Flow and Heat Transfer Analysis in a Single Row Narrow Impingement Channel: Comparison of Particle Image Velocimetry, Large Eddy Simulation, and RANS to Identify RANS Limitations

2017 ◽  
Vol 140 (3) ◽  
Author(s):  
Jahed Hossain ◽  
Erik Fernandez ◽  
Christian Garrett ◽  
Jayanta Kapat
Keyword(s):  
Heat Transfer ◽  
Experimental Data ◽  
Mean Flow ◽  
Single Row ◽  
Eddy Simulation ◽  
Flow And Heat Transfer ◽  
Turbulent Statistics ◽  
Image Velocimetry ◽  
Flow Phenomena ◽  
The Mean

The present study aims to understand the flow, turbulence, and heat transfer in a single row narrow impingement channel for gas turbine heat transfer applications. Since the advent of several advanced manufacturing techniques, narrow wall cooling schemes have become more practical. In this study, the Reynolds number based on jet diameter was ≃15,000, with the jet plate having fixed jet hole diameters and hole spacing. The height of the channel is three times the impingement jet diameter. The channel width is four times the jet diameter of the impingement hole. The dynamics of flow and heat transfer in a single row narrow impingement channel are experimentally and numerically investigated. Particle image velocimetry (PIV) was used to reveal the detailed information of flow phenomena. PIV measurements were taken at a plane normal to the target wall along the jet centerline. The mean velocity field and the turbulent statistics generated from the mean flow field were analyzed. The experimental data from the PIV reveal that the flow is highly anisotropic in a narrow impingement channel. To support experimental data, wall-modeled large eddy simulation (LES) and Reynolds-averaged Navier–Stokes (RANS) simulations (shear stress transport k–ω, ν2−f, and Reynolds stress model (RSM)) were performed in the same channel geometry. Mean velocities calculated from the RANS and LES were compared with the PIV data. Turbulent kinetic energy budgets were calculated from the experiment, and were compared with the LES and RSM model, highlighting the major shortcomings of RANS models to predict correct heat transfer behavior for the impingement problem. Temperature-sensitive paint (TSP) was also used to experimentally obtain a local heat transfer distribution at the target and the side walls. An attempt was made to connect the complex aerodynamic flow behavior with the results obtained from heat transfer, indicating heat transfer is a manifestation of flow phenomena. The accuracy of LES in predicting the mean flow field, turbulent statistics, and heat transfer is shown in the current work as it is validated against the experimental data through PIV and TSP.

Flow and Heat Transfer Analysis in a Single Row Narrow Impingement Channel: Comparison of PIV, LES, and RANS to Identify RANS Limitations

2017 ◽  
Author(s):  
Jahed Hossain ◽  
Erik Fernandez ◽  
Christian Garrett ◽  
Jay Kapat
Keyword(s):  
Heat Transfer ◽  
Experimental Data ◽  
Flow Field ◽  
Mean Flow ◽  
Single Row ◽  
Flow And Heat Transfer ◽  
Turbulent Statistics ◽  
Flow Phenomena ◽  
The Mean ◽  
Channel Configuration

