, the leakage for the in-line worn land case (i.e. tooth under rub) is higher compared to unworn. Furthermore, the intended rotor/teeth forward axial offset/shift with respect to the rubbed land reduces the seal leakage. The overall leakage of a rubbed seal with axial tooth offset is observed to be considerably reduced, and it can become even less than a small clearance seal designed not to rub. The reduced leakage during steady state is due to a targeted smaller running gap because of tooth offset under the intended/worn land groove shape, higher blockages, higher turbulence and flow deflection as compared to worn seal model without axial tooth offset.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A005. doi:10.1115/GT2017-63201.

Windage loss in small, high speed electrical machinery is often predicted using fairly simple quasi-empirical correlations. Many of the correlations used are primarily based on testing performed with larger test articles, at lower speeds, and often with liquid lubricants. This paper presents a new set of air gap windage loss test data for test articles that are more nearly representative of small, high speed electrical machinery. These data were obtained using a unique new test rig. This rig was designed around test articles that are representative of 50 to 200 kW machinery operating up to 60 krpm with air as the fluid in the rotor-stator clearance. This paper describes the new test facility and presents data for a smooth surface 72.4 mm rotor with both a smooth stator and a stator with simulated winding slots, for a range of clearances. The smooth surface results are shown to be in reasonable agreement with previously published results for annular gap windage power loss.

Topics: Machinery
Commentary by Dr. Valentin Fuster
2017;():V05BT15A006. doi:10.1115/GT2017-63423.

Because of the superior sealing characteristics compared to labyrinth seals, brush seals found an increased spread in turbomachinery in recent years. Their outstanding sealing performance results mainly from their flexibility. Thus, a very small gap between the rotor and bristle package can be obtained without running the risk of severe detrimental deterioration in case of rubbing. Rubbing between rotor and seal during operation might occur as a result of e.g. an unequal thermal expansion of the rotor and stator or a rotor elongation due to centrifugal forces or manoeuvre forces. Thanks to the flexible structure of the brush seal the contact forces during a rubbing event are reduced, however the frictional heat input can still be considerable. Particularly in aircraft engines with their thin and lightweight rotor structures the permissible material stresses can easily be exceeded by an increased heat input and thus harm the engine’s integrity. The geometry of the seal has a decisive influence on the resulting contact forces and consequently the heat input. The complex interactions between the geometric parameters of the seal and the heat input and leakage characteristics are not yet fully understood. This paper presents the investigation of the influence of the geometric parameters of a brush seal on the heat input into the rotor and the leakage behaviour. Two seals with different packing densities were tested under relevant engine conditions with pressure differences ranging from 1 to 7 bar, relative surface speeds ranging from 30 to 180 m/s and radial overlaps ranging from 0.1 to 0.4 mm. The transient temperature rise during the rub event was recorded with 24 thermocouples in close proximity to the rub contact embedded in the rotor structure. By comparing the temperature curves with the results of a thermal finite element analysis of the rotor the heat input into the rotor was calculated iteratively. It could be shown that the packing density has a decisive influence on the overall operating behaviour of a brush seal. Furthermore, results are obtained for the heat flux distribution between seal and rotor are shown.

Topics: Heat , Leakage
Commentary by Dr. Valentin Fuster
2017;():V05BT15A007. doi:10.1115/GT2017-63505.

The modern gas turbine is widely applied in the aviation propulsion and power generation. The rim seal is usually designed at the periphery of the wheel-space and prevented the hot gas ingestion in modern gas turbines. The high sealing effectiveness of rim seal can improve the aerodynamic performance of gas turbines and avoid of the disc overheating. Effect of outer fin axial gap of radial rim seal on the sealing effectiveness and fluid dynamics was numerically investigated in this work. The sealing effectiveness and fluid dynamics of radial rim seal with three different outer fin axial gaps was conducted at different coolant flow rates using three-dimensional Reynolds-Averaged Navier-Stokes (RANS) and SST turbulent model solutions. The accuracy of the presented numerical approach for the prediction of the sealing performance of the turbine rim seal was demonstrated. The obtained results show that the sealing effectiveness of radial rim seal increases with increase of coolant flow rate at the fixed axial outer fin gap. The sealing effectiveness increases with decrease of the axial outer fin gap at the fixed coolant flow rate. Furthermore, at the fixed coolant flow rate, the hot gas ingestion increases with the increase of the axial outer fin gap. This flow behavior intensifies the interaction between the hot gas and coolant flow at the clearance of radial rim seal. The preswirl coefficient in the wheel-space cavity is also illustrated to analyze the flow dynamics of radial rim seal at different axial outer fin gaps.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A008. doi:10.1115/GT2017-63512.

Rotating air inside the wheel-space creates a radial gradient of pressure which drives the gas ingress through the rim seal. This kind of reason for the gas ingestion is called rotationally induced ingress (RI). The minimum sealing flow rate was proportional to the seal-clearance. The geometric structure, including the position of the seal-clearance, is also important to predict the minimum sealing flow rate for RI ingestion. This paper gets the sealing efficiency and the flow results of different geometric structure through the method of 3D steady compressible CFD (Computational Fluid Dynamics). Because the analysis of the influence of geometry is given under the condition of RI ingestion, a 3D model without turbine blades has been chosen.

Some experiments initially revealed that the different seal-clearance positions have different sealing efficiency. However, what position would have best sealing efficiency was not given. If the position of seal-clearance is selected in the rotor disc or the static disc, the effect of the “pump” of the rotor disc is more obvious, which makes the gas ingestion serious. When the position of seal-clearance is near the rotor disc, the gas is fully mixed with the cooling air after the ingestion and then flows to the side of the static disc. Therefore, the sealing efficiency of the structure, whose seal-clearance position is near the rotor disc, will be higher than that, whose seal-clearance position is close to the static disc. When the fluid flows to the static disc, the velocity triangle shows that a barrier will be created between the cavity and mainstream in a particular seal-clearance position, which makes the efficiency higher than other positions.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A009. doi:10.1115/GT2017-63531.

Recent experimental work from the present authors demonstrated that interactions between the mainstream and cavity/rim seal flows lead to ingestion mechanisms with a range of length scales. In addition to the (well known) effect of the vane and blade pressure fields, it was demonstrated that the shear layer instabilities between the mainstream and rim seal flows can affect ingress. Building upon these observations and the understanding in the literature, this paper presents a model which relates rotor-stator cavity seal effectiveness to purge flow rate based on turbulent transport. The main assumption is that all length scales of ingress lead to an effective eddy diffusivity. This eddy diffusivity drives ingress across the seal concentration gradient. Following Prandtl’s mixing length hypothesis for eddy viscosity, the model uses an empirical constant representing an equivalent mixing length. This assumption is shown to be sufficient across a limited range of dimensionless flow rates. An extension of the model is presented to account for the reduction in turbulent mixing in the rim seal recirculation region as it becomes washed out with increasing purge flow. The rate at which the effect of the rim seal recirculation region gets washed out is modelled with a purge-to-mainstream blowing ratio term and the volume fraction of the seal occupied by the rim seal recirculation. The differences in volume fraction and blowing ratio between the different experiments in literature are defined by the geometry and flow condition only. By fitting, it is shown that the model is sufficient to capture a wide variety of experimental data in the literature and that of the present authors. The results and the model derivation provide an encouraging first step and a framework towards a model that is sensitized to both geometry and flow conditions.

Topics: Turbulence
Commentary by Dr. Valentin Fuster
2017;():V05BT15A010. doi:10.1115/GT2017-63562.

Annular gas seals for compressors and turbines are designed to operate in a nominally centered position in which the rotor and stator are at concentric condition, but due to the rotor-stator misalignment or flexible rotor deflection, many seals usually are suffering from high eccentricity. The centering force (represented by static stiffness) of an annular gas seal at eccentricity plays a pronounced effect on the rotordynamic and static stability behavior of rotating machines.

