Abstract

The flow physics of impinging jet is very complex and is not fully understood yet. The flow field in an impingement problem comprised of three different distinct regions: a free jet with a potential core, a stagnation region where the velocity goes to zero as the jet impinges onto the wall and a creation of wall jet region where the boundary layer grows radially outward after impinging. Since impingement itself is a broad topic, effort is being made in the current study to narrow down on three particular geometric configurations (a narrow wall, an array impingement configuration and a curved surface impingement configuration) that shows up in a typical gas turbine impingement problem in relation to heat transfer. Impingement problems are difficult to simulate numerically using conventional RANS models. It is worth noting that the typical RANS model contains a number of calibrated constants and these have been formulated with respect to relatively simple shear flows. As a result typically these isotropic eddy viscosity models fail in predicting the correct heat transfer value and trend in impingement problem where the flow is highly anisotropic. The common RANS-based models over predict stagnation heat transfer coefficients by as much as 300% when compared to measured values. Even the best of the models, the v^2-f model, can be inaccurate by up to 30%. Even though there is myriad number of experimental and numerical work published on single jet impingement; the knowledge gathered from these works cannot be applied to real engineering impingement cooling application as the dynamics of flow changes completely. This study underlines the lack of experimental flow physics data in published literature on multiple jet impingement and the author emphasized how important it is to have experimental data to validate CFD tools and to determine the suitability of Large Eddy Simulation (LES) in industrial application. In the open literature there is not enough study where experimental heat transfer and flow physics data are combined to explain the behavior for gas turbine impingement cooling application. Often it is hard to understand the heat transfer behavior due to lack of time accurate flow physics data hence a lot of conjecture has been made to explain the phenomena. The problem is further exacerbated for array of impingement jets where the flow is much more complex than a single round jet. The experimental flow field obtained from Particle Image Velocimetry (PIV) and heat transfer data obtained from Temperature Sensitive Paint (TSP) from this work will be analyzed to understand the relationship between flow characteristics and heat transfer for the three types of novel geometry mentioned above. There has not been any effort made on implementing LES technique on array impingement problem in the published literature. Nowadays with growing computational power and resources CFD are widely used as a design tool. To support the data gathered from the experiment, LES is carried out in narrow wall impingement cooling configuration. The results will provide more accurate information on impingement flow physics phenomena where experimental techniques are limited and the typical RANS models yield erroneous result The objective of the current study is to provide a better understanding of impingement heat transfer in relation to flow physics associated with it. As heat transfer is basically a manifestation of the flow and most of the flow in real engineering applications is turbulent, it is very important to understand the dynamics of flow physics in an impingement problem. The work emphasis the importance of understanding mean velocities, turbulence, jet shear layer instability and its importance in heat transfer application. The present work shows detailed information of flow phenomena using Particle Image Velocimetry (PIV) in a single row narrow impingement channel. Results from the RANS and LES simulations are compared with Particle Image Velocimetry (PIV) data. The accuracy of LES in predicting the flow field and heat transfer of an impingement problem is also presented the in the current work as it is validated against experimental flow field measured through PIV. Results obtained from the PIV and LES shows excellent agreement for predicting both heat transfer and flow physics data. Some of the key findings from the study highlight the shortcomings of the typical RANS models used for the impingement heat transfer problem. It was found that the stagnation point heat transfer was over predicted by as much as 48% from RANS simulations when compared to the experimental data. A lot of conjecture has been made in the past for RANS' ability to predict the stagnation point heat transfer correctly. The length of the potential core for the first jet was found to be ~ 2D in RANS simulations as oppose to 1D in PIV and LES, confirm the possible underlying reason for this discrepancy. The jet shear layer thickness was underpredicted by ~ 40% in RANS simulations proving the model is not diffusive enough for a flow like jet impingement. Turbulence production due to shear stress was over predicted by ~130% and turbulence production due to normal stresses were underpredicted by ~40 % in RANS simulation very close to the target wall showing RANS models fail where both strain rate and shear stress plays a pivotal role in the dynamics of the flow. In the closing, turbulence is still one of the most difficult problems to solve accurately, as has been the case for about a century. A quote below from the famous mathematician, Horace Lamb (1849-1934) express the level of difficulty and frustration associated with understanding turbulence in fluid mechanics. "I am an old man now, and when I die and go to heaven there are two matters on which I hope for enlightenment. One is quantum electrodynamics, and the other is the turbulent motion of fluids. And about the former I am rather optimistic." Source: http://scienceworld.wolfram.com/biography/Lamb.html This dissertation is expected to shed some light onto one specific example of turbulent flows.

Graduation Date

2016

Semester

Fall

Advisor

Kapat, Jayanta

Degree

Doctor of Philosophy (Ph.D.)

College

College of Engineering and Computer Science

Department

Mechanical and Aerospace Engineering

Degree Program

Mechanical Engineering

Format

application/pdf

Identifier

CFE0006463

URL

http://purl.fcla.edu/fcla/etd/CFE0006463

Language

English

Release Date

December 2017

Length of Campus-only Access

1 year

Access Status

Doctoral Dissertation (Campus-only Access)

Share

COinS