The present study aims to understand the flow, turbulence, and heat transfer in a single row narrow impingement channel for gas turbine heat transfer applications. Since the advent of several advanced manufacturing techniques, narrow wall cooling schemes have become more practical. In this study, the Reynolds number based on jet diameter was ≃15,000, with the jet plate having fixed jet hole diameters and hole spacing. The height of the channel is 3 times the impingement jet diameter. The channel width is 4 times the jet diameter of the impingement hole. The channel configuration was chosen such that the crossflow air is drawn out in the streamwise direction (maximum crossflow configuration). The impinging jets and the wall jets play a substantial role in removing heat in this kind of configuration. Hence, it is important to understand the evolution of flow and heat transfer in a channel of this configuration. The dynamics of flow and heat transfer in a single row narrow impingement channel are experimentally and numerically investigated. Particle Image Velocimetry (PIV) was used to reveal the detailed information of flow phenomena. The detailed PIV experiment was performed on this kind of impingement channel to satisfy the need for experimental data for this kind of impingement configuration, in order to validate turbulence models. PIV measurements were taken at a plane normal to the target wall along the jet centerline. The mean velocity field and turbulent statistics generated from the mean flow field were analyzed. The experimental data from the PIV reveals that flow is highly anisotropic in a narrow impingement channel. To support experimental data, wall-modeled Large Eddy Simulation (LES), and Reynolds Averaged Navier-Stokes (RANS) simulations (SST k-ω, v2–f, and Reynolds Stress Model (RSM)) were performed in the same channel geometry. The Wall-Adapting Local Eddy-viscosity SGS mdoel (WALE) [1] is used for the LES calculation. Mean velocities calculated from the RANS and LES were compared with the PIV data. Turbulent kinetic energy budgets were calculated from the experiment, and were compared with the LES and RSM model, highlighting the major shortcomings of RANS models to predict correct heat transfer behavior for the impingement problem. Temperature Sensitive Paint (TSP) was also used to experimentally obtain a local heat transfer distribution at the target and the side walls. An attempt was made to connect the complex aerodynamic flow behavior with results obtained from heat transfer, indicating heat transfer is a manifestation of flow phenomena. The accuracy of LES in predicting the mean flow field, turbulent statistics, and heat transfer is shown in the current work as it is validated against the experimental data through PIV and TSP.

A Comparative Study of DES and URANS in a Two-Pass Internal Cooling Duct With Normal Ribs

2005 ◽  
Author(s):  
Aroon K. Viswanathan ◽  
Danesh K. Tafti
Keyword(s):  
Heat Transfer ◽  
Secondary Flows ◽  
Navier Stokes ◽  
Internal Cooling ◽  
Detached Eddy Simulation ◽  
Mean Flow ◽  
Streamline Curvature ◽  
Eddy Simulation ◽  
Flow And Heat Transfer ◽  
Developing Flow

The capabilities of the Detached Eddy Simulation (DES) and the Unsteady Reynolds Averaged Navier-Stokes (URANS) versions of the 1988 κ-ω model in predicting the turbulent flow field and the heat transfer in a two-pass internal cooling duct with normal ribs is presented. The flow is dominated by the separation and reattachment of shear layers; unsteady vorticity induced secondary flows and strong streamline curvature. The techniques are evaluated in predicting the developing flow at the entrance to the duct and downstream of the 180° bend, fully-developed regime in the first pass, and in the 180° bend. Results of mean flow quantities, secondary flows, friction and heat transfer are compared to experiments and Large-Eddy Simulations (LES). DES predicts a slower flow development than LES, while URANS predicts it much earlier than LES computations and experiments. However it is observed that as fully developed conditions are established, the capability of the base model in predicting the flow and heat transfer is enhanced by switching to the DES formulation. DES accurately predicts the flow and heat transfer both in the fully-developed region as well as the 180° bend of the duct. URANS fails to predict the secondary flows in the fully-developed region of the duct and is clearly inferior to DES in the 180° bend.

Large-Eddy Simulation With Zonal Near Wall Treatment of Flow and Heat Transfer in a Ribbed Duct for the Internal Cooling of Turbine Blades

2013 ◽  
Vol 135 (3) ◽  
Author(s):  
Sunil Patil ◽  
Danesh Tafti
Keyword(s):  
Heat Transfer ◽  
Turbine Blades ◽  
Reynolds Numbers ◽  
Large Eddy Simulations ◽  
Internal Cooling ◽  
Mean Flow ◽  
Flow And Heat Transfer ◽  
Turbulent Statistics ◽  
Large Eddy ◽  
Near Wall