The paper deals with the leakage and static stability behavior of a fully-partitioned pocket damper seal (FPDS) at high eccentricity ratios. The present work introduces a novel mesh generation method for the full 360° mesh of annular gas seals with eccentric rotor, based on the mesh deformation technique. The leakage flow rates, static fluid-induced response forces and static stiffness coefficients were solved for the FPDS at high eccentricity ratios, using the steady Reynolds-Averaged Navier-Stokes (RANS) solution approach. The calculations were performed at typical operating conditions including seven rotor eccentricity ratios up to 0.9 for four rotational speeds (0 rpm, 7 000 rpm, 11 000 rpm and 15 000 rpm) including the non-rotating condition, three pressure ratios (0.17, 0.35 and 0.50) including the choked exit flow condition, two inlet preswirl velocities (0 m/s, 60 m/s). The numerical method was validated by comparisons to the experiment data of static stiffness coefficients at choked exit flow conditions. The static direct and cross-coupling stiffness coefficients are in reasonable agreement with the experiment data. An interesting observation stemming from these numerical results is that the FPDS has a positive direct stiffness as long as it operates at subsonic exit flow conditions, no matter the eccentricity ratio and rotational speed are high or low. For the choked exit condition, the FPDS shows negative direct stiffness at low eccentricity ratio and then crosses over to positive value at the crossover eccentricity ratio (0.5–0.7) following a trend indicative of a parabola. Therefore, the negative static direct stiffness is limited to the specific operating conditions: choked exit flow condition and low eccentricity ratio less than the crossover eccentricity ratio, where the pocket damper seal would be statically unstable.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A011. doi:10.1115/GT2017-63565.

Active use of dry gas seals for gas turbine engines is constrained by several important factors. One of them is a significant deformation of the sealing rings. This paper is dedicated to the creation of a technique of designing of a dry gas seal with spiral grooves with a consideration of arbitrary gap shape. A large number of studies for this type of seal have been conducted. But the technique of the seal designing which combines sealing leakage calculation with the calculation of the actual rings deformation has not been implemented. This article proposes a solution for this significant problem. Indeed, the increase in temperature and pressure drop results in a deformation of the rings surfaces. For the small gap, the impact of force and thermal deformation is critical. The seal designing without consideration of the surfaces deformation can lead to significant errors, but also to the failure of the seal in operation in the worst case. An improved mathematical model for calculating the leakage is proposed. On its basis, the designing technique has been developed. This technique combines the analytical calculation and calculation of deformation by finite element method. Implementation of this technique has a good practical result. The seal was created for a gas pumping unit. Experimental results have confirmed the computational results.

Topics: Design , Shapes
Commentary by Dr. Valentin Fuster
2017;():V05BT15A012. doi:10.1115/GT2017-63647.

The progressive upgrading of heavy-duty gas turbines, aimed at increased performance, can ultimately introduce more onerous operating conditions, to the point that original design limits can be approached.

An increased gas turbine pressure ratio together with compression and expansion line adjustments can directly affect the rotor axial thrust. Other than the individual forces acting on the rotor, a key component to be taken into account is the fluid film thrust bearing, which should assure safe and reliable operation during the worst case operating conditions. Typically, such bearings are designed with large safety margins, yet it is possible that the new and more challenging conditions require a bearing capability upgrade, especially when field retrofit needs pose additional constraints.

A succession of performance upgrades have been carried out on Ansaldo Energia’s AE94.2 E-Class GT. An accurate understanding of the thrust-related phenomena proved necessary and drove improvements in the thrust bearing design along with hardware adjustments to lower the rotor thrust.

This paper addresses calculations and experimental arrangements for the rotor axial thrust evaluation on the aforementioned GT and considers both the matters related to the secondary air system for the thrust generation and the mechanical/functional matters for the bearing upgrade.

It is shown that issues such as uneven load sharing across the thrust bearing, or the variability of rotor thrust from engine to engine within the fleet, strongly affect the maximum thrust assessment and thus the requirements used in the process of selecting a suitable bearing. A predictive calculation method is described considering the main thrust contributions. Field experimental setups and main observations are reported. Measurements have been carried out using thermocouples and load cells placed on many of the thrust bearing pads. Moreover, the engine cavities carrying the highest and/or the most uncertain thrust share have been instrumented and characterized by pressure sensors.

The development of an upgraded thrust bearing is finally depicted through the main issues addressed, such as improved thrust pad lining material, increased load sharing efficiency and enlarged thrust bearing active surface area. Waukesha Bearings test results on the upgraded lining material, a high-tin aluminium alloy are reported as well.

A multidisciplinary approach is presented as necessary to manage the crucial challenge of improving the thrust balancing system, especially in the case of a formerly designed engine which receives a powerful upgrade.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A013. doi:10.1115/GT2017-63734.

Fast response pressure data acquired in a high-speed 1.5-stage turbine Hot Gas Ingestion Rig shows the existence of pressure oscillation modes in the rim-seal-wheelspace cavity of a high pressure gas turbine stage with purge flow. The experimental results and observations are complemented by computational assessments of pressure oscillation modes associated with the flow in canonical cavity configurations. The cavity modes identified include shallow cavity modes and Helmholtz resonance. The response of the cavity modes to variation in design and operating parameters are assessed. These parameters include cavity aspect ratio, purge flow ratio, and flow direction defined by the ratio of primary tangential to axial velocity. Scaling the cavity modal response based on computational results and available experimental data in terms of the appropriate reduced frequencies appears to indicate the potential presence of a deep cavity mode as well. While the role of cavity modes on hot gas ingestion cannot be clarified based on the current set of data, the unsteady pressure field associated with turbine rim cavity modal response can be expected to drive ingress/egress.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A014. doi:10.1115/GT2017-63841.

This paper is a continuation of a previous comparison dealing with URANS-based validation of the ASU-Honeywell turbine stage mainstream/disc-cavity interaction rig data. Here, the validation is with a CFD code named PowerFLOW which is based on the Lattice Boltzmann Method (or LBM). Transient LBM simulations were conducted across the previously published purge flows (Cw of 1540 to 6161), and at the higher mainstream flow condition of 2300 cfm (1.086m3/s). Sensitivity of convergence on results was investigated by increasing the number of revolutions, as well as by varying the passive scalar and temperature difference assumptions between mainstream and purge flow. Results indicate that at lower purge flow, LBM was able to significantly improve validation of sealing effectiveness measurements. For the intermediate purge flows, however, there is a departure from what the data shows. Finally, at the higher purge flow cases, LBM prediction improves at the outer radial location as compared to URANS. Moreover, on pressure validation, it has closed the gap in matching the measured steady pressures inside the lower disc cavity except at the highest purge flow. In the critical upper rim cavity, the gap between the two methods closes as purge flow increases. The outcome from this comparative tool validation study is that at the low critical purge flow case where ingestion is most critical as well as at the upper rim cavity location, sealing effectiveness predictions were significantly improved. The paper also discusses the current limitations of LBM.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A015. doi:10.1115/GT2017-63844.

The interaction between the mainstream and disc cavity purge flows in a turbine stage is an unsteady 360° phenomenon. Most of the current rotating rigs have used steady pressure transducers to measure the mainstream annulus pressure distributions as well as the pressure distribution in the disc cavity. Unsteady static pressure measurements in these regions using fast-response transducers have also been reported but to a much lesser degree, mainly at ASU, OSU, VKI, and ETH. To gain better insight into the prevailing unsteady flow phenomena, and to assess the difference between steady and time-averaged unsteady pressure data, new unsteady static pressure measurements were recently carried out at three locations in an ASU-Honeywell turbine stage, namely, in the main gas path on the outer shroud near vane trailing edge as well as on the vane platform lip, and on the stator surface rim seal. They are reported in this paper along with the comparative results of the corresponding URANS CFD simulation reported in an earlier publication. Experiments were carried out at five different purge air flow conditions for each of the two mainstream air flow rate and rotor speed combinations. The current unsteady measurements indicate that the rim cavity pressure frequency is governed by the blade passage frequency. The unsteadiness amplitude increases with purge flow in the main gas path, but decreases with increase in purge flow for the rim cavity where the sensitivity to change in purge flow is smaller at the lower mainstream flow rate. The difference in the ambient-corrected time-averaged static pressures between those evaluated from the current unsteady measurements and the previously published steady measurements are found to be within the measurement uncertainties.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A016. doi:10.1115/GT2017-63910.

