Advances in Heat Transfer 52 1St Edition J P Abraham Full Chapter

Advances in heat transfer 52 1st Edition
J.P. Abraham
Visit to download the full and correct content document:
https://ebookmass.com/product/advances-in-heat-transfer-52-1st-edition-j-p-abraham/
Academic Press is an imprint of Elsevier
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
525 B Street, Suite 1650, San Diego, CA 92101, United States
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
125 London Wall, London, EC2Y 5AS, United Kingdom
First edition 2020
Copyright © 2020 Elsevier Inc. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by any means, electronic
or mechanical, including photocopying, recording, or any information storage and retrieval system,
without permission in writing from the publisher. Details on how to seek permission, further
information about the Publisher’s permissions policies and our arrangements with organizations such
as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website:
www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the
Publisher (other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience
broaden our understanding, changes in research methods, professional practices, or medical
treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating
and using any information, methods, compounds, or experiments described herein. In using such
information or methods they should be mindful of their own safety and the safety of others, including
parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume
any liability for any injury and/or damage to persons or property as a matter of products liability,
negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas
contained in the material herein.
ISBN: 978-0-12-820737-6
ISSN: 0065-2717
For information on all Academic Press publications

visit our website at https://www.elsevier.com/books-and-journals
Publisher: Zoe Kruze

Acquisitions Editor: Ashlie Jackman
Editorial Project Manager: Chris Hockaday
Production Project Manager: Abdulla Sait
Cover Designer: Christian Bilbow
Typeset by SPi Global, India
List of contributors
J.P. Abraham
University of St. Thomas, School of Engineering, St. Paul, MN, United States
A. Andreozzi
Dipartimento di Ingegneria Industriale, Università degli Studi di Napoli Federico II, Napoli,
Italy
Filippo de Monte
Department of Industrial and Information Engineering and Economics, University of
L’Aquila, L’Aquila, AQ, Italy
Zhipeng Duan
School of Mechanical, Electronic and Control Engineering, Beijing Jiaotong University,
Beijing, China
J.M. Gorman
Department of Mechanical Engineering, University of Minnesota, Minneapolis, MN,
United States
P. Alex Greaney
Department of Mechanical Engineering, University of California Riverside—Riverside,
Riverside, CA, United States
A. Haji-Sheikh
Department of Mechanical and Aerospace Engineering, University of Texas at Arlington,
Arlington, TX, United States
Jackson R. Harter
Reactor Physics Analysis, Idaho National Laboratory, Idaho Falls, ID, United States
M. Iasiello
Dipartimento di Ingegneria Industriale, Università degli Studi di Napoli Federico II, Napoli,
Italy
Erfan Kosari
Mechanical Engineering Department, University of California, Riverside, CA, United States
Hao Ma
School of Mechanical, Electronic and Control Engineering, Beijing Jiaotong University,
Beijing, China
W.J. Minkowycz
Department of Mechanical and Industrial Engineering, University of Illinois at Chicago,
Chicago, IL, United States
Todd S. Palmer
School of Nuclear Science and Engineering, Oregon State University, Corvallis, OR,
United States
ix
x List of contributors
Matthew Regnier
Department of Mechanical Engineering, University of Minnesota, Minneapolis, MN,
United States
C. Tucci
Dipartimento di Medicina Traslazionale, Università degli Studi del Molise, Campobasso,
Italy
Kambiz Vafai
Mechanical Engineering Department, University of California, Riverside, CA, United States
Tie Wei
Department of Mechanical Engineering, New Mexico Institute of Mining and Technology,
Socorro, NM, United States
CHAPTER ONE
Analyses of buoyancy-driven
convection
Tie Wei*
Department of Mechanical Engineering, New Mexico Institute of Mining and Technology, Socorro,
NM, United States
*Corresponding author: e-mail address: Tie.Wei@nmt.edu
Contents
1. Introduction 2
2. Prediction of Nusselt number 6
2.1 Lessons from forced convection 12
3. Governing equations 14
3.1 Reynolds averaged equations for the mean flow and heat transport 15
3.2 Laminar DHVC solution 18
4. Dimensional analysis of buoyancy-driven convection 20
4.1 Dimensional analysis of laminar DHVC 21
4.2 Dimensional analysis of turbulent DHVC 23
5. Review of scaling patch approach 27
5.1 Layer structure of turbulent channel flow 31
5.2 Steps in scaling patch approach 33
6. Scaling analysis of laminar DHVC 44
7. Scaling analysis of the mean momentum equation in turbulent DHVC 48
7.1 Layer structure of the mean momentum balance equation 48
7.2 Properties of the Reynolds shear stress 51
7.3 Outer scaling of the mean momentum equation 53
7.4 Inner scaling of the mean momentum equation 59
7.5 Meso scaling of the mean momentum equation 62
8. Scaling analysis of the mean heat equation 67
8.1 Layer structure of the mean heat equation 68
8.2 Properties of the turbulent temperature flux Rwθ 70
8.3 Outer scaling of the mean heat equation 71
8.4 Inner scaling of the mean heat equation 75
8.5 Scaling patches in the mean heat equation 78
9. New prediction of Nusselt number 82
10. Summary and conclusions 86
Acknowledgments 89
References 89
Advances in Heat Transfer, Volume 52 # 2020 Elsevier Inc. 1

ISSN 0065-2717 All rights reserved.
https://doi.org/10.1016/bs.aiht.2020.09.002
2 Tie Wei
Abstract
This article investigates the multilayer structure of turbulent flow and heat transport in
buoyancy-driven convection, and in particular, introduces a relatively new scaling patch
approach. A differentially heated vertical channel (DHVC) is used as an example of
buoyancy-driven convection, and its multilayer structure is first qualitatively investi-
gated by dimensional analysis. In the near-wall region of turbulent DHVC, flow and heat
transport is strongly influenced by the molecular diffusion, and the kinematic viscosity
and thermal diffusivity are important control parameters in the dimensional analysis.
Flow and heat transport in the inner layer is controlled by two nondimensional num-
bers: the Prandtl number of the fluid and an inner Richardson number. Away from the
wall, flow and heat transport are dominated by eddy motions, largely independent of
molecular diffusion. The controlling nondimensional parameter in the outer layer of tur-
bulent DHVC is an outer Richardson number. The multilayer structure in turbulent DHVC
is then elucidated quantitatively by the scaling patch analysis. Based on the character-
istics of force balance, the mean momentum equation is divided into three layers: an
inner layer, a meso layer, and an outer layer. The inner and outer Richardson numbers,
derived from the dimensional analysis, appear naturally in the properly scaled mean
momentum equation. Another nondimensional number that appears naturally from
the scaling patch analysis is the friction Reynolds number. The characteristic length scale
in the inner layer is directly influenced by the friction Reynolds number, distinctively
different from that in forced convection. The characteristic length scale in the meso
layer is an Obukhov-style length scale. The mean heat equation can also be divided into
multiple layers. In fact, an inherent hierarchy of layer structure (scaling patches) is rev-
ealed through a simple transformation of the turbulent temperature flux. A new predic-
tion of the Nusselt number is developed based on the insight gained from the
dimensional analysis and scaling patch analysis. The new prediction is directly con-
nected to the multilayer structure of heat transport in turbulent DHVC and is fundamen-
tally different from the traditional power-law correlations.
1. Introduction
A cornerstone in the study of turbulence is the recognition that the
dynamics of turbulent flow and scalar transport operate on a great many
space and time scales (see, e.g., Monin and Yaglom [1], Tennekes and
Lumley [2]). A better understanding of the multiscale structure of turbulence
is critical in improving our predictive capabilities of turbulence, i.e., predic-
tion of skin friction and heat transport rate. Two powerful tools to uncover
the multiscale nature of turbulence are dimensional analysis and scaling anal-
ysis. This article applies these tools to elucidate the multilayer structure of
turbulent buoyancy-driven convection. In particular, one aim of this article
is to introduce a relatively new scaling approach to buoyancy-driven con-
vection. The new approach, based on the “search of scaling patches,” was
Analyses of buoyancy-driven convection 3
originally developed for forced wall-bounded turbulence such as shear-

driven turbulent flow over a flat plate or pressure-driven turbulent pipe
and channel flow, in a series of papers by Fife, Klewicki, McMurtry,
Wei, and coworkers [3–17]. Some concepts and ideas in the scaling patch
approach are similar to previous scaling approaches, however, the logical
trains of thought in the new approach are distinctly different.
Generally speaking, a fluid expands when being heated, resulting in a
density decrease. Density inhomogeneity can also be generated from the
mixing of fluids with different densities. In the presence of a gravitational
field, there is a net force that pushes upward a light fluid surrounded by a
heavier fluid, and this upward force is called the buoyancy force. The buoy-
ancy force gives rise to the ascending of the lighter fluid and descending of
the heavier fluid. The bulk fluid motion induced by the buoyancy force is
commonly called buoyancy-driven convection. If the fluid motion is driven
solely by buoyancy, the convection is also called natural convection or free
convection. If the fluid motion is driven by both buoyancy and shear or
pressure, the convection is called mixed convection.
Buoyancy-driven convections are encountered in a variety of natural
phenomena and industry applications. Examples include atmospheric and
oceanic convection, space heating and cooling, smoke and fire spreading,
nuclear reactor containment, and solar collectors. There have been an enor-
mous amount of studies about the effects of buoyancy on flow and heat (or
mass) transport. It is impossible to give in one article an exhaustive review of
buoyancy-driven flow and heat transport, which can be found in books
devoted to the subject, e.g. by Turner [18], Jaluria [19], Gebhart et al.
[20], Kakac et al. [21], Martynenko and Khramtsov[22], and Verman [23].
How buoyancy affects the flow and heat or mass transport is of important
practical interest, and is also of great theoretical interest. To this day, there
are still no reliable tools for predicting the heat or mass transfer coefficient in
buoyancy-driven turbulence [24]. One of the earliest studies of turbulent
flow affected by buoyancy is about the convective atmospheric boundary
layer (ABL) by Taylor [25] and Schmidt [26]. A key quantity in the under-
standing and prediction of stratified ABL is the vertical transport of momen-
tum, water vapor, sensible heat, or heat in latent form [27]. One of the
landmarks in the development of our understanding of atmospheric turbu-
lence was the formation of the concept of the Austausch coefficient by
Schmidt and Taylor. The idea was built on the knowledge of pressure-
or shear-driven wall turbulence developed by Prandtl et al. [27]. While rea-
sonable for shear dominated turbulence, the concept of eddy viscosity based
4 Tie Wei
on local properties becomes meaningless when the buoyancy effect is