Large eddy simulations of flow and heat transfer in a square ribbed duct with rib height to hydraulic diameter of 0.1 and 0.05 and rib pitch to rib height ratio of 10 and 20 are carried out with the near wall region being modeled with a zonal two layer model. A novel formulation is used for solving the turbulent boundary layer equation for the effective tangential velocity in a generalized co-ordinate system in the near wall zonal treatment. A methodology to model the heat transfer in the zonal near wall layer in the large eddy simulations (LES) framework is presented. This general approach is explained for both Dirichlet and Neumann wall boundary conditions. Reynolds numbers of 20,000 and 60,000 are investigated. Predictions with wall modeled LES are compared with the hydrodynamic and heat transfer experimental data of (Rau et al. 1998, “The Effect of Periodic Ribs on the Local Aerodynamic and Heat Transfer Performance of a Straight Cooling Channel,”ASME J. Turbomach., 120, pp. 368–375). and (Han et al. 1986, “Measurement of Heat Transfer and Pressure Drop in Rectangular Channels With Turbulence Promoters,” NASA Report No. 4015), and wall resolved LES data of Tafti (Tafti, 2004, “Evaluating the Role of Subgrid Stress Modeling in a Ribbed Duct for the Internal Cooling of Turbine Blades,” Int. J. Heat Fluid Flow 26, pp. 92–104). Friction factor, heat transfer coefficient, mean flow as well as turbulent statistics match available data closely with very good accuracy. Wall modeled LES at high Reynolds numbers as presented in this paper reduces the overall computational complexity by factors of 60–140 compared to resolved LES, without any significant loss in accuracy.

Large Eddy Simulation of Flow and Heat Transfer in a 90° Ribbed Duct With Rotation: Effect of Coriolis Forces

2004 ◽  
Author(s):  
Samer Abdel-Wahab ◽  
Danesh K. Tafti
Keyword(s):  
Heat Transfer ◽  
Large Eddy Simulation ◽  
Rotation Number ◽  
Mean Flow ◽  
Heat Transfer Augmentation ◽  
Eddy Simulation ◽  
Flow And Heat Transfer ◽  
A Value ◽  
Rotation Numbers ◽  
Large Eddy

This paper presents results from large eddy simulation (LES) of fully developed flow in a 90° ribbed duct with rib pitch-to-height ratio P/e = 10 and a rib height to hydraulic diameter ratio e/Dh = 0.1. Three rotation numbers Ro = 0.18, 0.35 and 0.67 are studied at a nominal Reynolds number based on bulk velocity of 20,000. Mean flow and turbulent quantities, together with heat transfer and friction augmentation data are presented. Turbulence and heat transfer are augmented on the trailing surface and reduced at the leading surface. The heat transfer augmentation ratio on the trailing surface asymptotes to a value of 3.7 ± 5% and does not show any further increasing trend as the rotation number increases beyond 0.2. On the other hand, augmentation ratios on the leading surface keep decreasing with an increase in rotation number with values ranging from 1.7 at Ro = 0.18 to 1.2 at Ro = 0.67. Secondary flow cells augment the heat transfer coefficient on the smooth walls by 20% to 30% over a stationary duct. An increase in rotation number from 0.35 to 0.67 decreases the frictional losses from an augmentation ratio of 9.6 to 8.75 and is a consequence of decrease in form drag and wall shear. Overall augmentation compared with a non-rotating duct ranges from +15% to +20% for heat transfer, and +10% to +15% for friction over the range of rotation numbers studied. Comparison of heat transfer augmentation with previous experimental results in the literature shows very good agreement.

Capability Assessment of Five Different RANS-Based Turbulence Models to Simulate the Various Regions of Slot Turbulent Impingement Jet Flow

2017 ◽  
Author(s):  
Mahmoud Charmiyan ◽  
Ahmed-Reza Azimian ◽  
Ebrahim Shirani ◽  
Fethi Aloui
Keyword(s):  
Experimental Data ◽  
Turbulence Models ◽  
Jet Flows ◽  
Two Dimensional ◽  
Mean Flow ◽  
Impingement Jet ◽  
Image Velocimetry ◽  
The Mean ◽  
Bottom To Top ◽  
Plate Distance

In this paper, impingement of a turbulent rectangular flow to a fixed wall is investigated. The jet flows from bottom-to-top and the output jet Reynolds is 16000. The nozzle-to-plate distance is equal to 10 (H/e = 10). Five turbulence models, including k-ε, RNG k-ε, k-ω SST, RSM and v2f model have been used for two-dimensional numerical simulation of the turbulent flow. Because of the complexities of the impingement flow, such as curved streamlines, flow separation, normal strains and sudden deceleration in different areas, different turbulence models are proposed to simulate different regions of the flow. To investigate the capability of these turbulence models in simulating different regions of the impinging jet, the mean flow velocity field and turbulent kinetic energy are extracted and compared with the experimental data of a two-dimensional particle image velocimetry (PIV). The calculated error of these five turbulence models was presented for the various flow regions, while it have not been clearly investigated earlier. Results indicate the highest conformity of the v2f model with the experimental data at the jet centerline. However, this model does not predict well the flow at the shear layer and wall-jet areas. RSM Gibson and Lander model has the highest conformity with the experimental data in these regions.