Secondary air is bled from the compressor in a gas turbine engine to cool turbine components and seal the cavities between stages. Unsealed cavities can lead to hot gas ingestion, which can degrade critical components or, in extreme cases, can be catastrophic to engines. For this study, a 1.5 stage turbine with an engine-realistic rim seal was operated at an engine-relevant axial Reynolds number, rotational Reynolds number, and Mach number. Purge flow was introduced into the inter-stage cavity through distinct purge holes for two different configurations. This paper compares the two configurations over a range of purge flow rates. Sealing effectiveness measurements, deduced from the use of CO2 as a flow tracer, indicated that the sealing characteristics were improved by increasing the number of uniformly distributed purge holes and improved by increasing levels of purge flow. For the larger number of purge holes, a fully sealed cavity was possible while for the smaller number of purge holes, a fully sealed cavity was not possible. For this representative cavity model, sealing effectiveness measurements were compared with a well-accepted orifice model derived from simplified cavity models. Sealing effectiveness levels at some locations within the cavity were well-predicted by the orifice model, but due to the complexity of the realistic rim seal and the purge flow delivery, the effectiveness levels at other locations were not well-predicted.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A017. doi:10.1115/GT2017-63951.

Turbine blades and the disks are connected by tenons. There is a pair of jagged assembly clearance between each tenon and corresponding mortise. In practical engineering applications, flow and heat transfer characteristics in assembly clearance used to be simplified. In order to obtain more accurate temperature fields of the turbine blades and disks, detailed study of the flow and heat transfer mechanism in tenon joint gap is necessary.

In this paper, two typical assembly clearances under the stationary and rotating conditions were investigated numerically, including double S-shaped and double Crescent-shaped. The inlet Reynolds numbers range from 5,500 to 50,000 and the Rotation numbers range from 0 to 0.005. The results show that the fluids in the two branches of the double S-shaped channel have different flow characteristics under rotating conditions. A vortex is formed at the corner of the left branch and the vortex scale can be influenced by Re and Ro. The large vortex decreases the local heat transfer coefficient. In the right branch, the three-dimensional flow from the flat wall to the concave wall increases the local heat transfer coefficient of different regions. For the double Crescent-shaped channel, the region with higher velocity is offset to the right of the channel which leads to higher local heat transfer coefficient under rotating conditions. The simulation results have great significance to the heat transfer analysis of turbine blades and disks.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A018. doi:10.1115/GT2017-64143.

The cooling air in the secondary air system of gas turbines is controlled and metered by numerous restrictors, mainly in the shape of orifices. The ability to understand and predict the associated pressure losses are important in order to improve the air flow in the secondary air system. This experimental study investigates the behavior of the discharge coefficient for circular orifices with inlet cross flow which is a common flow case in gas turbines. Examples of this are at the inlet of a film cooling hole or the feeding of air to a blade through an orifice in a rotor disc.

Measurements were conducted for a total number of 38 orifices, covering a wide range of length-to-diameter ratios, including short and long orifices with varying inlet geometries. Up to five different chamfer-to-diameter and radius-to-diameter ratios were tested per orifice length. Furthermore, the static pressure ratio across the orifice was varied between 1.05 and 1.6 for all examined orifices.

The results of this comprehensive investigation demonstrate the beneficial influence of rounded inlet geometries and the ability to decrease pressure losses, which is especially true for higher cross flow ratios where the reduction of the pressure loss in comparison to sharp edged holes can be as high as 54%. With some exceptions, the chamfered orifices show a similar behavior as the rounded ones but with generally lower discharge coefficients. Nevertheless, a chamfered inlet yields lower pressure losses than a sharp edged inlet. The obtained experimental data was used to develop two correlations for the discharge coefficient as a function of geometrical as well as flow properties.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A019. doi:10.1115/GT2017-64169.

Under actual operating conditions of gas turbine, centrifugal and thermal growth of disc in radial direction result in dislocation of inflow boundary at the disc mid-radius height, and a radial step of platform at high radius height. In this paper, flow and heat transfer characteristics in dislocated rim seal region are analyzed by the conjugate and thermal mechanical numerical methods.

The calculated radial growths of turbine discs reach approximately 14–20 % of turbine platform structure thickness. Dislocation of rim seal structure directly affects the flow characteristic of externally-induced (EI) ingress and rotationally-induced (RI) ingress, and aggravates overheat of stator disc due to induced hot gas ingestion, further affects the loss of mixture of mainstream gas and cooling sealant air flow in rim-seal and wheelspace regions. Radial step between rotor and stator platforms exacerbates the area and depth of hot gas ingestion in seal clearance, along with a 2–7 % decrease in seal efficiency.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A020. doi:10.1115/GT2017-64297.

The existence and causes of the deep ingress of the annulus flow into the core region of a turbine rotor-stator disc cavity, or core penetration flow, have been investigated experimentally. In addition, the effects of annulus flow coefficient, rotational Reynolds number, and non-dimensional purge air flow rate on the core penetration flow have been examined. Using the low–speed, low expansion ratio single-stage cold turbine test facility at Seoul National University (SNU), time-resolved tangential and radial velocities in the cavity have been measured with 2-D hot-wire anemometers. In addition, time-resolved static pressures on the stator disc have been measured with fast response pressure transducers, and the unsteady cavity velocity field in the absolute frame has been measured using Particle Image Velocimetry (PIV). Geometric non-axisymmetry (e.g. eccentricity of a rotor disc cover in this study) can change the cavity exit pressure, and thus the radial pressure gradient in the cavity. A time lag in the tangential velocity adjustment to the variation in the radial pressure gradient results in a net radial force, leading to core penetration flow. The core penetration flow occurs twice when the cavity exit pressure increases, and once when the cavity exit pressure decreases. In this study, with a once per revolution geometric non-axisymmetry, the core penetration flow occurs three times per revolution, revolving at the disc’s rotational speed. Variations in the annulus flow coefficient or rotational Reynolds number do not affect the core penetration flow, but increasing the purge air flow rate weakens the core penetration flow.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A021. doi:10.1115/GT2017-64388.

Modeling of hot gas ingestion in a gas turbine engine is critical because its accuracy directly affects performance as well as turbine durability. In this paper, ASU ingestion test rig data accompanied by its published ingress/egress discharge coefficients (Cdi and Cde , formulation) are used to propose a simplified 1D ingestion model embedded in the secondary air system software (Network). The proposed externally induced ingress model includes separate boundary nodes with equal static pressure in the annulus hub, and distinct circumferential pressure variation in the form of normalized annulus pressure at the hub (P1P1avg)/(P1maxP1min). The corresponding Cdi and Cde for the engine conditions are scaled based on rig-to-engine non-dimensional minimum purge, Cwmin where engine Cwmin uses the actual (P1P1avg)/(P1maxP1min) derived from previously published CFD data along with the effective rim-seal overlap clearance. The vane pitch integrated driving pressure difference at the hub for the ingestion used in the orifice model comes from an embedded saw-tooth assumption on the circumferential pressure profile. Recirculation of ingested hot gas from the upper rim cavity to the lower wheel space is considered by comparing the supplied purge flow to the rotating disc entrainment requirement. The proposed model is compared with another model based on constant Cdi / Cde ratio of 0.14 published by the University of Bath. Engine test data from a previously published engine configuration is used to assess the appropriate model for engine. The probability of failure in violating the lower rim cavity sealing effectiveness limit based on analysis of variation (AOV) was conducted under both formulations and the results are presented.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A022. doi:10.1115/GT2017-64410.

An aero-engine bearing chamber is a structure that is used to contain and collect oil used in lubricating and cooling the bearings supporting the high-speed engine shafts. There are various bearings in an aero-engine. Within the bearing chambers, there are typically the bearings, rotating shafts, seals and gears (in some designs). The walls of the bearing chamber are stationary and there are vents and sumps to take out the oil, via an offtake pipe, and the sealing air. The oil collected via the sump and vents is recycled and used again in the loop. To prevent oil degradation and reduce chance of coking in the chamber, it is desired that all of the oil goes through the recycling loop, with no oil staying longer than necessary in the chamber. The sealing air is used to maintain a positive pressure to keep the oil within the chamber. The flow inside a bearing chamber is highly turbulent and consists of a rotating mixture of oil and air.

A smaller amount of the oil, mostly as oil-droplets, exits at the vents and is separated from the air using de-aerators [1]. It is expected that by gravity, most of the oil collects at the sump and can be easily scavenged. This is provided the sump can be large enough. The geometry of a bearing chamber is, however, complex largely because of space limitations. It is very important that oil is not resident longer than necessary to prevent over-heating and therefore deterioration or coking. Experimental observations by Chandra & Simmons [2], have shown that bearing chambers with deep sumps perform better that those with shallow sumps.