strong, e.g., in convective atmospheric boundary layer [28]. A better under-
standing of the underlying physics in buoyancy-driven turbulence is critical
in developing more robust models.
To understand the essential physics of buoyancy-driven turbulence,
convection with simple geometry is typically used in physical experiments
or numerical simulations. Two buoyancy-driven convections with simple
geometry are illustrated in Fig 1: differentially heated vertical channel
(DHVC) and Rayleigh-Benard convection (RBC). In both cases, flow
and heat transport occurs between two parallel plates, which are maintained
at different temperatures. In DHVC, the temperature gradient is perpendic-
ular to the gravity, but in RBC, the temperature gradient is aligned with the
gravity. In DHVC, hotter fluid ascends along the hot plate side, and at the
same time, colder fluid descends along the cold plate side. In turbulent
RBC, a prominent feature is the rising up of hot fluid as plumes and falling
down of cold fluid as inverted plumes.
RBC has been extensively studied in physical laboratories for more than
one hundred years. During the past thirty years, numerical simulation, espe-
cially direct numerical simulation (DNS), has become an important tool in
the study of turbulent flows, including buoyancy-driven convection, RBC
Fig. 1 Canonical configurations of buoyancy-driven convection. (A) Differentially

heated vertical channel (DHVC). The hotter fluid rises up on the left side, and the colder
fluid descends on the right side. (B) Rayleigh–Benard convection. The mean flow is zero
def
and the mean temperature is antisymmetric about the mid-plane. Θ ¼ T hot T is the
mean transformed temperature.
and DHVC. Analysis of turbulent quantities, especially those involved

derivatives, are often only feasible with DNS data, due to the accuracy
and resolution requirement.
In the interests of simplicity and clarity, buoyancy-driven DHVC is cho-
sen as an example to introduce the newly developed scaling patch approach,
but the approach can be readily adapted to other buoyancy-driven con-
vection. DHVC is an interesting buoyancy-driven convection to test tools
and concepts developed originally for forced convection. In turbulent
DHVC, flow is driven by buoyancy, but the turbulent kinetic energy is
produced by two mechanisms: shear-generation and buoyancy-generation
[29]. From a practical point of view, a better understanding of DHVC can
help our prediction of heat transfer from a hot or cold vertical wall, as in the
heating or cooling of building spaces. Last, DHVC is selected because there
have been direct numerical simulation (DNS) data from three independent
studies by Versteegh [30], Kiš [31], and Ng [32]. As in the analysis of forced
wall turbulence [3], high-quality DNS data are essential in the application
and evaluation of the scaling patch approach.
More details on DHVC were given in dissertations by Versteegh [30],
Kiš [31], and Ng [32]. Here, only a brief review of previous studies on tur-
bulent DHVC is provided. One of the earliest studies of buoyancy-driven
turbulence next to a vertical plate was by George and Capp [33]. They used
classical scaling arguments and proposed a three-layer structure of buoyancy-
driven turbulence: an inner layer adjacent to the solid wall, an outer layer
away from the wall, and a buoyant sublayer in between. Applying an asymp-
totic matching approach, they proposed a power-law variation of the mean
velocity profile and mean temperature profile in the buoyant sublayer. Their
scaling analysis also leads to an explicit relationship between the Nusselt
number and the Rayleigh number.
Some of the early numerical simulations of buoyancy-driven convec-
tion in a vertical slot were by Phillips [34] and Boudjemadi et al. [35]. Due
to the limit of computing power, early DNS typically used a small computa-
tional domain. The first benchmark DNS study of DHVC was by Versteegh
[30,36], who used a large domain of 24 12 2 (Lx Ly Lz). Using the
simulation data, they investigated the scaling behavior of the mean temper-
ature, the mean velocity profile, and the profiles of various turbulence
statistics.
Kiš [31,37] performed direct numerical simulations of DHVC over a
wider range of Rayleigh number on a large computation domain of 24
12 2. The effect of domain size, spatial resolution, time-averaging period,
6 Tie Wei
and error analysis was examined in detail. Budgets for Reynolds stress and
turbulent heat flux were documented, and entropy generation was also
investigated.
Ng and coworkers [32,38–41] have performed the highest Rayleigh
number simulation of DHVC so far. They used the simulation data to
explore the scaling of the flow and heat transport, the structures of the flow
field, and the spectral properties of the turbulence.
The rest of the article is organized as follows. In Section 2, the current
strategy in the prediction of Nusselt number is reviewed, and its shortcom-
ing is pointed out. In Section 3, governing equations for flow and heat
transport are presented. In Section 4, dimensional analysis is applied to both
the laminar DHVC and turbulent DHVC. In Section 5, the scaling patch
approach is reviewed and demonstrated using a pressure-driven turbulent
channel flow. In Section 6, scaling analysis is applied to laminar DHVC.
In Section 7, the scaling patch analysis is applied to the mean momentum
balance equation in turbulent DHVC. In Section 8, the scaling patch analysis
is applied to the mean heat equation in turbulent DHVC. In Section 9, a
new prediction of Nusselt number is proposed. Section 10 summarizes
the article.
2. Prediction of Nusselt number

A centerpiece in many studies of convection, forced or buoyancy-
driven, is the prediction of heat transport rate. The prediction of flow
and heat transport often starts with a dimensional analysis [42]. The first step
in dimensional analysis is to identify parameters that affect the quantity being
predicted. In buoyancy-driven convection, DHVC, or RBC, the conven-
tionally used control parameters are listed in Table 1.
There are a total of nine control parameters, and three primary dimen-
sions are entailed: length L, time t, and temperature T. Following the
Buckingham-Pi theorem [43], flow and heat transport in DHVC or
RBC are governed by six nondimensional numbers. For example,
selecting ΔT, ν, and Lz (the domain size in the direction of temperature
gradient) as the repeating variables, the following six nondimensional
numbers are found accordingly
ν
Π1 ¼ Πα ¼ ¼ Pr; (1a)
α
Π2 ¼ Πβ ¼ β ΔT ; (1b)
Table 1 Parameters that affect flow and heat transport in buoyancy-driven convection.
Material ν Kinematic viscosity [L2/t]

properties
α Thermal diffusivity [L2/t]
β Thermal expansion coefficient [1/T]
!
Force g Gravitational acceleration [L/t2]
Thermal BC ΔT Temperature difference [T]
Geometry ∡ !
Angle between the temperature gradient and g [1]
Lx, Ly, Lz Domain size [L]
The last column lists the primary dimensions of the parameters: length L, time t, and temperature T.
g L 3z
Π3 ¼ Πg ¼ ; (1c)
ν2
Π4 ¼ Π∡ ¼ ∡; (1d)
L
Π5 ¼ Πx ¼ x ¼ Γ x ; (1e)
Lz
Ly
Π6 ¼ Πy ¼ ¼ Γy , (1f)
Lz
where Π1 is the ratio between the kinematic viscosity and thermal diffusiv-
ity, called the Prandtl number. For the ideal configurations of DHVC and
RBC within infinite parallel plates, the aspect ratios are Γx ≫ 1 and Γy ≫ 1.
The angle between the temperature gradient and the gravitational accel-
eration is ∡ ¼ π=2 for DHVC and ∡ ¼ 0 for RBC. Thus, three non-
dimensional control parameters remain for DHVC or RBC: Π1 ¼ Pr,
Π2 ¼ β ΔT, and Π3 ¼ g L 3z =ν2 .
In traditional analyses of buoyancy-driven convection, the non-
dimensional numbers Π2 and Π3 are combined to form a Grashof number
(the justification is given in Section 4). In this article, the Grashof number
is defined using the channel half-width δ and the temperature difference
between the wall and channel mid-plane Θmp ¼ 0.5ΔT as
def gβΘmp δ3 1 gβΔT ð2δÞ

3
Gr ¼ ¼ : (2)
ν2 16 ν2
The reason for using δ and Θmp is that the multilayer structures of flow and
heat transport in DHVC or RBC are symmetrical about the channel
8 Tie Wei
mid-plane. Such a defined Grashof number is compatible with the friction

Reynolds number used in pressure-driven wall-bounded turbulence
def
through a pipe or channel Reτ ¼ δuτ =ν, where δ is the channel half-width
or pipe radius and uτ is the friction velocity [3]. The traditional Grashof
number is defined using the channel whole width Lz ¼ 2δ and the tem-
perature difference across the whole channel ΔT ¼ 2Θmp. Thus, the pres-
ently defined Grashof number is 1/16 of the traditionally defined Grashof
number.
Traditionally, Grashof number is interpreted as a ratio between the
buoyancy force and the viscous force [44]. However, Grashof number
can also be related to the ratio of length scales by rearranging as follows
δ
Gr1=3 ¼ : (3)
ðν2 =gβΘ mp Þ
1=3
Thus, one-third power of Grashof number can be interpreted as the ratio of

the channel half-width to a “viscous length scale” defined as (ν2/gβΘmp)1/3.
In the study of buoyancy-driven convection, a popular nondimensional
number is the Rayleigh number defined as
def gβΘmp δ3
Ra ¼ ¼ Gr Pr: (4)
να
One of the most important quantities in the study of convection is the heat
transport rate. A commonly used nondimensional number for heat transport
efficiency is the Nusselt number, which is defined as the ratio between the
convective transport rate and conductive transport rate:

dΘ
def qt
Nu ¼ Θ ¼ Θ dz wall
, (5)
mp mp
α
δ δ
def
where qt ¼ Qwall =ρref c p ¼ αdΘ=dzjwall is the temperature flux at the wall,
Qwall is the wall heat flux, ρref is the fluid density, and cp is the heat capacity.
Mathematically, Nusselt number can be interpreted as a normalized second
kind boundary condition [44], i.e., the temperature gradient at the wall nor-
malized by a temperature scale Θmp and a length scale δ.
Accurate prediction of the Nusselt number is of great importance for
engineering applications. Based on the dimensional analysis, it is obvious
that Nusselt number in DHVC or RBC is a function of the Prandtl number
of the fluid and the Rayleigh number:
Nu ¼ f ðPr, RaÞ: (6)
Historically, power-law is often used to correlate the Nusselt number with

the Rayleigh number and Prandtl number as
Nu C Prn Ram : (7)
While experimental measurements of Nusselt number for DHVC are scant,

abundant experimental measurements of Nusselt number are available for
RBC. Hence, we will use the Nusselt number prediction in RBC to illus-
trate the current strategy of predicting the Nusselt number for buoyancy-
driven convection. Fig. 2 presents the Nusselt number data for RBC over
a wide range of Rayleigh numbers. Numerous correlations have been pro-
posed for the Nusselt number in RBC, and Table 2 lists just a few of them.
Nearly all the correlations used a power-law for the Rayleigh number
dependence as Ram. However, a long-standing, still heated debated, issue
in the study of RBC is the value of the exponent m.
In the 1950s, Malkus [45] analyzed turbulent RBC in terms of marginal
stability of the mean flow and derived an exponent of m ¼ 1/3. The work
was further developed by Howard [46]. In the 1960s, Kraichnan [47] refined
Fig. 2 Nusselt number vs Rayleigh number in Rayleigh–B enard convection. Inset shows
the trend at ultra-high Rayleigh numbers. Data of helium (U) are from experimental
measurement Urban et al. [50]. Data of helium (N) are from the experimental measure-
ment Niemela and Sreenivasan [51]. Data of water and silicone oil AK3 are from exper-
imental measurement of Silveston [52].
10 Tie Wei
Table 2 Examples of correlations for Nusselt number in RBC.