Large Eddy Simulation of Flow and Heat Transfer in a Staggered 45° Ribbed Duct

2004 ◽  
Author(s):  
Samer Abdel-Wahab ◽  
Danesh K. Tafti
Keyword(s):  
Heat Transfer ◽  
Large Eddy Simulation ◽  
Secondary Flow ◽  
Mean Flow ◽  
Cross Sectional ◽  
Heat Transfer Augmentation ◽  
Eddy Simulation ◽  
Flow And Heat Transfer ◽  
Large Eddy ◽  
Rib Vortex

Results from large eddy simulation (LES) of fully developed flow in a staggered 45° ribbed duct are presented with rib pitch-to-height ratio P/e = 10 and a rib height to hydraulic diameter ratio e/Dh = 0.1. The nominal Reynolds number based on bulk velocity is 47,300. Mean flow and turbulent quantities, together with heat transfer and friction augmentation results are presented. The flow is characterized by a helical vortex behind each rib and a complementary cross-sectional secondary flow, both of which result from the angle of the rib with respect to the mean flow. Averaged velocity profiles at the duct center show excellent agreement with experiments and heat transfer predictions agree well with experiments. Turbulent kinetic energy, shear stress, and heat transfer augmentation ratios show a strong correlation to the rib vortex and the secondary flow. Overall, heat transfer is augmented by a factor of 2.3 compared with a smooth duct and matches experimental data within 2%.

Large Eddy Simulation in a Duct With Rounded Skewed Ribs

2005 ◽  
Author(s):  
Aroon K. Viswanathan ◽  
Danesh K. Tafti
Keyword(s):  
Heat Transfer ◽  
Large Eddy Simulation ◽  
Cross Section ◽  
Mean Flow ◽  
Cross Sectional ◽  
Square Duct ◽  
Eddy Simulation ◽  
Large Eddy ◽  
The Mean ◽  
Experimental Values

Results from Large Eddy Simulation (LES) of fully developed flow in a ribbed duct are presented with rib pitch-to-height ratio (P/e) is 10 and a rib height to hydraulic diameter ratio (e/Dh) is 0.1. Computations are carried out on a square duct with 45° ribs on the top and bottom walls arranged in a staggered fashion. The ribs have a rounded cross-section and are skewed at 45° to the main flow. The Reynolds number based on bulk velocity is 25,000. Mean flow and turbulent quantities, together with heat transfer and friction augmentation results are presented for a stationary case. The flow is characterized by a helical vortex behind each rib and a complementary cross-sectional secondary flow, both of which result from the angle of the rib with respect to the mean flow and result in a spanwise variation of the heat transfer. The mean flow, the turbulent quantities and the heat transfer in the duct show similar trends as in the duct with square cross-section ribs. However the results show that there is lesser friction in the ducts with rounded ribs. The overall heat transfer on the ribbed wall was augmented by 2.85 times that of a smooth duct, at the cost of friction which increases by a factor of 10. The computed values compare well with the experimental values.

On the behaviour of impinging zero-net-mass-flux jets

2016 ◽  
Vol 810 ◽  
pp. 25-59 ◽  
Author(s):  
Carlo Salvatore Greco ◽  
Gennaro Cardone ◽  
Julio Soria
Keyword(s):  
Flow Field ◽  
Vortex Ring ◽  
Strouhal Number ◽  
Mass Flux ◽  
Mean Flow ◽  
Turbulent Statistics ◽  
Image Velocimetry ◽  
The Mean ◽  
Zero Net Mass Flux ◽  
Plate Distance

