The Dynamics of the Horseshoe Vortex and Associated Endwall Heat Transfer: Part 2 — Time-Mean Results

Author(s):  
T. J. Praisner ◽  
C. R. Smith

Time-mean endwall heat transfer and flow-field data in the endwall region are presented for a turbulent juncture flow formed with a symmetric bluff body. The experimental technique employed allowed the simultaneous recording of instantaneous particle image velocimetry flow field data, and thermochromic liquid-crystal-based endwall heat transfer data. The time-mean flow field on the symmetry plane is characterized by the presence of primary (horseshoe), secondary, tertiary, and corner vortices. On the symmetry plane the time-mean horseshoe vortex displays a bimodal vorticity distribution and a stable-focus streamline topology indicative of vortex stretching. Off the symmetry plane, the horseshoe vortex grows in scale, and ultimately experiences a bursting, or breakdown, upon experiencing an adverse pressure gradient. The time-mean endwall heat transfer is dominated by two bands of high heat transfer, which circumscribe the leading edge of the bluff body. The band of highest heat transfer occurs in the corner region of the juncture, reflecting a 350% increase over the impinging turbulent boundary layer. A secondary high heat-transfer band develops upstream of the primary band, reflecting a 250% heat transfer increase, and is characterized by high levels of fluctuating heat load. The mean upstream position of the horseshoe vortex is coincident with a region of relatively low heat transfer that separates the two bands of high heat transfer.

2005 ◽  
Vol 128 (4) ◽  
pp. 755-762 ◽  
Author(s):  
T. J. Praisner ◽  
C. R. Smith

Time-mean endwall heat transfer and flow-field data in the endwall region are presented for a turbulent juncture flow formed with a symmetric bluff body. The experimental technique employed allowed the simultaneous recording of instantaneous particle image velocimetry flow field data, and thermochromic liquid-crystal-based endwall heat transfer data. The time-mean flow field on the symmetry plane is characterized by the presence of primary (horseshoe), secondary, tertiary, and corner vortices. On the symmetry plane the time-mean horseshoe vortex displays a bimodal vorticity distribution and a stable-focus streamline topology indicative of vortex stretching. Off the symmetry plane, the horseshoe vortex grows in scale, and ultimately experiences a bursting, or breakdown, upon experiencing an adverse pressure gradient. The time-mean endwall heat transfer is dominated by two bands of high heat transfer, which circumscribe the leading edge of the bluff body. The band of highest heat transfer occurs in the corner region of the juncture, reflecting a 350% increase over the impinging turbulent boundary layer. A secondary high heat-transfer band develops upstream of the primary band, reflecting a 250% heat transfer increase, and is characterized by high levels of fluctuating heat load. The mean upstream position of the horseshoe vortex is coincident with a region of relatively low heat transfer that separates the two bands of high heat transfer.


Author(s):  
D. R. Sabatino ◽  
C. R. Smith

The spatial-temporal flow-field and associated surface heat transfer within the leading edge, end-wall region of a bluff body were examined using both particle image velocimetry and thermochromic liquid crystal temperature measurements. The horseshoe vortex system in the end-wall region is mechanistically linked to the upstream boundary layer unsteadiness. Hairpin vortex packets, associated with turbulent boundary layer bursting behavior, amalgamate with the horseshoe vortex resulting in unsteady strengthening and streamwise motion. The horseshoe vortex unsteadiness exhibits two different natural frequencies: one associated with the transient motion of the horseshoe vortex, and the other with the transient surface heat transfer. Comparable unsteadiness occurs in the end-wall region of the more complex airfoil geometry of a linear turbine cascade. To directly compare the horseshoe vortex behavior around a turning airfoil to that of a simple bluff body, a length scale based on the maximum airfoil thickness is proposed.


Author(s):  
Shoichi Kodate ◽  
Tatsuya Kubo ◽  
Shinji Ebara ◽  
Hidetoshi Hashizume

In this study, the characteristic of the swirling flow was analyzed in detail in terms of flow field by means of a visualization experiment using matched refractive index PIV measurement to evaluate the applicability of the swirling flow generated downstream of a three-dimensionally connected dual elbow to the divertor cooling. The dual elbow used in the experiment comprises two 90-degree elbows with the same curvature connected directly in three-dimensional configuration. From the experiment, it was found that strong swirling velocity component appears locally near the pipe wall downstream of the second elbow. Moreover, although the strength of the swirling flow changed gradually as it flowed downstream, it attenuated little even 8D downstream of the dual elbow, where D was the diameter of the piping. Therefore, this swirling flow is expected to survive for a considerable distance downstream of the elbow, and the applicability of this flow field to divertor cooling can be promising. Furthermore turbulence quantities such as Reynolds stress were analyzed in terms of heat transfer performance. Since there were some regions where larger Reynolds stress than a developed turbulent pipe flow was observed near the pipe wall, high heat transfer is expected there.