Year Researchers Fluids Correlations
1959 Globe and Dropkin [53] Any fluids Nu ¼ 0.069 Ra0.33 Pr0.074
1969 Rossby [54] Mercury Nu ¼ 0.147 Ra0.257
1975 Threlfall [55] Helium Nu ¼ 0.173 Ra0.28
1989 Castaing et al. [48] Helium Nu ¼ 0.23 Ra0.282
1996 Takeshita et al. [56] Mercury Nu ¼ 0.155 Ra0.27
The applicable Rayleigh number ranges of the correlations are referred to the original papers.
More discussions on the correlations can be found in Ahlers et al. [57].
the similarity theory to include a double boundary layer. One interesting

result of his analysis is that at ultra-high Rayleigh number (say Ra ≳ 1018 ),
the exponent m will be larger than 1/3, because of the interaction of the
boundary layer with a horizontal fluctuating wind. Based on cryogenic
helium gas data over a wide range of Rayleigh numbers, Castaing et al.
[48] observed that the exponent is smaller than m ¼ 1/3. Moreover, they
developed a new scaling theory and suggested an exponent of m ¼ 2/7.
A more recent, and currently very popular model of RBC has been developed
by Grossmann and Lohse [49]. The model is built on a theoretical analysis of
the dissipation rate in both the boundary layer and the well-mixed core layer.
Depending on the ratio of the dissipation rate, the model provides a set of scal-
ing relations between the Nusselt number and Rayleigh number, Prandtl
number, and aspect ratio of the RBC cell.
The difference in the exponent, for example, m ¼ 1/3 vs m ¼ 2/7, can
lead to significant difference in the prediction of Nusselt number, up to a
factor of 10 at Ra 1025, as shown in the inset of Fig. 2. To better diagnose
the existence of a power law, it is common to plot the so-called compensated
Nusselt number as Nu Ram vs the Rayleigh number, as shown in Fig. 3.
The compensated Nusselt number with an exponent m ¼ 2/7 is shown in
Fig. 3A, and a constant region is observed for traditional Rayleigh numbers
between 105 and 1010. However, this compensated Nusselt number mono-
tonically increases for Rayleigh number larger than 1010. For the exponent of
m ¼ 1/3, the constant region of the compensated Nusselt number is at higher
Rayleigh numbers between 1010 and 1014, as shown in Fig. 3B. However,
the trend at ultra-high Rayleigh number (> 1014) is not clear presently.
Nevertheless, Fig. 3 shows that the dependence of Nusselt number on the
Rayleigh number in RBC can be approximated by a power-law, but over
A B
Fig. 3 Compensated Nusselt number vs Rayleigh number in RBC. (A) Nu Ra2/7 vs Ra.
(B) Nu Ra1/3 vs Ra.
A B
Fig. 4 Nusselt number in DHVC. (A) Nusselt number vs Rayleigh number.

(B) Compensated Nusselt number vs Rayleigh number.
a limited range of Rayleigh number, and the exponent varies with the
Rayleigh number. The nonconstancy of the exponent m in the power-
law prediction of Nusselt number is not just a nuisance in engineering appli-
cation, but, more importantly, it reflects an inherent drawback of the
power-law prediction: it lacks a connection to the underlying physics in
buoyancy-driven convection, i.e., the multilayer structure of turbulent flow
and heat transport.
For DHVC, the Nusselt number data are mainly from direct numerical
simulations. Fig. 4A presents the Nusselt number data vs Rayleigh number in
the turbulent DHVC regime. Like other buoyancy-driven convection,
power-law correlations are also popular in predicting the Nusselt number
in DHVC, as listed in Table 3. Compared with the RBC case, the Rayleigh
number range of DHVC data is much more limited. Curve-fitting within a
limited range of Rayleigh numbers can almost always produce a power-law
12 Tie Wei
Table 3 Correlations for Nusselt number in DHVC.

Year Researchers Correlations
1979 George and Capp Nu ¼ 0.059(Pr Ra)1/3
1998 Versteegh and Nieuwstadt Nu ¼ 0.071(Pr Ra)1/3
Ra1=3
2007 Balaji, Holling and Herwig Nu ¼
½0:101 lnðRaÞ+6:30634=3
2012 Kiš and Herwig Nu ¼ 0.186(Pr Ra)1/3.2

2014 Ng, Chuang and Ooi Nu ¼ 0.046(Pr Ra)1/3
correlation, as shown in Fig. 4A. The deviation between the data and
the power-law function seems small in Fig. 4A. However, when the com-
pensated Nusselt number is plotted, a systematic deviation is noticeable
in Fig. 4B as the Rayleigh increases, similar to the RBC case shown
in Fig. 3.
One goal of this article is to advocate that it is time to move beyond the
power-law prediction of Nusselt number in buoyancy-driven convection
using Eq. (7). In forced convection, the prediction of Nusselt number with
power-law correlation has long been superseded by more accurate equations
that were built on the underlying physics of heat transfer.
2.1 Lessons from forced convection

Power-law correlations with Reynolds number and Prandtl number were
popular in predicting Nusselt number in early studies of forced convection.
For example, for heat transfer (treated as passive scalar) through a turbulent
pipe flow, one of the earliest correlations is the well-known Dittus–Boelter
equation developed in the 1930s (see, e.g., [58–62])
Nu 0:023 Pr0:4 Re0:8

b , (8)
where Reb is the bulk Reynolds number defined by the diameter of the pipe
and the bulk mean velocity. The Dittus–Boelter equation can be found in
many textbooks on heat transfer, e.g., [44]. Now, it is generally agreed that
such power-law prediction is not accurate over a wide range of Reynolds
numbers, in particular at high Reynolds numbers. More accurate predictions
that are commonly used nowadays are the Petukhov equation [63] and the
Gnielinski equation [64]:
f
ðReb 1000ÞPr
Nu 8 rffiffiffiffi ; Petukhov equation (9a)
f 2=3
1 + 12:7 ðPr 1Þ
8
" #
23 0:11
f
ðReb 1000Þ Pr d Pr
Nu 8 rffiffiffi 1+ , Gnielinski equation
f L Prw
1 + 12:7 ðPr 1Þ
2=3
8
(9b)
where f is the friction coefficient. The Reynolds number and Prandtl num-
ber ranges for the validity of these equations can be found in standard heat
transfer textbooks, e.g., Ref. [44].
Recently, Wei [65] and Wei and Abraham [66] showed that the Kader–
Yaglom style equation [67] gives an excellent prediction of Nusselt number,
over a wide range of Prandtl number and Reynolds numbers. The Kader–
Yaglom style equation is presented as (see [65,66])
2Pr Reτ 2PrReτ
Nu ≡ , (10)
Θmix
+ 1
lnðPr Reτ Þ + Bθ,m
κθ
where Reτ is the friction Reynolds number defined using the channel half-
width or pipe radius, Θ+mix is the mixed temperature scaled by the friction
temperature.
It is known that the mean temperature in forced wall-bounded turbulence
can be robustly approximated by a logarithmic function, except in the near-
wall region and a small, and bounded, deviation in the core of the pipe [1].
The thermal “log-law” is analogous to the well-known “log-law” for the
mean velocity that was found in the 1930s (see, e.g., Monin and Yaglom
[1]). More importantly, the log-law approximation for the mean velocity dis-
tribution has been reliably and robustly observed in the high Reynolds num-
ber experiments such as Princeton Superpipe experiments [68–70].
The logarithmic function in the Kader–Yaglom style Eq. (10) is directly
related to the “log-layer” for the mean temperature distribution. The func-
tion Bθ,m represents the temperature increments in the thermal diffusion
sublayer, and the deviation in the core (outer layer) [65,66]. The effect of
the Prandtl number in the near-wall region is also accounted for in Bθ,m.
Thus, the Kader–Yaglom style equation is directly built on the multilayer
structure of mean
14 Tie Wei
In Section 9, a new prediction of the Nusselt number in buoyancy-

driven convection, similar to the Kader–Yaglom style equation, is devel-
oped. The new prediction equation is also directly built on the multilayer
structure of mean temperature in buoyancy-driven turbulence.
3. Governing equations
For most buoyancy-driven convection, the velocity of bulk motion is
sufficiently small, and the governing equations can be approximated by the
Boussinesq–Oberbeck equations [1]. In the Boussinesq–Oberbeck approx-
imation, the fluid is assumed to be incompressible, and the density variation
is negligible except in the buoyancy term [1]. The governing equations for
the instantaneous flow and heat transfer in the Cartesian coordinate systems
can be written as follows [1]
∂e
uj
¼ 0; (11a)
∂xj

∂e
ui ∂ðe
uj uei Þ ∂2 uei ∂ pe
+ ¼ν + gi ; (11b)
∂t ∂xj ∂xj ∂x j ∂xi ρref
∂θe ∂ðe e
uj θÞ ∂2 θe
+ ¼α : (11c)
∂t ∂xj ∂xj ∂xj
In this article, the instantaneous flow or heat variable is denoted by a tilde.
For example, uei is the instantaneous velocity in the i direction, and θe is the
instantaneous modified temperature defined as the difference between the
mean wall temperature Twall and the instantaneous temperature T e : θe def
¼
e
T wall T . gi is the component of the body force in the i direction.
Eq. (11a) is the continuity equation, Eqs. (11b) are the well-known
Navier-Stokes equation for the momentums in three directions, and Eq.
(11c) is the energy or heat equation. One fluid property, the kinematic vis-
cosity ν, appears in the momentum equations (11b), and another fluid prop-
erty, thermal diffusivity α appears in the heat equation (11c). In this article, the
material properties are assumed constant, and the variability of viscosity, ther-
mal diffusivity, or non-Boussinesq effects are not considered [71–74].
The coordinate system is set up that z is in the wall-normal direction. For
wall-bounded flow and heat transport, no-slip boundary conditions are
applied for the instantaneous velocity and temperature as uei jz¼0 ¼ 0 and
e
θjz¼0 ¼ 0 [1].
3.1 Reynolds averaged equations for the mean flow and heat
transport
General solutions to the instantaneous governing Eq. (11) lie beyond the
scope of existing methods. Moreover, in practical applications, the knowl-
edge of the instantaneous flow field is too cumbersome and is often not nec-
essary. Instead, we are mainly interested in the mean flow and scalar transport
properties, for example, the mean skin friction, the mean heat transfer rate,
and sometimes the average of the fluctuation level. Reynolds averaging pro-
vides an extremely common framework for studying the mean flow and sca-
lar transport [1]. It is a shortcut and erases a wealth of details about the fluid
motions being studied.
Conceptually, Reynolds averaging is ensemble averaging over a large
number of realizations of the flows under nominally the same boundary con-
ditions and initial conditions [1]. The Reynolds averaging decomposes an
instantaneous flow or heat variable into a mean component and a fluctuation
component. For example, the instantaneous velocity is decomposed as
uei ¼ U i + ui : (12)
Here, an upper case letter denotes a mean flow or heat variable, and a lower
case letter denotes its fluctuation. Similarly the instantaneous temperature is
decomposed as θe ¼ Θ + θ, where Θ is the mean temperature and θ is the
temperature fluctuation.
In an experimental or numerical study of statistically steady turbulence, a
more practical averaging is obtained by time averaging. Moreover, for tur-
bulence possessing statistically spatial homogeneity, the averaging is also
applied spatially over the homogeneous plane. For example, in turbulent
DHVC, mean flow and temperature vary only in the wall-normal z direction,
and are homogeneous in the vertical x–y plane. Therefore, the experimental
or simulation data of turbulent DHVC are both time-averaged and spatial
averaged in the x–y plane. The averaged governing equations for the mean
heat transport and mean flow in turbulent DHVC are as follows [30,75]
d2 Θ dRwθ
0¼α + ; (13a)
dz2 dz
d2 U dRwu
0¼ν 2 + + gβðΘmp ΘÞ, (13b)
dz dz
def
where Rwθ ¼ hwθi is the turbulent temperature flux in the vertical direc-
def
tion. Rwu ¼ hwui is the turbulent flux of the streamwise momentum in
the vertical direction, commonly called the kinematic Reynolds shear stress.
16 Tie Wei
Angle brackets h i denote the conceptual ensemble averaging. β is the ther-