Since shallow sumps are inevitable, a number of innovative studies have been done to improve bearing chamber designs. The presence of air in the oil (e.g. as bubbles) reduces the efficiency of the scavenging pump. Other factors such as oil momentum and windage can take oil away from the off-take pipe potentially increasing oil residence volume. Chandra & Simmons [2] placed inserts such as grille cover, perforated plate, etc, on a side of the bearing wall and improvements in the residence volume were seen. In this work, we are looking at a detailed computational fluid dynamics (CFD) simulation of one of the inserts that performed well. This will aid understanding of the flow characteristics of using an insert to improve oil residence in a bearing chamber.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A023. doi:10.1115/GT2017-64427.

With the major aim of gathering information on the machine lateral stability in high pressure-high density conditions, and of assessing the prediction capabilities of the in-house design tools and overall process, a back-to-back centrifugal compressor has been instrumented and tested in several operating conditions. The present paper focuses on the secondary flows across the interphase balance drum of the back-to-back compressor, where the sealing is accomplished with a honeycomb seal. The compressor interphase section has been instrumented with dedicated special probes for the clearance measurement associated to pressure and flow angle probes in order to characterize pressure distributions and swirl variations depending on the specific operating range. The experimental data acquired over the machine operation have been compared with a three-dimensional steady-state numerical analysis results obtained from the simulation, carried out with a Reynolds averaged Navier-Stokes (RANS) approach, of the flowfield in the complex interphase secondary system composed by the impeller cavities and the honeycomb seal. This paper addresses the comparison between numerical results and experimental data, which allowed the matching of models with experiments in terms of pressure distribution and the complex flowfield. Finally, all the data have been used to validate an in-house one-dimensional flow network solver for pressure distribution and leakage flow calculations along cavities and seals. Results have shown a general good agreement between measured data and calculation output. In particular, computational fluid dynamic analysis provided detailed pressure and velocity distributions that allowed gaining insight in the physics of such a complex region. The one-dimensional model has been demonstrated to be a fast and reliable tool to well predict local pressure variations inside cavities and seals and, consequently, the residual axial thrust.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A024. doi:10.1115/GT2017-64440.

Tight sealing lines are vital in large gas turbines (GT) to achieve high performance and efficiency. Leakage including rim purge air can sum up to 30% of the total cooling and leakage air consumption of a gas turbine. Leakage through static strip seals contributes about 1/3 to all leakage air. Considering the seal design as on drawings, sealing quality is generally influenced by the seal type, sealing groove curvature and the sealing groove roughness. In addition the sealing quality depends strongly on the geometric deviation of the groove compared to ideal design. This is caused by manufacturing deviations or relative movements of the grooves during operation of the parts containing the sealing.

In the article at hand, different seal designs and pertinent sealing quality is discussed. More in detail, it is discussed the geometric relation of seal, groove and misalignment to predict the seal position relative to its groove confinements. The risk of seal clamping can be judged and adaptation of seal or groove geometry can be derived. The effect of leakage increase due to misalignment is investigated by a test matrix varying seal length and curvature radius of groove as well as radial misalignment.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A025. doi:10.1115/GT2017-64503.

This paper deals with a numerical study aimed at the validation of a computational procedure for the aerothermal characterization of pre-swirl systems employed in axial gas turbines. The numerical campaign focused on an experimental facility which models the flow field inside a direct-flow pre-swirl system. Steady and unsteady simulation techniques were adopted in conjunction with both a standard two-equations RANS/URANS modelling and more advanced approaches such as the Scale-Adaptive-Simulation principle, the SBES and LES. The comparisons between CFD and experiments were done in terms of swirl number development, static and total pressure distributions, receiving holes discharge coefficient and heat transfer on the rotor disc surface. Several operating conditions were accounted for, spanning 0.78·106<Reφ<1.21·106 and 0.123<λt<0.376. Overall the steady-state CFD predictions are in good agreement with the experimental evidences even though it is not able to confidently mimic the experimental swirl and pressure behaviour in some regions. Although the use of unsteady sliding mesh and direct turbulence modelling, would in principle increase the insight in the physical phenomenon, from a design perspective the tradeoff between accuracy and computational costs is not always favourable.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A026. doi:10.1115/GT2017-64512.

The present work aims at investigating a new methodology developed at Ansaldo Energia, for the transient finite element modelling of the whole engine with an axisymmetric approach.

The strong coupling and non linearity in the heat transfer process during transient thermal analyses are handled by a partly coupled scheme. The 2D axisymmetric finite element model includes a dedicated thermal fluid network where fluid-metal temperatures are computed. In the overall procedure the selected finite element solver is a customized version of CalculiX®, while mass flow rates and pressure distributions in each thermal fluid network element are provided by external fluid network solvers in terms of customized time series.

This paper represents a first insight about a fully integrated WEM (Whole Engine Modelling) procedure currently under development. Geometrical changes during operation, lead to different fluid properties affecting heat transfer coefficients too. These modified conditions in their turn impact the material temperature and displacements. The future implementation steps will be oriented on the adoption of a customized version of the native CalculiX® fluid network solver with the aim of developing a fully integrated procedure able to take into account the interaction between the secondary air system and the modifications in the clearances and gaps due to the thermal and mechanical loads.

In this paper, a detailed description of the procedure will be reported with comprehensive discussions about some fundamental modelling aspects. Preliminary results, related to the first application of the new methodology to the transient thermal modelling of a simplified test case representative of real engine geometries, will be presented.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A027. doi:10.1115/GT2017-64530.

Adaptive lubricants involve binary mixture of synthetic oil and dissolved carbon dioxide (CO2). Unlike conventional lubricant oils, the lubricant viscosity not only varies with the temperature within the bearing, but also can be directly adjusted through the CO2 concentration in the system. In this study, we investigated the performance of adaptive lubricants in a hybrid journal bearing considering the synthetic oil to be fully saturated by CO2. The adaptive lubricant analyzed for this study was the polyalkylene glycol (PAG) oils with low concentration of CO2 (< 30%). A three-dimensional computational fluid dynamic (CFD) model of the bearing was developed and validated against the experimental data. The mixture composition and the resultant mixture viscosity were calculated as a function of pressure and temperature using empirical equations.

The simulation results revealed that the viscosity distribution within the PAG/CO2-lubricated bearing is determined primarily by the pressure at the low operating speed. When the speed becomes higher, it is the temperature effect that dominates the viscosity distribution within the bearing. Moreover, the PAG/CO2-lubricated bearing can reduce up to 12.8% power loss than the PAG-lubricated bearing due to the low viscosity of PAG/CO2 mixture. Most importantly, we have found the PAG/CO2 can enhance the load capacity up to 19.6% when the bearing is operating at the high speed conditions.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A028. doi:10.1115/GT2017-64620.

In gas turbines, rim seals are fitted at the periphery of stator and rotor discs to minimise the purge flow required to seal the wheel-space between the discs. Ingestion (or ingress) of hot mainstream gases through rim seals is a threat to the operating life and integrity of highly-stressed components, particularly in the first-stage turbine. Egress of sealing flow from the first-stage can be re-ingested in downstream stages.

This paper presents experimental results using a 1.5-stage test facility designed to investigate ingress into the wheel-spaces upstream and downstream of a rotor disc. Re-ingestion was quantified using measurements of CO2 concentration, with seeding injected into the upstream and downstream sealing flows. Here a theoretical mixing model has been developed from first principles and validated by the experimental measurements. For the first time, a method to quantify the mass fraction of the fluid carried over from upstream egress into downstream ingress has been presented and measured; it was shown that this fraction increased as the downstream sealing flow rate increased. The upstream purge was shown to not significantly disturb the fluid dynamics but only partially mixes with the annulus flow near the downstream seal, with the ingested fluid emanating from the boundary layer on the blade platform. From the analogy between heat and mass transfer, the measured mass-concentration flux is equivalent to an enthalpy flux and this re-ingestion could significantly reduce the adverse effect of ingress in the downstream wheel-space. Radial traverses using a concentration probe in and around the rim seal clearances provide insight into the complex interaction between the egress, ingress and mainstream flows.

Topics: Gas turbines
Commentary by Dr. Valentin Fuster
2017;():V05BT15A029. doi:10.1115/GT2017-64632.

Engine designers require accurate predictions of ingestion (or ingress) principally caused by circumferential pressure asymmetry in the mainstream annulus. Cooling air systems provide purge flow designed to limit metal temperatures and protect vulnerable components from the hot gases which would otherwise be entrained into disc cavities through clearances between rotating and static discs. Rim seals are fitted at the periphery of these discs to minimise purge. The mixing between the efflux of purge (or egress) and the mainstream gases near the hub end-wall results in a deterioration of aerodynamic performance.