Author(s):  
Tadhg S. O’Donovan ◽  
Darina B. Murray

Impinging air jets are known as a method of achieving particularly high heat transfer coefficients and are employed in many applications including the cooling of electronics, manufacturing processes such as grinding, etc. The current investigation is concerned with acoustically exciting an impinging air jet to enhance its overall cooling capacity. Distributions of the heat transfer to an axially impinging air jet for a range of Reynolds numbers (Re) from 10000 to 30000, non-dimensional nozzle to impingement surface heights (H/D) from 0.5 to 2 and excitation frequencies (f) that range from 0.5 to 1 times the natural frequency of the jet are presented. For this low range of nozzle to impingement surface spacings it has been shown that the heat transfer distribution exhibits a peak at the stagnation point and secondary peaks at a radial location that is both excitation frequency and Reynolds number dependent. Distributions of the fluctuating component of the heat transfer coefficient are also presented for the range of parameters tested. These have been used, along with spectral analysis of the heat flux signal, to discern whether local variations in heat transfer are due to changes in the local vortex flow or to changes in the mean flow structure of the impinging jet.


2010 ◽  
Vol 133 (2) ◽  
Author(s):  
V. U. Kakade ◽  
G. D. Lock ◽  
M. Wilson ◽  
J. M. Owen ◽  
J. E. Mayhew

This paper investigates heat transfer in a rotating disk system using preswirled cooling air from nozzles at high and low radius. The experiments were conducted over a range of rotational speeds, flow rates, and preswirl ratios. Narrow-band thermochromic liquid crystal (TLC) was specifically calibrated for application to experiments on a disk, rotating at ∼5000 rpm and subsequently used to measure surface temperature in a transient experiment. The TLC was viewed through the transparent polycarbonate disk using a digital video camera and strobe light synchronized to the disk frequency. The convective heat transfer coefficient h was subsequently calculated from the one-dimensional solution of Fourier's conduction equation for a semi-infinite wall. The analysis was accounted for the exponential rise in the air temperature driving the heat transfer, and for the experimental uncertainties in the measured values of h. The experimental data was supported by “flow visualization,” determined from CFD. Two heat transfer regimes were revealed for the low-radius preswirl system: a viscous regime at relatively low coolant flow rates, and an inertial regime at higher flow rates. Both regimes featured regions of high heat transfer where thin, boundary layers replaced air exiting through receiver holes at high radius on the rotating disk. The heat transfer in the high-radius preswirl system was shown to be dominated by impingement under the flow conditions tested.


Author(s):  
James E. Mayhew ◽  
James W. Baughn ◽  
Aaron R. Byerley

The film-cooling performance of a flat plate in the presence of low and high freestream turbulence is investigated using thermochromic liquid crystal thermography. Full-surface distributions of the convective heat transfer coefficient are determined for three blowing rates on a model with three straight holes spaced three diameters apart. An increase in heat transfer coefficient due to mass injection is clearly observed in the images and is quantitatively determined for both the low and high freestream turbulence cases. The increase in heat transfer coefficient is greater than in previously published research, possibly due to the use of different, more representative thermal boundary conditions upstream of the injection location. These boundary conditions, along with high resolution images, may account for the appearance of “fork tine” patterns of high heat transfer due to the presence of these vortices, not previously seen. Although the driving potential for heat transfer is less, it is observed that in some instances film cooling may cause an increase in overall heat transfer due to the increase in heat transfer coefficient.