mal expansion coefficient, and Θmp is the mean temperature at the channel
mid-plane.
The Reynolds shear stress Rwu arises from the averaging of the nonlinear
inertial term in the Navier-Stokes equations, representing a force produced
by turbulent fluctuations. The turbulent temperature flux Rwθ arises from the
averaging of the nonlinear advective term in the instantaneous heat equation.
There are four unknowns Θ, U, Rwu, and Rwθ in two Eqs. (13a) and (13b).
This is the well-known “closure” problem in RANS modeling. Being
under-determined, the problem can yield no unique solutions by them-
selves. This conflicts with the usual image of asymptotic methods, which
provide a sequence of more and more accurate approximations to a unique
exact solution [5]. The scaling patch approach tacked this basic ill-posed
problem by adding reasonable and minimal assumptions about the qualitative
nature of the solutions [5].
The mean temperature Θ in Eq. (13a) and the mean velocity U in Eq.
(13b) are second-order, and Rwu and Rwθ are first-order. Thus, two bound-
ary conditions are required for U and Θ, and one boundary condition is
required for Rwu and Rwθ. The commonly used boundary conditions in
DHVC are
z¼0: U ¼ 0, Θ ¼ 0, Rwu ¼ 0, Rwθ ¼ 0: (14a)

z¼δ: U ¼ 0, Θ ¼ Θmp : (14b)
The boundary conditions in Eq. (14a) are the no-slip condition at the solid
surface. Note that all these boundary conditions are the first kind boundary
condition [44]. However, a second kind boundary condition can also be
applied at the solid surface as

dU q
¼ u; (15a)
dz z¼0 ν

dΘ qt
¼ : (15b)
dz z¼0 α
def
Here qu ¼ τwall =ρref denotes the kinematic momentum flux at the wall, where
def
τwall is the wall shear stress and ρref is the fluid density. qt ¼ Qwall =ρref c p denotes
the temperature flux at the wall. In Sections 4 and 7, qu and qt are shown to play
a key role in the multiscaling analysis of the mean equations using the dimen-
sional analysis approach and the scaling patch approach.
In the study of forced wall-bounded turbulence, two extremely impor-

tant wall variables are the friction velocity and friction temperature (see
Monin and Yaglom [1] and Tennekes and Lumley [2]) defined by the wall
shear stress and wall heat flux as
rffiffiffiffiffi
def τw pffiffiffiffi
uτ ¼ ¼ qu ; (16a)
ρ
def Q w q
θτ ¼ ¼ t: (16b)
ρref c p uτ uτ
Integrating the mean heat Eq. (13a) and the mean momentum Eq. (13b) in
the wall-normal z direction and applying boundary conditions produce a
relation for the total temperature flux and total momentum flux as [75]
dΘ
+ Rwθ ¼ uτ θτ ;
α (17a)
dz Z z
dU
ν + Rwu ¼ uτ gβ ðΘmp ΘÞdz:
2
(17b)
dz 0
The total temperature flux consists of the molecular diffusion (first term on
the left of Eq. (17a)) and the turbulent transport (second term). The physical
meaning of Eq. (17a) is that the total temperature flux (or heat flux when
multiplied by ρref c p) in a fully developed turbulent DHVC is a constant across
the channel, equal to the wall temperature (or heat) flux. This constancy of
total temperature flux is analogous to the constancy of total momentum flux
in turbulent plane Couette flow [76].
The total momentum flux also consists of two components: a viscous
transport (first term on the left of Eq. (17b)) and a turbulent transport (second
term). The physical meaning of Eq. (17b) is that the total momentum flux
equals to the wall momentum flux minus a spatial integral related to the tem-
perature deficit. The variation of total momentum flux in DHVC is distinc-
tively different from that in forced convection [3]. For example, in turbulent
plane Couette flow, the total momentum flux is a constant, equal to the wall
momentum flux [76]. In turbulent DHVC, the total momentum flux is
directly related to the buoyancy effect, which plays a fundamental role in
the shape and distribution of Reynolds shear stress, as shown in Section 7.
Next, we present the analytical solutions of laminar DHVC. The analytic
solutions themselves are trivial, but they provide a simple example to eval-
uate dimensional analysis and scaling patch approach.
18 Tie Wei
3.2 Laminar DHVC solution

The governing equation for laminar DHVC can be obtained by setting tur-
bulence terms to zero in the mean heat Eq. (13a) and the mean momentum
Eq. (13b). Integrating along the wall-normal z direction and applying
boundary conditions produce the analytical solution for the temperature
and velocity distribution in laminar DHVC as
z
Θ ¼ Θmp ; (18a)
δ

gβΘmp δ2 1 z 1 z2 1 z3
U¼ + : (18b)
ν 3δ 2 δ 6 δ
The analytical solutions for the temperature and velocity distributions in
laminar DHVC are shown in Fig. 5A. The temperature Θ varies linearly
in the wall-normal direction, like a pure conduction distribution. The ver-
tical velocity exhibits a maximum value near the hotter plate, then decreases
in the core of the channel, and exhibits a minimum near the colder plate.
The vertical velocity is antisymmetric about the channel mid-plane. For
comparison, the mean velocity and mean temperature distributions in
A B
Fig. 5 Temperature and vertical velocity profiles in a DHVC. (A) Laminar case.
(B) Turbulent cases. Data are from the DNS of Kiš [31] at two Grashof numbers: Gr ¼
4.75 104 (solid curves) and Gr ¼ 4.4 105 (dashed curves).
turbulent DHVC are presented in Fig. 5B at two Grashof numbers. In tur-

bulent DHVC, the mean temperature varies sharply in the near-wall region,
and the variation of temperature in the core is much smaller. On the flow
side, the variation of the mean velocity is also much sharper in the near-wall
region, and the peak location of the maximum velocity moves closer to the
wall with increasing Grashof number.
The maximum vertical velocity location in laminar DHVC can be found
by setting the velocity gradient to zero. The maximum location and value
are found as
zU max ¼ ð1 31=2 Þδ 0:423 δ; (19a)

gβqt δ3 gβΘmp δ2
U max 0:064 ¼ 0:064 : (19b)
να ν
In buoyancy-driven convection, a commonly used velocity scale is the
so-called free-fall velocity defined as
def pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
U ff ¼ gβΘmp δ: (20)
Hence, the maximum vertical velocity in laminar DHVC is related the free-
fall velocity as
δ U ff U max δ U ff
U max 0:064 U ff , or 0:064 : (21)
ν U ff ν
Thus, the so-called “free-fall” velocity is not a proper or “natural” scale for
the vertical velocity in laminar DHVC, because Umax/Uff is not a constant,
but increases as δUff/ν. In Section 4, it is shown that this “free-fall” velocity
is not a proper scale for the vertical velocity in turbulent DHVC either.
Although free-fall velocity is easy to compute from the inputs of the
buoyancy-driven convection and is commonly used to normalize equations
or numerical simulation data, such scaled variables are not appropriate and
should be avoided in the scaling of buoyancy-driven convection.
Based on Eq. (19b), a better velocity scale for laminar DHVC is defined as
def gβqt δ3
U c,lam ¼ : (22)
να
As shown in Section 4, the wall temperature and momentum fluxes qt and qu
are critical in understanding buoyancy-driven convection. For laminar
DHVC, the wall temperature flux and the wall momentum flux can be
obtained from the analytical solutions as
20 Tie Wei
Θmp
dΘ
qt ¼ α ¼α ; (23a)
dz z¼0 δ
gβqt δ2
dU
qu ¼ ν ¼ : (23b)
dz z¼0 3α
Eq. (23a) indicates that the Nusselt number in laminar DHVC is Nu ¼ 1. In
other words, in laminar DHVC, there is bulk fluid motion in the vertical
direction, but the flow direction is along the isothermal line, perpendicular
to the temperature gradient. Thus the bulk fluid motion does not advect
heat, and the heat transport from the hot plate to the cold plate is by pure
molecular diffusion.
From Eq. (23b), the friction velocity uτ in laminar DHVC is
rffiffiffiffiffiffiffiffiffiffiffiffi
gβqt δ2
uτ ¼ : (24)
3α
The friction Reynolds number for laminar DHVC can be obtained as
pffiffiffiffiffiffi
def δu Gr
Reτ ¼ τ ¼ pffiffiffi : (25)
ν 3
The ratio between the Umax and uτ in laminar DHVC depends on the
Grashof number and the friction Reynolds number as
U max pffiffiffiffiffiffi
0:11 Gr 0:19Reτ : (26)
uτ
4. Dimensional analysis of buoyancy-driven convection

Dimensional analysis is a powerful tool in the study of fluid dynamics
and heat transfer (see, e.g., Buckingham [43], Bridgman [77], Taylor [78],
Townsend [79], Sedov [80], Barenblatt [81]). According to Churchill [60],
the fundamental basis for dimensional analysis was established by Fourier in
1822. Rayleigh [82] demonstrated the power of dimensional analysis in a
short paper published in the Nature magazine in 1915, giving examples from
various fields. A fascinating example illustrating the power of dimensional
analysis is the prediction of the first atomic explosion yield by G. I.
Taylor [78], using a series of pictures published in a magazine.
We will first present a dimensional analysis of laminar DHVC, followed
by turbulent DHVC. In the dimensional analysis of laminar DHVC, we pre-
sent two options of selecting control parameters and discuss the common-
ality and difference.
Table 4 Dimensional analysis of laminar DHVC.