This paper presents experimental results using a turbine test rig with wheel-spaces upstream and downstream of a rotor disc. Ingress and egress was quantified using a CO2 concentration probe, with seeding injected into the upstream and downstream sealing flows. The probe measurements have identified an outer region in the wheel-space and confirmed the expected flow structure. For the first time, asymmetric variations of concentration have been shown to penetrate through the seal clearance and the outer portion of the wheel-space between the discs. For a given flow coefficient in the annulus, the concentration profiles were invariant with rotational Reynolds number. The measurements also reveal that the egress provides a film-cooling benefit on the vane and rotor platforms. Further, these measurements provide unprecedented insight into the flow interaction, and provide quantitative data for CFD validation, which should help reduce the use of purge and improve engine efficiency.

Topics: Turbines
Commentary by Dr. Valentin Fuster
2017;():V05BT15A030. doi:10.1115/GT2017-64703.

In order to reduce environmental and climate impact from air traffic, the main effort of aero-engine industry and research community is looking at a continuous increase in gearbox efficiency. With this kind of components every source of loss can be responsible for high heat loads; for this reason oil jet systems are used to provide proper cooling and lubrication of gears tooth surfaces. In the design phase it is important to predict the losses increase due to the lubricating oil jet impact on the spur gear, varying the different geometrical and working parameters such as the jet inclination, distance and the oil mass flow rate and temperature. An experimental investigation was carried out on a novel rotating test rig able to reproduce real engine working conditions in terms of speed, pressure and lubrication system, for a single spur gear. The rig consists of an electric spindle driving a shaft with a spur gear clamped on top. The gear is enclosed in a box where different air pressure conditions can be set and monitored. Pressure transducers and T-type thermocouples placed within the test box were used to measure the gear working conditions. The test box is also equipped with several optical accesses allowing flow field measurements or oil jet visualizations. The driving shaft is composed by two parts connected by a bearingless torquemeter equipped with a speedometer in order to perform torque losses and rotating velocity measurements. Tests were performed without the gear first, in order to separate the final value from the friction losses due to the driving shaft. Windage losses were characterized experimentally for every working condition and the results collected in a simple correlation that was used to separate the losses due to air windage from the ones due to the oil injection. An oil control unit allowed to impose the proper oil pressure and temperature conditions and to measure the mass flow rate. The oil jet was delivered by a spraybar placed within the gearbox, the jet to gear distance and relative angle were varied during the experiments. High speed visualizations were also performed for every test condition in order to deepen the physical understanding of the phenomena and to obtain more information on the lubrication capability of every jet condition. A high speed camera was placed in front of the gear exploiting an optical access while a halogen lamp was used to provide the proper lightening necessary due to the very low exposure time of the acquisitions. The wide experimental database provided, allowed the development of a simple numerical model able to well predict every losses contribution at the various working conditions.

Topics: Lubrication
Commentary by Dr. Valentin Fuster
2017;():V05BT15A031. doi:10.1115/GT2017-64864.

Developments in brush seal analyses tools have been covering advanced flow and structural analyses since brush seals are applied at elevated pressure loads, temperatures, surface speeds, and transients. Brush seals have dynamic flow and structural behaviors that need to be investigated in detail in order to estimate final leakage output and service life. Bristles move, bend and form a grift matrix depending on pressure load. The level of pressure load determines the tightness of the bristle pack, and thus, the leakage. In the CFD analyses of this work, the bristle pack is treated as a porous medium. Based on brush seal test data, the flow resistance coefficients (FRC) for the porous bristle pack are calibrated as a function of pressure load. A circular seal is tested in a static test rig under various pressure loads at room temperature. The FRC calibration is based on test leakage and literature based axial pressure distribution on the rotor surface and radial pressure distribution over the backing plate. The anisotropic FRC are treated as spatial dependent in axi-symmetrical coordinates. The fence height region and the upper region of bristle pack have different FRC since the upper region is supported by backing plate while bristles are free to move and bend at the fence height region. The FRC are found to be almost linearly dependent on the pressure load for investigated conditions. The blow-down is also calculated by incorporating test leakage and calibrated FRC.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A032. doi:10.1115/GT2017-64884.

This paper presents new experimental measurements, at conditions representative of an aero engine, of heat transfer from the inner peripheral surface (shroud) of a rotating cavity. The results are taken from the University of Sussex Multiple Cavity Rig, which is designed to be similar to a gas turbine high pressure compressor internal air system. The shroud Nusselt numbers are shown to be dependent on the shroud Grashof number and insensitive to throughflow axial Reynolds number. The magnitude of the shroud Nusselt numbers are consistent with accepted correlations for turbulent free convection from a horizontal plate, yet show a trend (gradient of Nusselt to Grashof numbers) that is similar to laminar free convection.

A supporting high-resolution 3D unsteady RANS simulation was conducted to investigate the cavity flow structure with particular attention paid to the near shroud region. This demonstrated flow structures that are consistent with published work but also show the existence of a type of Rayleigh-Bénard flow that manifests as a series of streaks that propagate along the periphery of the cavity. These structures can be found in the literature albeit in different circumstances. Whilst these streaks have been shown in the simulation their existence cannot be ratified without experimental confirmation.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A033. doi:10.1115/GT2017-64917.

Aside from increasing the overall engine pressure ratio, current trends in the design of innovative aircraft propulsion systems focus also on improving the propulsive efficiency. In this context, research efforts have been devoted towards the implementation of a new family of Ultra-High Bypass Ratio (up to 20) engines. In order to take into account the new constraints set by this innovative fan concept, configurations able to guarantee an independent optimization of the low pressure spool modules are more promising. This framework however introduces new challenges for the transmission system in terms of heat management and power losses, since the amount of power transferred through the gearbox is largely increased. Among the various types of losses occurring within the gearbox, whose relative magnitude largely vary depending on the operating conditions, one of the most significant at high rotational speeds is certainly windage.

This paper studies windage losses within an epicyclic gear train by means of CFD simulations. Superposition effect is assumed for the various losses contributions hence gear meshing is avoided to simplify the computational domain and separate the effects. Results obtained in this computational campaign highlight the influence of most relevant parameters such as oil concentration, chamber pressure and rotational speed. The response of the system to increased oil concentration and chamber pressure is linear with density variation, while power losses on both sun and planetary gears are shown to scale well with the third power of rotational speed. Furthermore the effects of geometrical modifications, introduced to avoid gear meshing, are studied and guidelines to scale the effective power losses is provided.

Multi-phase simulations have been carried out in order to confirm obtained trends from single-phase analysis verifying the rationality of such a computational approach.

Commentary by Dr. Valentin Fuster
2017;():V05BT15A034. doi:10.1115/GT2017-64948.

Enhancing the efficiency of gearing systems is an important topic for the development of future aero-engines with low specific fuel consumption. The transmission system in fact has a direct impact on the engine overall efficiency by means of its weight contribution, internal power losses and lubrication requirements. Thus, an evaluation of its structure and performance is mandatory in order to optimize the design as well as maximize its efficiency. Gears are among the most efficient power transmission systems, whose efficiencies can exceed 99 %, nevertheless in high speed applications power losses are anything but negligible. All power dissipated through losses is converted into heat that must be dissipated by the lubrication system. More heat leads to a larger cooling capacity, which results in more oil, larger heat exchangers which finally means more weight. Mechanical power losses are usually distinguished in two main categories: load-dependent and load-independent losses. The former are all those associated with the transmission of torque, while the latter are tied to the fluid-dynamics of the environment which surrounds the gears, namely windage, fluid trapping and squeezing between meshing gear teeth and inertial losses resulting by the impinging oil jets, usually adopted in high speed transmission for cooling and lubrication purposes. The relative magnitude of these phenomena is strongly dependent on the operative conditions of the transmission. While load-dependent losses are predominant at slow speeds and high torque conditions, load-independent mechanisms become prevailing in high speed applications, like in turbomachinery. Among fluid-dynamic losses, windage is extremely important and can dominate the other mechanisms. In this context, a new test rig was designed for investigating windage power losses resulting by a single spur gear rotating in a free oil environment. The test rig allows the gear to rotate at high speed within a box where pressure and temperature conditions can be set and monitored. An electric spindle, which drives the system, is connected to the gear through a high accuracy torque meter, equipped with a speedometer providing the rotating velocity. The test box is fitted with optical accesses in order to perform particle image velocimetry measurements for investigating the flow-field surrounding the rotating gear. The experiment has been computationally replicated, performing RANS simulations in the context of conventional eddy viscosity models. The numerical results were compared with experimental data in terms of resistant torque as well as PIV measurements, achieving a good agreement for all of the speed of rotations.