Author(s):  
Yumin Xiao ◽  
R. S. Amano

A numerical study has been performed to predict a three-dimensional turbulent flow and end-wall heat transfer in a blade passage. The complex three-dimensional flow in the end-wall region has an important impact on the local heat transfer. The leading edge horseshoe vortex, the leading edge corner vortices, the passage vortex, and the trailing edge wake cause large variations in the entire end-wall region. The heat transfer distributions in the end-wall region are calculated and the mechanism for the high heat transfer region has been revealed. The calculations show that the algebraic turbulence model lacks the ability to predict the heat transfer in the transition region, but it is valid in other flow region. The local high heat transfer downstream of the trailing edge is enhanced by the wake downstream of the trailing edge. The horseshoe vortex results a high heat transfer region near the leading edge and induces the leading edge corner vortices which cause high heat transfer on the end-wall at both sides of blade end-wall corner.


Author(s):  
F. Montomoli ◽  
P. Adami ◽  
S. Della Gatta ◽  
F. Martelli

A reliable and accurate prediction of temperature field in hot components plays a key role in design process of modern gas turbines. The first stages of turbine and the combustor basket are usually subjected to high heat transfer rates and hot gas temperatures exceed the melting point of the employed alloys. The accurate knowledge of temperature distribution could extend the life of critical components through an accurate design of coolant systems. The present work concerns the upgrade of the finite volume CFD (Computational Fluid Dynamic) solver HybFlow, (see Adami et al.[1]) to simulate heat transfer in gas turbine cooling devices. In particular, the conjugate simulation of flow field heat transfer and metal heat conduction has been considered. To this aim, the original solver has been coupled to a routine solving the Fourier equation in solid domain. This modification allows the “conjugate heat transfer” investigation of heat transfer in fluid flow and solid domain simultaneously. The code has been validated through two different test-case applications. The first is a laminar flow over a flat plate, while the second is a film-cooled plate. Finally, a complete 3D film cooled NGV (Nozzle Guide Vane) has been investigated as an example for a more complex application. The simulation couples the thermal field inside the metal and the flow field in the vane, in the two plenum channels and in the six rows of cooling channels as well.


2008 ◽  
Vol 131 (1) ◽  
Author(s):  
D. R. Sabatino ◽  
C. R. Smith

The spatial-temporal flow field and associated surface heat transfer within the leading edge, end-wall region of a bluff body were examined using both particle image velocimetry and thermochromic liquid crystal temperature measurements. The horseshoe vortex system in the end-wall region is mechanistically linked to the upstream boundary layer unsteadiness. Hairpin vortex packets, associated with turbulent boundary layer bursting behavior, amalgamate with the horseshoe vortex resulting in unsteady strengthening and streamwise motion. The horseshoe vortex unsteadiness exhibits two different natural frequencies: one associated with the transient motion of the horseshoe vortex and the other with the transient surface heat transfer. Comparable unsteadiness occurs in the end-wall region of the more complex airfoil geometry of a linear turbine cascade. To directly compare the horseshoe vortex behavior around a turning airfoil to that of a simple bluff body, a length scale based on the maximum airfoil thickness is proposed.


Author(s):  
Jian Liu ◽  
Safeer Hussain ◽  
Lei Wang ◽  
Gongnan Xie ◽  
Bengt Sundén ◽  
...  

A pocket cavity is generated at the connection of two parts, such as the transition part between the low pressure turbine (LPT) and outlet guide vane (OGV) in a gas turbine engine. A bluff body, working as a heat transfer enhancement part or supporting strength part, has tremendous engineering applications in turbomachinery. In the present work, the effect of the pocket on the heat transfer of endwall with a bluff body in the rear part of gas turbine is investigated. A simplified triangular pocket cavity is built in a rectangular channel and two bluff bodies, a cylinder or a cuboid is attached downstream on the endwall. The heat transfer and fluid flow on the endwall are investigated experimentally and numerically. Liquid Crystal Thermography (LCT) is employed to measure the heat transfer over the pocket surface with Reynolds number ranging from 87,597, to 218,994. The turbulent flow details are provided by numerically calculations based on the commercial software Fluent 17.0. Based on the results, high heat transfer areas are usually found at the boundary of the pocket cavity and vortex street shedding regions around the bluff body. When a pocket cavity is placed in the upstream of a bluff body, the endwall heat transfer around the bluff body is obviously decreased due to the disturbance by the pocket. There are no recirculating flows in front of the tested cylinder while this is not applicable for the cuboid case. The recirculating flow behind the bluff bodies forms a three-dimensional flow structure rotating in two directions.


Sign in / Sign up

Export Citation Format

Share Document