Option 1 Option 2
Control parameters gβ, ν, α, δ, Θmp gβ, ν, α, δ, qt, qu
Repeating variables gβ, ν, δ gβ, ν, δ
Nondim. parameters Π1 ¼ Πα ¼ αν ¼ Pr Π1 ¼ Πα ¼ αν ¼ Pr
gβΘmp δ3 gβqt δ4
Π2 ¼ ΠΘmp ¼ ν2 ¼ Gr Π2 ¼ Πqt ¼ ν3 ¼ Gr
Pr
qu δ2
Π3 ¼ Πqu ¼ ν2
Length scale lc ¼ δ Ψ(Pr, Gr) lc ¼ δ Ψ(Pr, Gr)

Velocity scale uc ¼ νδ ΨðPr, GrÞ uc ¼ νδ ΨðPr, GrÞ
Temperature scale θc ¼ ΘmpΨ(Pr, Gr) θc ¼ ΘmpΨ(Pr, Gr)
The generic nondimensional functions Ψ in the last three rows are, in general, different from each other,
and cannot be determined by the dimensional analysis itself.
4.1 Dimensional analysis of laminar DHVC

Proper selection of control parameters is of utmost importance in a suc-
cessful dimensional analysis of any flow or heat transport problem. For
laminar DHVC, two sets of control parameters can be selected for dimen-
sional analysis, as listed in Table 4. In both options, only three primary
dimensions are entailed: length, time, and temperature. However, there
are five control parameters in option 1, and six in option 2. Hence, based
on the Buckingham Pi theorem, option 1 produces two nondimensional
control parameters, and option 2 produces three.
The common control parameters in option 1 and option 2 are gβ, ν, α, δ,
which arise from the momentum and heat equation, and the domain size
(part of the boundary condition). The parameter g and β are grouped
together based on the observation of Eq. (13b): the effect of gravity on
the buoyancy force is always coupled with the thermal expansion coeffi-
cient. There is nowhere in the governing equations that g and β appear sep-
arately. Therefore, gβ should be treated as a single parameter, instead of as
two parameters as in Table 1.
The difference between option 1 and option 2 in Table 4 lies in the
parameter(s) related to heat. In option 1, Θmp is used. Note that Θmp can
be interpreted as the temperature difference between the wall and the chan-
nel mid-plane. In other words, Θmp is the difference of the first kind thermal
boundary conditions at z ¼ 0 and z ¼ δ. More importantly, Θmp in general
does not reflect any local rate of temperature change, at the wall, or near the
22 Tie Wei
channel mid-plane. In contrast, the two control parameters qt and qu in

option 2 represent the local rate of change at the wall (a second kind bound-
ary condition).
Applying the standard dimensional analysis procedure [43], the two con-
trol parameters from option 1 in Table 4 are the Prandtl number of the fluid
and the Grashof number. For example, based on the dimensional analysis
results, the maximum vertical velocity in laminar DHVC can be presented as
U max
¼ ΨðPr, GrÞ, (27)
ν=δ
where Ψ denotes a generic nondimensional function that depends on Pr and
Gr, which has to be determined from analytical solutions or experimental
and numerical data.
The three nondimensional parameters resulting from option 2 in Table 4
are Pr, Π2 ¼ gβqtδ4/ν3, and Π3 ¼ quδ2/ν2. However, in laminar DHVC, the
boundary conditions of qu and qt are not independent, but directly related to
each other as shown in Eq. (23b):
qu δ2 Pr gβqt δ4
¼ : (28)
ν2 3 ν2 α
Thus, the third nondimensional number Π3 from option 2 in Table 4 is
redundant, and both options in Table 4 produce identical nondimensional
control parameters.
An advantage of option 2 in Table 4 is that it leads directly to the def-
inition of a flux Grashof number. The flux Grashof number is defined using
the wall temperature flux qt as (e.g., Bejan [42])
def gβqt δ4
Gr ¼ ¼ Gr Nu: (29)
f
ν2 α
Similarly, a flux Rayleigh number can be defined as
def gβqt δ4
Ra f ¼ ¼ Pr Gr f : (30)
να2
Note that this flux Rayleigh number is similar to the H number defined by
George and Capp [33] for natural convection along a vertical plate under
constant heat flux. For laminar DHVC, Nu ¼ 1 and the flux Grashof num-
ber equals the traditional Grashof number Gr ¼ Grf and Ra ¼ Raf.
The flux Grashof number is related to the ratio of two length scales, as
showing in the following form
1=4 δ
Gr f ¼ (31)
ðν2 α=gβqt Þ1=4
Thus, one-fourth power of the flux Grashof number can be interpreted as
the ratio between the channel half-width δ and a “viscous length scale”
defined as (ν2α/gβqt)1/4. Such defined viscous length scale is similar to the
inner length scale defined by Townsend [79] for buoyancy-driven convec-
tion over a single horizontal plane or Rayleigh–Benard convection.
To recap, laminar DHVC is controlled by two nondimensional num-
bers. Two options are valid for the selection of control parameters. The dif-
ference between the two options lies in the selection of parameter(s) with a
local rate of temperature change or a global temperature difference. In lam-
inar DHVC, the temperature varies linearly, hence, the global temperature
difference Θmp is directly related to the local rate of temperature change qt.
Thus, the two options in Table 4 are equivalent. However, in turbulent
DHVC, the mean temperature variation is not linear, and the selection of
Θmp becomes inappropriate.
4.2 Dimensional analysis of turbulent DHVC

In pressure- or shear-driven wall-bounded turbulence, it is well known that
the flow and heat transport can be divided into several layers, and the char-
acteristic scales may vary with respect to given layers [1,2]. For example, tur-
bulent flow over a flat plate is traditionally divided into four layers: a viscous
sublayer, a buffer layer, a log-layer, and an outer layer (see, e.g., Tennekes
and Lumley [2]). In the viscous sublayer, the characteristic length scale is the
viscous length scale ν/uτ. However, in the outer layer, the characteristic
length scale is the boundary layer thickness.
In turbulent DHVC, flow and heat transport can also be divided into
several layers, and the control parameters in each layer may be different.
For example, adjacent to the wall, the velocity and temperature profiles have
large gradients (see Fig. 5B). As a result, the molecular diffusion is important,
and ν and α should be selected as the control parameters in the near-wall
region. On the other hand, in the core of the channel, flow and heat trans-
port are dominated by eddy motions, and molecular diffusion is negligible.
For example, in the core of the channel in Fig. 5B, the mean temperature
variation is small, meaning that the molecular temperature flux αdΘ/dz is
small. Therefore, the kinematic viscosity ν and thermal diffusivity α should
not be selected as control parameters for the flow and heat transport in the
core of the channel. At a sufficiently high Grashof number, the separation
24 Tie Wei
between the inner layer and the outer layer becomes large, and a meso layer
emerges between the inner and outer layers.
Based on physical reasoning [83], a turbulent DHVC is divided into an
inner layer, a meso layer, and an outer layer, as illustrated in Fig. 6. The con-
trol parameters for each layer are shown in Fig. 6 and listed in Table 5.
Fig. 6 Inner, meso, and outer layer structure of turbulent DHVC, and the control param-
eters in each layer [83].
Table 5 Summary of dimensional analysis for the inner layer, meso layer, and outer
layers in turbulent DHVC.
Inner layer Meso layer Outer layer
Control parameters gβ, ν, α, qu, qt. gβ, qu, qt. gβ, qu, qt, δ.
Repeating variables gβ, ν, qu. gβ, qu, qt. gβ, qu, δ.
gβqt ν gβqt δ
Nondim. parameters Πqt ¼ q2u ¼ Rii Πqt ¼ 3=2 ¼ Rio
qu
Πα ¼ αν ¼ Pr
Length scale l i ¼ uντ ΨðPr, Rii Þ L ¼ gβq
qu
3=2
lo ¼ δΨ(Rio)
t
Velocity scale ui ¼ uτΨ(Pr, Rii) uo ¼ uτΨ(Rio)

Temperature scale θi ¼ θτΨ(Pr, Rii) θo ¼ θτΨ(Rio)
The generic nondimensional functions Ψ in the last three rows are, in general, different from each other
and must be determined from experimental or numerical data.
It is obvious that Θmp cannot be an appropriate control parameter in the

inner layer, because Θmp is not a local quantity in the near-wall region. In
some previous studies of turbulent DHVC, only qt was selected as a control
parameter, but not qu. Here, both qt and qu are used. There are five control
parameters for the inner layer: gβ, qu, qt, ν, α, and the standard dimensional
analysis following Buckingham’s Pi theorem produces two nondimensional
controlling parameters: one is the Prandtl number of the fluid, and the other
one is called the inner Richardson number defined as [83]
def gβqt ν gβθτ ν Gr Nu Gr f

Rii ¼ 2 ¼ 3 ¼ ¼ : (32)
qu uτ Pr Reτ Pr Re4τ
4
In the outer layer of highly turbulent flow and highly turbulent heat transfer
DHVC, flow and heat transport is dominated by eddy motions, and the
effect of molecular diffusion is negligible. As the channel width is a geomet-
ric constraint on the size of the eddies, channel half-width δ is chosen as a
control parameter in the outer layer. So there are four control parameters for
the outer layer gβ, qu, qt, δ, and dimensional analysis produces one non-
dimensional number, called the outer Richardson number defined as [83]
def gβqt δ gβθτ δ Gr Nu Gr f

Rio ¼ ¼ 2 ¼ ¼ ¼ Rii Reτ : (33)
qu
3=2 uτ Pr Reτ Pr Re3τ
3
For a turbulent DHVC at a sufficiently high Rayleigh number, there is a

meso layer that is far away from the wall and the molecular diffusion does
not affect the flow or heat transport. At the same time, the meso layer is also
far away from the outer layer and is not affected by the channel half-width δ
either. Thus, the meso layer is controlled by three parameters gβ, qu, qt,
which leads to the definition of an Obukhov-style length scale [83,84]
q3=2 u2
¼ τ :
def
L ¼ u
(34)
gβqt gβθτ
Therefore, the inner Richardson number can be interpreted as the ratio

between an inner length scale and the Obukhov-style length scale:
ν=uτ ν=uτ
Rii ¼ ¼ , (35)
u2τ =gβθτ L
and the outer Richardson number can also be interpreted as a length scale
ratio: the channel half-width to the Obukhov length scale
26 Tie Wei
δ δ
Rio ¼ ¼ : (36)
u2τ =gβθτ L
Fundamentally, flow and heat transport in turbulent DHVC is controlled by

two nondimensional parameters. One is the Prandtl number of the fluid,
and the other can be any one of the following: Gr, Grf, Ra, Raf, Rii,
Rio, or Reτ. The nondimensional numbers that can be used in the study
of buoyancy-driven convection in DHVC are summarized in Table 6.
These nondimensional parameters are not independent of each other. For
example, if Gr and Pr are known, then Rii or Rio or Reτ each has a unique
value. If Rio and Pr are known, then each of the Gr, Rii or Reτ also has a
unique value. Fig. 7 shows the relationship among these nondimensional
numbers. Over the range of DNS data, these nondimensional numbers
can be approximated by simple power-laws.
The solid curve at low Grashof numbers in Fig. 7 represents the analytical
pffiffiffiffiffiffi
Eq. (26), U max =uτ 0:11 Gr, for laminar DHVC. The gray vertical band
indicate the transitional Grashof numbers, based on the stability analysis of
Versteegh [30]. Also included in the figure is the data of the maximum ver-
tical velocity scaled by the friction velocity Umax/uτ, because it is a conve-
nient parameter in the scaling analysis of the mean equations as shown in
Sections 7 and 8.
Table 6 Nondimensional numbers that can be used in the study of DHVC.