Topics: Gears
Commentary by Dr. Valentin Fuster

General Computational Heat Transfer

2017;():V05BT22A001. doi:10.1115/GT2017-63032.

The lifetime of the modern gas turbines greatly depends on the durability of hot section components operating at high temperatures. Film cooling is key to air cooling technologies in modern gas turbine and widely used in high-temperature and high-pressure blades as an active cooling scheme.

The requirements of accurate prediction of aerodynamic flow and heat transfer in gas turbine blades lay the essential foundation of cooling effectiveness improvement and working life estimation. In recent days, Large Eddy Simulations (LES) is considered as a useful tool to predict turbulent flows and heat transfer around gas turbine blades, but, comparing to the Reynolds-Averaged Navier–Stokes (RANS) methods, the LES method usually needs more computing resource and depends on computational power and mesh quality.

In this paper, LES/DES (Detached Eddy Simulation) predictions were compared to RANS prediction with interest in the accuracy and improvement of turbulent flow and heat transfer phenomena around NASA’s C3X high-pressure gas turbine vane with leading edge cooling film. RANS/LES/DES were detailed and further investigated to assess their ability to predict flow and heat transfer in boundary layer around C3X vane.

The current predictions showed that the mix between film cooling injections and free stream resulted in complex flow and heat transfer in the boundary layer on the external vane surface. The predictions of the aerodynamic load along the C3X vane with RANS/LES/DES were almost identical and agreed well with the experimental results. However, the heat transfer predictions with RANS/LES/DES were different. The transition prediction showed the best agreement with the experiment data in the most region. The LES prediction only partially agreed with the experimental data before separation point on the suction side and mild pressure gradient region on the pressure side. The DES and RANS predictions agreed with the experiment data after separation point on the suction side and most region on the pressure side.

Commentary by Dr. Valentin Fuster
2017;():V05BT22A002. doi:10.1115/GT2017-63243.

This article presents a parametric numerical study to analyze the sensitivity of wall heat fluxes on an academic acoustic liner to inlet conditions. Large Eddy Simulation (LES) is used to simulate an array of 20 aligned honeycomb cells on a flat plate with 10% porosity characteristic of installed liners. The computational domain is periodic in the span direction comprising 2 honeycomb cells. The operating conditions are representative of cruise with a Mach number of 0.5 at ambient pressure and temperature. Comparisons of heat fluxes obtained on a none perforated flat plate with the honeycomb liner are proposed with different inlet conditions: steady laminar boundary layer profile, turbulence injection and acoustic perturbation injection at different frequencies. Results show that for the operating condition and the boundary layer thickness used, large differences are observed on the first cells of the liners resulting from different transition to turbulence processes. A first important difference exists from laminar and turbulent conditions where turbulent conditions exhibits higher heat fluxes as expected. Then, case pulsed at the resonant frequency of the honeycomb shows higher heat fluxes than other frequencies. Finally, after a given number of cells, the heat fluxes reach an asymptotic behavior at the same level which seems to be controlled by the turbulence generated by the interaction of the flow and the perforations whatever the inlet conditions.

Commentary by Dr. Valentin Fuster
2017;():V05BT22A003. doi:10.1115/GT2017-63306.

Modern aeroengines utilized effusion cooling technology to further protect the components from degrading at the operating temperatures. Most studies did not address the influence of the manufacturing process used to form the cooling holes on the flow physics where percussion laser drilling was a common technique that produced irregularly shaped holes with roughened surfaces. The investigated as-drilled hole surface was statistically homogeneous, non-isotropic, and generally composed of gradually transitioning plateaus that had imperfections with an average height of 0.32 hole diameters. A conjugate heat transfer CFD study was completed on cylindrical, conical nozzle, and as-drilled holes, all yielding the same hole mass flow rates, with the realizable k-ε turbulence model at representative engine conditions. The cylindrical hole had higher film cooling effectiveness due to lower effluent velocity, and better in-hole heat transfer performance due to higher on-average in-hole flow velocities. The as-drilled hole had nominally better film cooling than the conical nozzle hole due to the higher in-hole turbulence production caused by the roughened surface texture. Ultimately, the hole area profile more significantly influenced the averaged metal temperature.

Topics: Cooling , Lasers
Commentary by Dr. Valentin Fuster
2017;():V05BT22A004. doi:10.1115/GT2017-63480.

Cooling of turbine hot-gas-path components can increase engine efficiency, reduce emissions, and extend engine life. As cooling technologies evolved, numerous blade cooling geometries have been, and continue to be proposed by researchers and engine builders for internal and external blade and vane cooling. However, the impact of these improved cooling configurations on overall engine performance is the ultimate metric. There is no assurance that obtaining higher cooling performance for an individual cooling technique will result in better turbine performance because of the introduction of additional second law losses, e.g. exergy loss from blade heat transfer, cooling air friction losses, fluid mixing, etc. and thus the higher cooling performance might not always be the best solution to improve efficiency.

To quantify the effect of different internal and external blade cooling techniques and their combinations on engine performance, a cooled engine model has been developed for industrial gas turbines and aero-engines using MATLAB Simulink®. The model has the flexibility to be used for both engine types, and consists of uncooled on-design, turbomachinery design and a cooled off-design analysis in order to evaluate the engine performance parameters by using operating conditions, polytropic efficiencies, material information and cooling system information. The cooling analysis algorithm involves a Second Law analysis to calculate losses from the cooling technique applied.

The effects of variations in engine parameters such as turbine inlet temperature, by-pass ratio, and operating temperature are studied. The impact of variations in metal Biot number, thermal barrier coating Biot number, film cooling effectiveness, internal cooling effectiveness and maximum allowable blade temperature on engine performance parameters are analyzed. Possible design recommendations based on these variations, and direction of use of this tool for new cooling design validation, are presented.

Topics: Cooling , Gas turbines
Commentary by Dr. Valentin Fuster
2017;():V05BT22A005. doi:10.1115/GT2017-63504.

Endwall 2D contouring is a typical design to reduce the strength of secondary flows within the passage. Such contouring can lead to significant changes in the passage flow. A leakage slot at the combustor-turbine interface is a typical turbine endwall design. The leakage flow can be used to cool the endwall and vane surface. Moreover, the leakage flow interacts with the main flow and results in the change of aerodynamic loss. A 3D numerical method was used to investigate endwall adiabatic effectiveness and passage total pressure loss coefficient (TPLC) on a NGV with 2D contoured endwall under a series of mass flow ratios (MFRs). The numerical method was validated by comparision with the experiment data. The results indicate that under the condition of this study, When MFR<0.625%, there is ingestion, and when 0.625%<MFR<1.0%, the TPLC is high. When 0.875%<MFR<1.0%, the area averaged adiabatic effectiveness (AAAE) decreases and the TPLC stays high as the MFR increases. When 1.0%<MFR<1.5%, the adiabatic effectiveness is high and the TPLC is low. To accomplish high adiabatic effectiveness, low aerodynamic loss and slight ingestion, the recommended range of MFR is 1.0%<MFR<1.5%.

Commentary by Dr. Valentin Fuster
2017;():V05BT22A006. doi:10.1115/GT2017-63581.

A GPU based 3D RANS multi-block slover with the capability of predicting heat transfer in fully resolved boundary layer flows based on the turbomachinery CFD code TBLOCK has been developed. A diabatic wall treatment and the scalar dissipation scheme have been integrated. The standard one-equation Spalart-Allmaras turbulence model with trip term is used for turbulence modeling. For validation, steady CFD results from the developed GPU version of TBLOCK are compared to a well-documented experiment of a highly loaded turbine cascade. The experiment covers a range of different Mach numbers and Reynolds numbers with static pressure and heat flux distributions on suction and pressure side of the blade. The predicted pressure and heat flux distribution show good agreement to the experimental data. A significant reduction of calculation time could be achieved compared to the authors’ CPU based MPI version of TBLOCK.

Commentary by Dr. Valentin Fuster
2017;():V05BT22A007. doi:10.1115/GT2017-63842.