This paper reports on an experimental study of the influence of the Strouhal number (0.011, 0.022 and 0.044) and orifice-to-plate distances (2, 4 and 6 orifice diameters) on the flow field of an impinging zero-net-mass-flux jet at a Reynolds number equal to 35 000. These jets are generated by a reciprocating piston that oscillates in a cavity behind a circular orifice. Instantaneous two-dimensional in-plane velocity fields are measured in a plane containing the orifice axis using multigrid/multipass cross-correlation digital particle image velocimetry. These measurements have been used to investigate the mean flow quantities and turbulent statistics of the impinging zero-net-mass-flux jets. In addition, the vortex ring behaviour is analysed via its trajectory and azimuthal vorticity as well as the saddle point excursion, the flow rate and entrainment. The behaviour of all these quantities depends on the Strouhal number and the orifice-to-plate distance because the former governs the presence and the relative importance of the vortex ring and the trailing jet on the flow field and the latter delimits the downstream evolution of these structures.

Large-Eddy Simulation With Zonal Near Wall Treatment of Flow and Heat Transfer in a Ribbed Duct for the Internal Cooling of Turbine Blades

2011 ◽  
Author(s):  
Sunil Patil ◽  
Danesh Tafti
Keyword(s):  
Heat Transfer ◽  
Boundary Layer Equation ◽  
Reynolds Numbers ◽  
Wall Region ◽  
Mean Flow ◽  
Flow And Heat Transfer ◽  
Turbulent Statistics ◽  
Wall Boundary Conditions ◽  
Large Eddy ◽  
Near Wall

Large eddy simulations of flow and heat transfer in a square ribbed duct with rib height to hydraulic diameter of 0.1 and 0.05 and rib pitch to rib height ratio of 10 and 20 are carried out with the near wall region being modeled with a zonal two layer model. A novel formulation is used for solving the turbulent boundary layer equation for the effective tangential velocity in a generalized co-ordinate system in the near wall zonal treatment. A methodology to model the heat transfer in the zonal near wall layer in the LES framework is presented. This general approach is explained for both Dirichlet and Neumann wall boundary conditions. Reynolds numbers of 20,000 and 60,000 are investigated. Predictions with wall modeled LES are compared with the hydrodynamic and heat transfer experimental data of Rau et al. [1], and Han et al. [2], and wall resolved LES data of Tafti [3]. Friction factor, heat transfer coefficient, mean flow as well as turbulent statistics match available data closely with very good accuracy. Wall modeled LES at high Reynolds numbers as presented in this paper reduces the overall computational complexity by factors of 60–140 compared to resolved LES, without any significant loss in accuracy.

Flow Visualization Experiment of a Swirling Flow Formed Downstream of a Piping With Successive Three Elbow to be Applied to Divertor Cooling

2014 ◽  
Author(s):  
Shoichi Kodate ◽  
Shinji Ebara ◽  
Hidetoshi Hashizume
Keyword(s):  
Heat Transfer ◽  
Swirling Flow ◽  
Pipe Wall ◽  
Mean Flow ◽  
Transfer Enhancement ◽  
Cooling Systems ◽  
Turbulent Statistics ◽  
Visualization Experiment ◽  
The Mean ◽  
3D Layout

As one of potential divertor cooling systems, three-dimensionally connected elbow piping has been proposed. In this study, a visualization experiment was conducted for swirling flows generated downstream of five kinds of piping with successive three three-dimensionally connected elbows in order to evaluate the applicability of these systems to divertor cooling by comparing with the dual elbow case previously obtained in terms of strength of the swirling flow and turbulent statistics. From the experimental results, it was found that the triple elbow piping in which all elbows were connected three–dimensionally, referred to as 3D+3D layout, generated strong swirling velocity components than those of the dual elbow case and it became up to 70 % of the mean flow velocity. Moreover, it did not attenuate even 5D downstream of the triple elbow, where D was the diameter of the piping and the applicability of this flow field to divertor cooling can be promising. In addition, when heat transfer was evaluated in terms of turbulent statistics in the 3D+3D layout, heat transfer enhancement is expected because larger turbulent kinetic energy was observed near the pipe wall than a straight piping case.

Sign in / Sign up

Export Citation Format

Share Document