Definition
Grashof number def
Gr ¼ gβΘmp δ3 =ν2
Flux Grashof number Gr
def
¼ gβqt δ4 =ðν2 αÞ ¼ Gr Nu
f
Rayleigh number def

Ra ¼ gβΘmp δ3 =ðναÞ
Flux Rayleigh number Ra
def
¼ gβΘmp δ4 =ðνα2 Þ ¼ Ra Nu
f
Inner Richardson number def

Rii ¼ gβqt ν=q2u ¼ gβθτ ν=u3τ ¼ Gr Nu=ðPr Re4τ Þ
Outer Richardson number def
Rio ¼ gβqt δ=q3=2
u ¼ gβθ τ δ=uτ ¼ Gr Nu=ðPr Reτ Þ
2 3
Friction Reynolds number def

Reτ ¼ δuτ =ν
Friction Peclet number def
Peτ ¼ δuτ =α ¼ Pr Reτ
Another random document with
no related content on Scribd:
singular, the said Frederick Baily alias
Douglass unto the said Hugh Auld, his
executors and administrators, and against all and
every other person or persons whatsoever, shall
and will warrant and forever defend by these
presents. In witness whereof, I set my hand and
seal, this thirteenth day of November, eighteen
hundred and forty-six (1846.)
Thomas Auld.
“Signed, sealed, and delivered, in presence of
Wrightson Jones, John C. Lear.”
The authenticity of this bill of sale is attested
by N. Harrington, a justice of the peace of the
state of Maryland, and for the county of Talbot,
dated same day as above.
* * * * *
“To all whom it may concern: Be it known that
I, Hugh Auld, of the city of Baltimore, in Baltimore
county in the state of Maryland, for divers good
causes and considerations, me thereunto
moving, have released from slavery, liberated,
manumitted, and set free, and by these presents
do hereby release from slavery, liberate,
manumit, and set free, my negro man, named
Frederick Baily, otherwise called Douglass,
being of the age of twenty-eight years, or
thereabouts, and able to work and gain a
sufficient livelihood and maintenance; and him
the said negro man, named Frederick
Douglass, I do declare to be henceforth free,
manumitted, and discharged from all manner of
servitude to me, my executors and
administrators forever.
“In witness whereof, I the said Hugh Auld,
have hereunto set my hand and seal the fifth of
December, in the year one thousand eight
hundred and forty-six.
Hugh Auld.
“Sealed and delivered in presence of T.
Hanson Belt, James N. S. T. Wright.”
Having remained abroad nearly two years, and being about to

return to America, not as I left it—a slave, but a freeman, prominent
friends of the cause of emancipation intimated their intention to make
me a testimonial both on grounds of personal regard to me, and also
to the cause to which they were so ardently devoted. How such a
project would have succeeded I do not know, but many reasons led
me to prefer that my friends should simply give me the means of
obtaining a printing press and materials, to enable me to start a
paper, advocating the interests of my enslaved and oppressed
people. I told them that perhaps the greatest hindrance to the
adoption of abolition principles by the people of the United States,
was the low estimate everywhere in that country placed upon the
negro as a man: that because of his assumed natural inferiority,
people reconciled themselves to his enslavement and oppression, as
being inevitable if not desirable. The grand thing to be done,
therefore, was to change this estimation, by disproving his inferiority
and demonstrating his capacity for a more exalted civilization than
slavery and prejudice had assigned him. In my judgment a tolerably
well conducted press in the hands of persons of the despised race,
would by calling out and making them acquainted with their own
latent powers, by enkindling their hope of a future, and developing
their moral force, prove a most powerful means of removing
prejudice and awakening an interest in them. At that time there was
not a single newspaper regularly published by the colored people in
the country, though many attempts had been made to establish
such, and had from one cause or another failed. These views I laid
before my friends. The result was, that nearly two thousand five
hundred dollars were speedily raised towards my establishing such a
paper as I had indicated. For this prompt and generous assistance,
rendered upon my bare suggestion, without any personal effort on
my part, I shall never cease to feel deeply grateful, and the thought
of fulfilling the expectations of the dear friends who had given me
this evidence of their confidence, was an abiding inspiration for
persevering exertion.
Proposing to leave England, and turning my face toward America
in the spring of 1847, I was painfully reminded of the kind of life
which awaited me on my arrival. For the first time in the many
months spent abroad, I was met with proscription on account of my
color. While in London I had purchased a ticket, and secured a berth,
for returning home in the Cambria—the steamer in which I had come
from thence—and paid therefor the round sum of forty pounds,
nineteen shillings sterling. This was first cabin fare; but on going on
board I found that the Liverpool agent had ordered my berth to be
given to another, and forbidden my entering the saloon. It was rather
hard after having enjoyed for so long a time equal social privileges,
after dining with persons of great literary, social, political, and
religious eminence, and never, during the whole time, having met
with a single word, look, or gesture, which gave me the slightest
reason to think my color was an offense to anybody—now to be
cooped up in the stern of the Cambria, and denied the right to enter
the saloon, lest my presence should disturb some democratic fellow-
passenger. The reader can easily imagine what must have been my
feelings under such an indignity.
This contemptible conduct met with stern rebuke from the British
press. The London Times, and other leading journals throughout the
United Kingdom, held up the outrage to unmitigated condemnation.
So good an opportunity for calling out British sentiment on the
subject had not before occurred, and it was fully embraced. The
result was that Mr. Cunard came out in a letter expressive of his
regret, and promising that the like indignity should never occur again
on his steamers, which promise I believe has been faithfully kept.
CHAPTER VII.
TRIUMPHS AND TRIALS.
New Experiences—Painful Disagreement of Opinion with old Friends—Final

Decision to Publish my Paper in Rochester—Its Fortunes and its Friends
—Change in my own Views Regarding the Constitution of the United
States—Fidelity to Conviction—Loss of Old Friends—Support of New
Ones—Loss of House, etc., by Fire—Triumphs and Trials—Underground
Railroad—Incidents.
PREPARED as I was to meet with many trials and perplexities on