The present study pays attention to the pressure effect and geometric effect on heat transfer deterioration onset to supercritical hydrocarbon fuel. Numerical simulation about heat transfer deterioration of hydrocarbon fuel flowing upward in vertical round tubes with various diameter at supercritical pressure was performed. In the simulation, a four-species surrogate model of RP-3 based on the generalized corresponding states law was used and LS low-Reynolds number eddy viscosity turbulence model was selected. For the boundary conditions, inlet temperature was 623K, pressure ranged from 3 to 4MPa, tube diameter varied from 3 to 9mm, and wall heat flux to mass flux ratio changed from 0.07 to 3.18kJ/kg. Comparative analyses between the predicted results and the experimental data revealed the accuracy of thermophysical property model and numerical method. The results indicated that the operating pressure and tube diameter have significant effect to the heat transfer deterioration onset of supercritical hydrocarbon fuel: heat transfer deterioration aggravates and heat transfer deterioration onset moves upstream when the diameter increases. With the increase of operating pressure, heat transfer deterioration becomes weak and the heat transfer deterioration onset moves downstream. Based on current results, several existing correlations of the heat transfer deterioration onset were reviewed and assessed, showing different prediction performance. A new correlation of the threshold value for the ratio between heat flux and mass flux for determining the boundary for heat transfer deterioration under various tube diameter and operating pressure was obtained. The effect of length to diameter ratio on safety margin was discussed. The present study provides the optimization design of regenerative cooling on reducing heat transfer deterioration.

Commentary by Dr. Valentin Fuster
2017;():V05BT22A008. doi:10.1115/GT2017-63875.

An intercooling technique using convective cooling channels in the compressor stator vanes has been proposed in recent years. In this paper, two cooling methods are presented and conjugate heat transfer method is used in the numerical simulation to study the effect of cooling on the laminar boundary layer and turbulent boundary layer in compressors. The overall performance of the compressor is also analyzed.

A flat plate in T3C series experiments under adverse pressure gradient has been simulated to verify the aerodynamic simulation and preliminarily investigate the cooling effect. Subsequently, a two-dimensional compressor vane NACA65-(12A2I8b)10 has been numerically simulated to study the cooling effect on two-dimensional boundary layer of the curve surface. The numerical simulation results of the vane without cooling channel are in good agreement with the experiment data by NASA. By comparing it with the case which has convective cooling channels, it can be found that the cooling decreases the size of laminar separation bubble and delays the turbulent separation, which reduces the loss at both the design and off-design angle of attack.

A three-dimensional highly-loaded five-stage axial compressor whose stator vanes have cooling channels and cooling endwalls has also been numerically simulated. The cooling channels and endwalls decrease the temperature rising of the main stream with a slight increase of the pressure rising, which indicates that this intercooling method can be used in the intercooled and recuperated (ICR) cycle. Cooling channels decrease the temperature of the stator vanes and protect them from the high temperature. Besides, the effect of cooling on the turbulent separation in the corner region has also been investigated. The cooling channels decrease the total pressure loss, which indicates that cooling has a beneficial effect on the aerodynamic performance of the compressor.

Topics: Cooling , Compressors
Commentary by Dr. Valentin Fuster
2017;():V05BT22A009. doi:10.1115/GT2017-63908.

A new empirical correlation for upward flowing supercritical aviation kerosene RP-3 in the vertical tubes is proposed. In order to obtain the database, numerical simulation with a four-component surrogate model on RP-3 and LS low Reynolds turbulence model in vertical circular tube has been performed. Tubes of diameter 2mm to 10mm are studied and operating conditions cover pressure from 3MPa to 6MPa. Heat flux is 500KW/m2, mass flow rate is 700kg/(m2·s). The numerical results on wall temperature distribution under various conditions are compared with experimental data and a good agreement is achieved. The existing correlations are summarized and classified into three categories. Three representative correlations of each category are selected out to evaluate the applicability in heat transfer of supercritical RP-3. The result shows that correlations concluded from water and carbon-dioxide do not perform well in predicting heat transfer of hydrocarbon fuel. The mean absolute deviation of them is up to 20% and predict about 80% of the entire database within 30% error bands. So a new correlation which is applicable to different working conditions for supercritical RP-3 is put forward. Gnielinski type has been adapted as the basis of the new correlation for its higher accuracy. In consideration of major influence factors of supercritical heat transfer, correction terms of density and buoyancy effect are added in. The new correlation has a MAD of 9.26%, predicting 90.6% of the entire database within ±15% error bands. The comparisons validate the applicability of the new correlation.

Topics: Heat transfer
Commentary by Dr. Valentin Fuster
2017;():V05BT22A010. doi:10.1115/GT2017-63949.

Cooling uniformity is of crucial importance for the cooling structure design of turbine vane and rotor blade. However, there exist no veritable quantification of the uniformity. This paper, for the first time, reveals the correlation between temperature distribution uniformities of the temperature image and its textural features, and five texture measures are employed to characterize these features and quantify the temperature distribution uniformity from different angles. Finally, their linear weighted combinations are proposed to obtain a comprehensive evaluation, and thus a systematic and quantitative methodology for cooling uniformity evaluation is established. Moreover, this method is exemplified through a flat-plate jet array impingement cooling model, where cooling uniformities over the impingement target surface under five cases are compared. Results show that both the cooling uniformity sequence and the relative changes of the uniformity composite index between the ranking agree well with those of the corresponding thermal stress analyses. This method shows a promising application in the design and optimization of cooling structures for gas turbine hot components, such as film cooling and impingement cooling.

Topics: Cooling
Commentary by Dr. Valentin Fuster
2017;():V05BT22A011. doi:10.1115/GT2017-64080.

The complex heat and mass transfer across the turbine tip gap requires a detailed analysis which cannot be expressed using the classical Newton heat convection approach. In this portion of the turbine, characterized by tight moving clearances, pressure gradients are counterbalanced by viscous effects. Hence, non-dimensional analysis, based on the boundary layer, is inadequate and therefore the use of an adiabatic wall temperature is questionable. In this paper, we propose an alternative approach to predict the convective heat transfer problem across the turbine rotor tip using Discrete Green Functions.

The linearity of the energy equation can be applied with a superposition technique to measure the data extracted from flow simulations to determine the Green’s function distribution. The Discrete Green Function is a matrix of coefficients that relate the increment of temperature observed in a surface with the heat flux integrated on the same surface. These coefficients are independent on the inlet temperature of the flow and are associated to the geometry. The controlled surface is discretized into cells and each cell is associated to a vector of coefficients. The Discrete Green Function coefficients are calculated using the temperature response of the cell to a heat flux pulse imposed at different locations.

The methodology was previously applied to a backward facing step to prove its validity. Several simulations were performed applying a representative pulse of heat flux in different locations in the bottom wall of the backward facing step. From these simulations, the increment of temperature in each node of the geometry was retrieved and the Discrete Green Function coefficients associated to the bottom wall were calculated. A numerical validation was performed imposing a random pattern of heat flux and predicting the increment of temperature on the bottom wall under different inlet flow conditions. The final aim of this paper is to demonstrate the method in the rotor turbine tip.

A turbine stage at engine-like conditions was assessed using a CFD software. The heat flux pulses were applied at different locations in the rotor tip geometry, and the increment of temperature in this zone was evaluated for different clearances, with a consequent variation of the Discrete Green Function coefficients. A validation of the rotor tip heat flux was accomplished by imposing different heat flux distributions in the studied region. Ultimately, a detailed uncertainty analysis of the methodology was included based in the magnitude of the heat flux pulses used in the Discrete Green Function coefficients calculation and the uncertainty in the experimental measurements of the wall temperature.

Topics: Turbines , Heat flux
Commentary by Dr. Valentin Fuster
2017;():V05BT22A012. doi:10.1115/GT2017-64205.

This paper presents two different numerical methods to predict the thermal load of a convection-cooled gas-turbine blade under realistic operating temperature conditions. The subject of the investigation is a gas-turbine rotor blade equipped with an academic convection-cooling system and investigated at a cascade test-rig. It consists of three cooling channels, which are connected outside the blade, so allowing cooling air temperature measurements.

Both methods use FE models to obtain the temperature distribution of the solid blade. The difference between these methods lies in the generation of the heat transfer coefficients along the cooling channel walls which serve as a boundary condition for the FE model. One method, referred to as the FEM1D method, uses empirical one-dimensional correlations known from the available literature. The other method, the FEM2D method, uses three-dimensional CFD simulations to obtain two-dimensional heat transfer coefficient distributions. The numerical results are compared to each other as well as to experimental data, so that the benefits and limitations of each method can be shown and validated.