reaching home, one of which I little dreamed was awaiting me. My
plans for future usefulness, as indicated in the last chapter, were all
settled, and in imagination I already saw myself wielding my pen as
well as my voice in the great work of renovating the public mind, and
building up a public sentiment, which should send slavery to the
grave, and restore to “liberty and the pursuit of happiness” the
people with whom I had suffered.
My friends in Boston had been informed of what I was intending,
and I expected to find them favorably disposed toward my cherished
enterprise. In this I was mistaken. They had many reasons against it.
First, no such paper was needed; secondly, it would interfere with my
usefulness as a lecturer; thirdly, I was better fitted to speak than to
write; fourthly, the paper could not succeed. This opposition from a
quarter so highly esteemed, and to which I had been accustomed to
look for advice and direction, caused me not only to hesitate, but
inclined me to abandon the undertaking. All previous attempts to
establish such a journal having failed, I feared lest I should but add
another to the list, and thus contribute another proof of the mental
deficiencies of my race. Very much that was said to me in respect to
my imperfect literary attainments, I felt to be most painfully true. The
unsuccessful projectors of all former attempts had been my
superiors in point of education, and if they had failed how could I
hope for success? Yet I did hope for success, and persisted in the
undertaking, encouraged by my English friends to go forward.
I can easily pardon those who saw in my persistence an
unwarrantable ambition and presumption. I was but nine years from
slavery. In many phases of mental experience I was but nine years
old. That one under such circumstances should aspire to establish a
printing press, surrounded by an educated people, might well be
considered unpractical if not ambitious. My American friends looked
at me with astonishment. “A wood-sawyer” offering himself to the
public as an editor! A slave, brought up in the depths of ignorance,
assuming to instruct the highly civilized people of the north in the
principles of liberty, justice, and humanity! The thing looked absurd.
Nevertheless I persevered. I felt that the want of education, great as
it was, could be overcome by study, and that wisdom would come by
experience; and further (which was perhaps the most controlling
consideration) I thought that an intelligent public, knowing my early
history, would easily pardon the many deficiencies which I well knew
that my paper must exhibit. The most distressing part of it all was the
offense which I saw I must give my friends of the old Anti-Slavery
organization, by what seemed to them a reckless disregard of their
opinion and advice. I am not sure that I was not under the influence
of something like a slavish adoration of these good people, and I
labored hard to convince them that my way of thinking about the
matter was the right one, but without success.
From motives of peace, instead of issuing my paper in Boston,
among New England friends, I went to Rochester, N. Y., among
strangers, where the local circulation of my paper—“The North
Star”—would not interfere with that of the Liberator or the Anti-
Slavery Standard, for I was then a faithful disciple of Wm. Lloyd
Garrison, and fully committed to his doctrine touching the pro-slavery
character of the Constitution of the United States, also the non-
voting principle of which he was the known and distinguished
advocate. With him, I held it to be the first duty of the non-
slaveholding States to dissolve the union with the slaveholding
States, and hence my cry, like his, was “No union with slaveholders.”
With these views I came into western New York, and during the first
four years of my labors here I advocated them with pen and tongue,
to the best of my ability. After a time, a careful reconsideration of the
subject convinced me that there was no necessity for dissolving the
“union between the northern and southern States;” that to seek this
dissolution was no part of my duty as an abolitionist; that to abstain
from voting was to refuse to exercise a legitimate and powerful
means for abolishing slavery; and that the Constitution of the United
States not only contained no guarantees in favor of slavery, but on
the contrary, was in its letter and spirit an anti-slavery instrument,
demanding the abolition of slavery as a condition of its own
existence, as the supreme law of the land.
This radical change in my opinions produced a corresponding
change in my action. To those with whom I had been in agreement
and in sympathy, I came to be in opposition. What they held to be a
great and important truth I now looked upon as a dangerous error. A
very natural, but to me a very painful thing, now happened. Those
who could not see any honest reasons for changing their views, as I
had done, could not easily see any such reasons for my change, and
the common punishment of apostates was mine.
My first opinions were naturally derived and honestly
entertained. Brought directly, when I escaped from slavery, into
contact with abolitionists who regarded the Constitution as a
slaveholding instrument, and finding their views supported by the
united and entire history of every department of the government, it is
not strange that I assumed the Constitution to be just what these
friends made it seem to be. I was bound not only by their superior
knowledge to take their opinions in respect to this subject, as the
true ones, but also because I had no means of showing their
unsoundness. But for the responsibility of conducting a public
journal, and the necessity imposed upon me of meeting opposite
views from abolitionists outside of New England, I should in all
probability have remained firm in my disunion views. My new
circumstances compelled me to re-think the whole subject, and study
with some care not only the just and proper rules of legal
interpretation, but the origin, design, nature, rights, powers, and
duties of civil governments, and also the relations which human
beings sustain to it. By such a course of thought and reading I was
conducted to the conclusion that the Constitution of the United
States—inaugurated “to form a more perfect union, establish justice,
insure domestic tranquility, provide for the common defense,
promote the general welfare, and secure the blessings of liberty”—
could not well have been designed at the same time to maintain and
perpetuate a system of rapine and murder like slavery, especially as
not one word can be found in the Constitution to authorize such a
belief. Then, again, if the declared purposes of an instrument are to
govern the meaning of all its parts and details, as they clearly
should, the Constitution of our country is our warrant for the abolition
of slavery in every State of the Union. It would require much time
and space to set forth the arguments which demonstrated to my
mind the unconstitutionality of slavery; but being convinced of the
fact my duty was plain upon this point in the farther conduct of my
paper. The North Star was a large sheet, published weekly, at a cost
of $80 per week, and an average circulation of 3,000 subscribers.
There were many times, when in my experience as editor and
publisher, I was very hard pressed for money, but by one means or
another I succeeded so well as to keep my pecuniary engagements,
and to keep my anti-slavery banner steadily flying during all the
conflict from the autumn of 1847 till the union of the States was
assured and emancipation was a fact accomplished. I had friends
abroad as well as at home who helped me liberally. I can never be
too grateful to Rev. Russell Laut Carpenter and to Mrs. Carpenter,
for the moral and material aid they tendered me through all the
vicissitudes of my paper enterprise. But to no one person was I more
indebted for substantial assistance than to Mrs. Julia Griffiths Crofts.
She came to my relief when my paper had nearly absorbed all my
means, and was heavily in debt, and when I had mortgaged my
house to raise money to meet current expenses; and by her
energetic and effective management, in a single year enabled me to
extend the circulation of my paper from 2,000 to 4,000 copies, pay
off the debts and lift the mortgage from my house. Her industry was
equal to her devotion. She seemed to rise with every emergency,
and her resources appeared inexhaustible. I shall never cease to
remember with sincere gratitude the assistance rendered me by this
noble lady, and I mention her here in the desire in some humble
measure to “give honor to whom honor is due.” During the first three
or four years my paper was published under the name of the North
Star. It was subsequently changed to Frederick Douglass’ Paper in
order to distinguish it from the many papers with “Stars” in their titles.
There were “North Stars,” “Morning Stars,” “Evening Stars,” and I
know not how many other stars in the newspaper firmament, and
some confusion arose naturally enough in distinguishing between
them; for this reason, and also because some of these stars were
older than my star I felt that mine, not theirs, ought to be the one to
“go out.”
Among my friends in this country, who helped me in my earlier
efforts to maintain my paper, I may proudly count such men as the
late Hon. Gerrit Smith, and Chief Justice Chase, Hon. Horace Mann,
Hon. Joshua R. Giddings, Hon. Charles Sumner, Hon. John G.
Palfry, Hon. Wm. H. Seward, Rev. Samuel J. May, and many others,
who though of lesser note were equally devoted to my cause.
Among these latter ones were Isaac and Amy Post, William and
Mary Hallowell, Asa and Hulda Anthony, and indeed all the
committee of the Western New York Anti-Slavery Society. They held
festivals and fairs to raise money, and assisted me in every other
possible way to keep my paper in circulation, while I was a non-
voting abolitionist, but withdrew from me when I became a voting
abolitionist. For a time the withdrawal of their coöperation
embarrassed me very much, but soon another class of friends were
raised up for me, chief amongst whom were the Porter family of
Rochester. The late Samuel D. Porter and his wife Susan F. Porter,
and his sisters, Maria and Elmira Porter, deserve grateful mention as
among my steadfast friends, who did much in the way of supplying
pecuniary aid.
Of course there were moral forces operating against me in
Rochester, as well as material ones. There were those who regarded
the publication of a “Negro paper” in that beautiful city as a blemish
and a misfortune. The New York Herald, true to the spirit of the
times, counselled the people of the place to throw my printing press
into Lake Ontario and to banish me to Canada, and while they were
not quite prepared for this violence, it was plain that many of them
did not well relish my presence amongst them. This feeling, however,
wore away gradually, as the people knew more of me and my works.
I lectured every Sunday evening during an entire winter in the
beautiful Corinthian Hall, then owned by Wm. R. Reynolds, Esq.,
who though he was not an abolitionist, was a lover of fair-play and
was willing to allow me to be heard. If in these lectures I did not
make abolitionists I did succeed in making tolerant the moral
atmosphere in Rochester; so much so, indeed, that I came to feel as
much at home there as I had ever done in the most friendly parts of
New England. I had been at work there with my paper but a few
years before colored travelers told me that they felt the influence of
my labors when they came within fifty miles. I did not rely alone upon
what I could do by the paper, but would write all day, then take a
train to Victor, Farmington, Canandaigua, Geneva, Waterloo,
Batavia, or Buffalo, or elsewhere, and speak in the evening,
returning home afterwards or early in the morning, to be again at my
desk writing or mailing papers. There were times when I almost
thought my Boston friends were right in dissuading me from my
newspaper project. But looking back to those nights and days of toil
and thought, compelled often to do work for which I had no
educational preparation, I have come to think that, under the
circumstances it was the best school possible for me. It obliged me
to think and read, it taught me to express my thoughts clearly, and
was perhaps better than any other course I could have adopted.
Besides it made it necessary for me to lean upon myself, and not
upon the heads of our Anti-Slavery church. To be a principal, and not
an agent. I had an audience to speak to every week, and must say
something worth their hearing, or cease to speak altogether. There is
nothing like the lash and sting of necessity to make a man work, and
my paper furnished this motive power. More than one gentleman
from the south, when stopping at Niagara, came to see me, that they
might know for themselves if I could indeed write, having as they
said believed it impossible that an uneducated fugitive slave could
write the articles attributed to me. I found it hard to get credit in some
quarters either for what I wrote or what I said. While there was
nothing very profound or learned in either, the low estimate of Negro
possibilities induced the belief that both my editorials and my
speeches were written by white persons. I doubt if this scepticism
does not still linger in the minds of some of my democratic fellow-
citizens.
The 2d of June, 1872, brought me a very grievous loss. My
house in Rochester was burnt to the ground, and among other things
of value, twelve volumes of my paper, covering the period from 1848
to 1860, were devoured by the flames. I have never been able to
replace them, and the loss is immeasurable. Only a few weeks
before, I had been invited to send these bound volumes to the library
of Harvard University where they would have been preserved in a
fire-proof building, and the result of my procrastination attests the
wisdom of more than one proverb. Outside the years embraced in
the late tremendous war, there has been no period, more pregnant
with great events, or better suited to call out the best mental and
moral energies of men, than that covered by these lost volumes. If I
have at any time said or written that which is worth remembering or
repeating, I must have said such things between the years 1848 and
1860, and my paper was a chronicle of most of what I said during
that time. Within that space we had the great Free Soil Convention at
Buffalo, the Nomination of Martin Van Buren, the Fugitive Slave Law,
the 7th March Speech by Daniel Webster, the Dred Scott decision,
the repeal of the Missouri Compromise, the Kansas Nebraska bill,
the Border war in Kansas, the John Brown raid upon Harper’s Ferry,
and a part of the War against the Rebellion, with much else, well
calculated to fire the souls of men having one spark of Liberty and
Patriotism within them. I have only fragments now, of all the work
accomplished during these twelve years, and must cover this chasm,
as best I can from memory and the incidental items, which I am able
to glean from various sources. Two volumes of the North Star have
been kindly supplied me, by my friend, Marshall Pierce of Saco, Me.
He had these carefully preserved and bound in one cover and sent
to me in Washington. He was one of the most systematically careful
men of all my anti-slavery friends, for I doubt if another entire volume
of the paper exists.
One important branch of my anti-slavery work in Rochester, in
addition to that of speaking and writing against slavery, must not be
forgotten or omitted. My position gave me the chance of hitting that
old enemy some telling blows, in another direction than these. I was
on the southern border of Lake Ontario, and the Queen’s Dominions
were right over the way—and my prominence as an abolitionist, and
as the editor of an anti-slavery paper, naturally made me the station
master and conductor of the underground railroad passing through
this goodly city. Secrecy and concealment were necessary
conditions to the successful operation of this railroad, and hence its
prefix “underground.” My agency was all the more exciting and
interesting, because not altogether free from danger. I could take no
step in it without exposing myself to fine and imprisonment, for these
were the penalties imposed by the fugitive slave law, for feeding,
harboring, or otherwise assisting a slave to escape from his master;
but in face of this fact, I can say, I never did more congenial,
attractive, fascinating, and satisfactory work. True as a means of
destroying slavery, it was like an attempt to bail out the ocean with a
teaspoon, but the thought that there was one less slave, and one
more freeman,—having myself been a slave, and a fugitive slave—
brought to my heart unspeakable joy. On one occasion I had eleven
fugitives at the same time under my roof, and it was necessary for
them to remain with me, until I could collect sufficient money to get
them on to Canada. It was the largest number I ever had at any one
time, and I had some difficulty in providing so many with food and
shelter, but as may well be imagined, they were not very fastidious in
either direction, and were well content with very plain food, and a
strip of carpet on the floor for a bed, or a place on the straw in the
barn loft.
The underground railroad had many branches; but that one with
which I was connected had its main stations in Baltimore,
Wilmington, Philadelphia, New York, Albany, Syracuse, Rochester,
and St. Catharines (Canada). It is not necessary to tell who were the
principal agents in Baltimore; Thomas Garrett was the agent in
Wilmington; Melloe McKim, William Still, Robert Purvis, Edward M.
Davis, and others did the work in Philadelphia; David Ruggles, Isaac
T. Hopper, Napolian, and others, in New York city; the Misses Mott
and Stephen Myers, were forwarders from Albany; Revs. Samuel J.
May and J. W. Loguen, were the agents in Syracuse; and J. P.
Morris and myself received and dispatched passengers from
Rochester to Canada, where they were received by Rev. Hiram
Wilson. When a party arrived in Rochester, it was the business of Mr.
Morris and myself to raise funds with which to pay their passages to
St. Catharines, and it is due to truth to state, that we seldom called in
vain upon whig or democrat for help. Men were better than their
theology, and truer to humanity, than to their politics, or their offices.
On one occasion while a slave master was in the office of a
United States commissioner, procuring the papers necessary for the
arrest and rendition of three young men who had escaped from
Maryland, (one of whom was under my roof at the time, another at
Farmington, and the other at work on the farm of Asa Anthony just a
little outside the city limits,) the law partner of the commissioner, then
a distinguished democrat, sought me out, and told me what was
going on in his office, and urged me by all means to get these young
men out of the way of their pursuers and claimants. Of course no
time was to be lost. A swift horseman was dispatched to Farmington,
eighteen miles distant, another to Asa Anthony’s farm about three
miles, and another to my house on the south side of the city, and
before the papers could be served, all three of the young men were
on the free waves of Lake Ontario, bound to Canada. In writing to
their old master, they had dated their letter at Rochester, though they
had taken the precaution to send it to Canada to be mailed, but this
blunder in the date had betrayed their whereabouts, so that the
hunters were at once on their tracks.
So numerous were the fugitives passing through Rochester that I
was obliged at last to appeal to my British friends for the means of
sending them on their way, and when Mr. and Mrs. Carpenter and
Mrs. Croffts took the matter in hand, I had never any further trouble
in that respect. When slavery was abolished I wrote to Mrs.
Carpenter, congratulating her that she was relieved of the work of
raising funds for such purposes, and the characteristic reply of that
lady was that she had been very glad to do what she had done, and
had no wish for relief.
My pathway was not entirely free from thorns in Rochester, and
the wounds and pains inflicted by them were perhaps much less
easily borne, because of my exemption from such annoyances while
in England. Men can in time become accustomed to almost anything,
even to being insulted and ostracised, but such treatment comes
hard at first, and when to some extent unlooked for. The vulgar
prejudice against color, so common to Americans, met me in several
disagreeable forms. A seminary for young ladies and misses, under
the auspices of Miss Tracy, was near my house on Alexander street,
and desirous of having my daughter educated like the daughters of
other men, I applied to Miss Tracy for her admission to her school.
All seemed fair, and the child was duly sent to “Tracy Seminary,” and
I went about my business happy in the thought that she was in the
way of a refined and Christian education. Several weeks elapsed
before I knew how completely I was mistaken. The little girl came
home to me one day and told me she was lonely in that school; that
she was in fact kept in solitary confinement; that she was not allowed
in the room with the other girls, nor to go into the yard when they
went out; that she was kept in a room by herself and not permitted to
be seen or heard by the others. No man with the feeling of a parent
could be less than moved by such a revelation, and I confess that I
was shocked, grieved, and indignant. I went at once to Miss Tracy to
ascertain if what I had heard was true, and was coolly told it was,
and the miserable plea was offered that it would have injured her
school if she had done otherwise. I told her she should have told me
so at the beginning, but I did not believe that any girl in the school
would be opposed to the presence of my daughter, and that I should
be glad to have the question submitted to them. She consented to
this, and to the credit of the young ladies, not one made objection.
Not satisfied with this verdict of the natural and uncorrupted sense of
justice and humanity of these young ladies, Miss Tracy insisted that
the parents must be consulted, and if one of them objected she
should not admit my child to the same apartment and privileges of
the other pupils. One parent only had the cruelty to object, and he
was Mr. Horatio G. Warner, a democratic editor, and upon his
adverse conclusion, my daughter was excluded from “Tracy
Seminary.” Of course Miss Tracy was a devout Christian lady after
the fashion of the time and locality, in good and regular standing in
the church.
My troubles attending the education of my children were not to
end here. They were not allowed in the public school in the district in
which I lived, owned property, and paid taxes, but were compelled, if
they went to a public school, to go over to the other side of the city,
to an inferior colored school. I hardly need say that I was not
prepared to submit tamely to this proscription, any more than I had
been to submit to slavery, so I had them taught at home for a while,
by Miss Thayer. Meanwhile I went to the people with the question
and created considerable agitation. I sought and obtained a hearing
before the Board of Education, and after repeated efforts with voice
and pen, the doors of the public schools were opened and colored
children were permitted to attend them in common with others.
There were barriers erected against colored people in most other
places of instruction and amusements in the city, and until I went
there they were imposed without any apparent sense of injustice or
wrong, and submitted to in silence; but one by one they have
gradually been removed and colored people now enter freely all
places of public resort without hindrance or observation. This change
has not been wholly effected by me. From the first I was cheered on
and supported in my demands for equal rights by such respectable
citizens as Isaac Post, Wm. Hallowell, Samuel D. Porter, Wm. C.
Bloss, Benj. Fish, Asa Anthony, and many other good and true men
of Rochester.
Notwithstanding what I have said of the adverse feeling exhibited
by some of its citizens at my selection of Rochester as the place to
establish my paper, and the trouble in educational matters just
referred to, that selection was in many respects very fortunate. The
city was, and still is, the center of a virtuous, intelligent, enterprising,
liberal, and growing population. The surrounding country is
remarkable for its fertility; and the city itself possesses one of the
finest water-powers in the world. It is on the line of the New York
Central railroad—a line that with its connections, spans the whole
country. Its people were industrious and in comfortable
circumstances; not so rich as to be indifferent to the claims of
humanity, and not so poor as to be unable to help any good cause
which commanded the approval of their judgment.
The ground had been measurably prepared for me by the labors
of others—notably by Hon. Myron Holley, whose monument of
enduring marble now stands in the beautiful cemetery at Mount
Hope, upon an eminence befitting his noble character. I know of no
place in the Union where I could have located at the time with less
resistance, or received a larger measure of sympathy and
coöperation, and I now look back to my life and labors there with
unalloyed satisfaction, and having spent a quarter of a century
among its people, I shall always feel more at home there than any
where else in this country.
Portrait of John Brown.