Overall, this paper provides an evaluation of the different methods which are used to predict temperature distributions in convection-cooled gas-turbines with regard to accuracy, numerical cost and the limitations of each method. The temperature profiles obtained in all methods generally show good agreement with the experiments. However, the more detailed methods produce more accurate results by causing higher numerical costs.

Commentary by Dr. Valentin Fuster
2017;():V05BT22A013. doi:10.1115/GT2017-64451.

The compressibility of flow field has an important effect on flow stability. However, when the compressibility is considered, the effect of Mach number is often considered while the effect of heat transfer is always neglected in the existing flow stability studies. Linear stability analysis tools based on compressible Orr-Sommerfeld (O-S) equations and linearized Navier-Stokes equations in cylindrical coordinate system are established in this paper. These equations are numerically solved by using Chebyshev spectral collocation method and pseudo-modes are eliminated. Linear stability analysis of pipe flow with heat transfer whose average flow field is obtained by CFD simulation is carried out. The results show that for spatial modes, the heating effect of the wall makes pipe flow more unstable, while cooling effect of the wall makes pipe flow more stable. For global modes of pipe flow, the frequency of global mode decreases when the wall cools the flow and the decrease of mean temperature of pipe flow leads to the improvement of global mode stability.

Commentary by Dr. Valentin Fuster
2017;():V05BT22A014. doi:10.1115/GT2017-64520.

When designing a gearbox it is important to consider the heat rise generated inside the gearbox due to the gear meshing action of gear teeth. Providing efficient lubrication helps keep the gearbox at lower temperatures and reduce friction, which in return leads to a longer lifespan. Given the difficulty in obtaining experimental data within the gearbox, the authors investigate and present the setup and methods using Computational Fluid Dynamics (CFD) modelling of the process. The main purpose of this work is to implement and demonstrate numerical techniques that are needed in order to perform CFD simulations on this subject. There are currently no widely used techniques known to the authors that would allow to carry out parametric CFD study of gearbox lubrication and cooling. There are only limited empirical models that are used to find a best design. When developed, CFD methods may allow to do parametric studies and therefore significantly improve the quality of the gearbox design.

In order to capture the fluid behaviour in a continuously changing topology around rotating gears, dynamic mesh technique with remeshing and smoothing is used. Dynamic mesh is a complex and expensive technique on its own; and becomes even more so when have to be implemented along with the two-phase flow and conjugated heat transfer. For that reason the development and implementation of this method requires an incremental approach with very gradual increase of difficulty and separation of the large task into small ones, which essentially what has been done in this work. Furthermore, investigation of how to reduce the cost of the simulation is an important part so that the method can then be used more widely.

Two types of lubrication are considered: partial dipping into oil (rotational submersion) and jet spraying. Rotational speeds of up to 8,000rpm are studied. Temperature of the gears and the surrounding fluids are initially defined as uniform. Additional heat sources are created in the solid cells of the gears where the teeth come into contact, also using a UDF. 2.5D dynamic remeshing is used for models with spur gears, whereas full 3D remeshing is used with helical gears. Simulations are performed using the Volume of Fluid method and the standard k-omega turbulence model. Simulations are run with varying degrees of complexity (low- and high-fidelity).

Some results of basic preliminary simulations are compared with available results from the literature, demonstrating a good agreement. Validation of the results demonstrate the ability of the presented methods to accurately predict the gear losses and the fluid flow in a gearbox. More complex simulations are run in order to observe and analyse both the fluid flow and the heat distribution in the gearbox. Main attention is given to the temperatures of the housing and the meshing teeth. Since all simulations with meshing gears require a small gap between the gears (i.e. with no direct contact of the gears), three different gap sizes are investigated. For these simulations a comparison of the oil flow is provided. This comparison is used to justify which model can be used most efficiently without significant loss of accuracy when modelling the temperature distribution at the housing.

Current work is an essential first step towards the detailed study that is currently of great interest of both research and industry. Future work is necessary to fully justify the methods, however the current work is essential and will hopefully provide an inspiration and encouraging of the topic advancement.

Commentary by Dr. Valentin Fuster
2017;():V05BT22A015. doi:10.1115/GT2017-64711.

The drive to further increase gas turbine thermal efficiency and specific power output continue to elevate core temperatures well beyond the natural capabilities of the metals employed in their manufacture necessitating increasingly complex cooling systems.

One such cooling mechanism is the double-wall, effusion-cooled system which combines in a very compact format, many cooling aspects already implemented in gas turbine cooling. To-date, thermomechanical stresses have provided one of the more significant challenges in the implementation of these systems and needs to be considered — alongside aerothermal performance — at the initial stages of design.

This paper presents a novel computational method that has been developed to allow an integrated assessment of both the aerothermal and thermomechanical performance of double-wall cooling geometries. A decoupled conjugate method was developed in which internal cooling performance was ascertained via a conjugate CFD model in which the mainstream flow was not simulated. Instead, external film cooling performance was assessed via a superposition method that was developed and applied to a two-dimensionally varying correlation allowing streamwise film development to be modelled. Results of both the internal and external cooling simulations were then utilised in a conduction model to develop a complete thermal assessment of the geometry. The calculated temperature distribution was used in a thermomechanical FEA analysis permitting an insight into the stress field developed within the double-wall geometry under thermal load.

The developed method was demonstrated in the assessment of seven circular pedestal, double-wall geometries in which a range of geometric parameters were investigated. The results provide an insight into the effect of varying these parameters on both the aerothermal performance of the selected geometries, along with the effect on the thermomechanical stress field developed.

Topics: Thermomechanics
Commentary by Dr. Valentin Fuster
2017;():V05BT22A016. doi:10.1115/GT2017-64913.

The prediction of temperature and heat transfer throughout the solid material of a gas-turbine combustor has driven interest in cooling technology which uses impingement/effusion (IE) cooling tiles on double-skinned combustor liners. The design of the IE tile system is simple but the aerodynamics are complex. The complexity of flow curvature, combined impingement and effusion cooling and heat transfer, poses a challenge to standard RANS CFD modelling. The IE combustor tile is numerically investigated using both URANS model with the SST-SAS model and Large Eddy Simulation (LES) in the Rolls-Royce in-house CFD code. The aim is to provide accurate CFD data and to test the viability of URANS approach to predict the impingement/effusion flow. Results of pressure, velocity and turbulence quantities are presented. It is found that the SST-SAS model, with high grid resolution, shows good agreement with LES. The current CFD results are used to resolve a substantial amount of very small impingement and effusion holes. The CFD results showed that every feature of the geometry has to be resolved by the numerical mesh, which makes the simulation impractical due to time consuming and high mesh resolution. These cooling holes can be omitted from the computational mesh and their effects captured on the flow via an impingement-effusion (IE) model which is based on defining the correct mass flow inside the holes as a function of the difference of pressure in the upstream and downstream regions of both impingement and effusion regions. The latter model takes the effect of pressure and velocity and it will be extended in future to take into account the heat transfer effects. The IE model is tested and validated for the 3-D combusor tile and results of pressure showed good agreement with the LES data.

Commentary by Dr. Valentin Fuster
2017;():V05BT22A017. doi:10.1115/GT2017-64982.

One of the approaches adopted to improve turbine efficiency and increase power to weight ratio is reducing vane count. In the current study, numerical analysis was performed for the heat transfer over the surface of nozzle guide vanes under the condition of reduced vane count using three dimensional computational fluid dynamics (CFD) models. The investigation has taken place in two stages: the baseline nonfilm-cooled nozzle guide vane, and the film-cooled nozzle guide vane. A finite volume based commercial code (ANSYS CFX 15) was used to build and analyze the CFD models. The investigated annular cascade has no heat transfer measurements available; hence in order to validate the CFD models against experimental data, two standalone studies were carried out on the NASA C3X vanes, one on the nonfilm-cooled C3X vane and the other on the film-cooled C3X vane. Different modelling parameters were investigated including turbulence models in order to obtain good agreement with the C3X experimental data, the same parameters were used afterwards to model the industrial nozzle guide vanes. Three Shear Stress Transport (SST) turbulence model variations were evaluated, the SST with Gamma-Theta transition model was found to yield the best agreement with the experimental results; model capabilities were demonstrated when the laminar to turbulent transition took place.

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