CHAPTER VIII.
JOHN BROWN AND MRS. STOWE.
My First Meeting with Capt. John Brown—The Free Soil Movement—Colored

Convention—Uncle Tom’s Cabin—Industrial School for Colored People—
Letter to Mrs. H. B. Stowe.
ABOUT the time I began my enterprise in Rochester I chanced to

spend a night and a day under the roof of a man whose character
and conversation, and whose objects and aims in life made a very
deep impression upon my mind and heart. His name had been
mentioned to me by several prominent colored men, among whom
were the Rev. Henry Highland Garnet and J. W. Loguen. In speaking
of him their voices would drop to a whisper, and what they said of
him made me very eager to see and know him. Fortunately I was
invited to see him in his own house. At the time to which I now refer
this man was a respectable merchant in a populous and thriving city,
and our first place of meeting was at his store. This was a substantial
brick building, on a prominent, busy street. A glance at the interior,
as well as at the massive walls without, gave me the impression that
the owner must be a man of considerable wealth. From this store I
was conducted to his house, where I was kindly received as an
expected guest. My welcome was all I could have asked. Every
member of the family, young and old, seemed glad to see me, and I
was made much at home in a very little while. I was, however, a little
disappointed with the appearance of the house and with its location.
After seeing the fine store I was prepared to see a fine residence, in
an eligible locality, but this conclusion was completely dispelled by
actual observation. In fact, the house was neither commodious nor
elegant, nor its situation desirable. It was a small wooden building,
on a back street, in a neighborhood chiefly occupied by laboring men
and mechanics; respectable enough to be sure, but not quite the
place, I thought, where one would look for the residence of a
flourishing and successful merchant. Plain as was the outside of this
man’s house, the inside was plainer. Its furniture would have
satisfied a Spartan. It would take longer to tell what was not in this
house than what was in it. There was an air of plainness about it
which almost suggested destitution. My first meal passed under the
misnomer of tea, though there was nothing about it resembling the
usual significance of that term. It consisted of beef soup, cabbage,
and potatoes; a meal such as a man might relish after following the
plow all day, or performing a forced march of a dozen miles over a
rough road in frosty weather. Innocent of paint, veneering, varnish, or
table-cloth, the table announced itself unmistakably of pine and of
the plainest workmanship. There was no hired help visible. The
mother, daughters, and sons did the serving and did it well. They
were evidently used to it, and had no thought of any impropriety or
degradation in being their own servants. It is said that a house in
some measure reflects the character of its occupants; this one
certainly did. In it there were no disguises, no illusions, no make
believes. Everything implied stern truth, solid purpose, and rigid
economy. I was not long in company with the master of this house
before I discovered that he was indeed the master of it, and was
likely to become mine too if I stayed long enough with him. He
fulfilled St. Paul’s idea of the head of the family. His wife believed in
him, and his children observed him with reverence. Whenever he
spoke his words commanded earnest attention. His arguments,
which I ventured at some points to oppose, seemed to convince all;
his appeals touched all, and his will impressed all. Certainly I never
felt myself in the presence of a stronger religious influence than
while in this man’s house.
In person he was lean, strong, and sinewy, of the best New
England mould, built for times of trouble, fitted to grapple with the
flintiest hardships. Clad in plain American woolen, shod in boots of
cowhide leather, and wearing a cravat of the same substantial
material, under six feet high, less than 150 pounds in weight, aged
about fifty, he presented a figure, straight and symmetrical as a
mountain pine. His bearing was singularly impressive. His head was
not large, but compact and high. His hair was coarse, strong, slightly
gray and closely trimmed, and grew low on his forehead. His face
was smoothly shaved, and revealed a strong square mouth,
supported by a broad and prominent chin. His eyes were bluish gray,
and in conversation they were full of light and fire. When on the
street, he moved with a long, springing race horse step, absorbed by
his own reflections, neither seeking or shunning observation. Such
was the man, whose name I had heard in whispers, such was the
spirit of his house and family, such was the house in which he lived,
and such was Captain John Brown, whose name has now passed
into history, as one of the most marked characters, and greatest
heroes known to American fame.
After the strong meal already described, Captain Brown
cautiously approached the subject which he wished to bring to my
attention; for he seemed to apprehend opposition to his views. He
denounced slavery in look and language fierce and bitter, thought
that slaveholders had forfeited their right to live, that the slaves had
the right to gain their liberty in any way they could, did not believe
that moral suasion would ever liberate the slave, or that political
action would abolish the system. He said that he had long had a plan
which could accomplish this end, and he had invited me to his house
to lay that plan before me. He said he had been for some time
looking for colored men to whom he could safely reveal his secret,
and at times he had almost despaired of finding such men, but that
now he was encouraged, for he saw heads of such rising up in all
directions. He had observed my course at home and abroad, and he
wanted my coöperation. His plan as it then lay in his mind, had much
to commend it. It did not, as some suppose, contemplate a general
rising among the slaves, and a general slaughter of the slave
masters. An insurrection he thought would only defeat the object, but
his plan did contemplate the creating of an armed force which should
act in the very heart of the south. He was not averse to the shedding
of blood, and thought the practice of carrying arms would be a good
one for the colored people to adopt, as it would give them a sense of
their manhood. No people he said could have self respect, or be
respected, who would not fight for their freedom. He called my
attention to a map of the United States, and pointed out to me the
far-reaching Alleghanies, which stretch away from the borders of
New York, into the Southern States. “These mountains,” he said, “are
the basis of my plan. God has given the strength of the hills to
freedom, they were placed here for the emancipation of the negro
race; they are full of natural forts, where one man for defense will be
equal to a hundred for attack; they are full also of good hiding
places, where large numbers of brave men could be concealed, and
baffle and elude pursuit for a long time. I know these mountains well,
and could take a body of men into them and keep them there despite
of all the efforts of Virginia to dislodge them. The true object to be
sought is first of all to destroy the money value of slave property; and
that can only be done by rendering such property insecure. My plan
then is to take at first about twenty-five picked men, and begin on a
small scale; supply them arms and ammunition, post them in squads
of fives on a line of twenty-five miles, the most persuasive and
judicious of whom shall go down to the fields from time to time, as
opportunity offers, and induce the slaves to join them, seeking and
selecting the most restless and daring.”
He saw that in this part of the work the utmost care must be
used to avoid treachery and disclosure. Only the most conscientious
and skillful should be sent on this perilous duty; with care and
enterprise he thought he could soon gather a force of one hundred
hardy men, men who would be content to lead the free and
adventurous life to which he proposed to train them, when these
were properly drilled, and each man had found the place for which
he was best suited, they would begin work in earnest; they would run
off the slaves in large numbers, retain the brave and strong ones in
the mountains, and send the weak and timid to the north by the
underground railroad; his operations would be enlarged with
increasing numbers, and would not be confined to one locality.
When I asked him, how he would support these men? he said
emphatically, he would subsist them upon the enemy. Slavery was a
state of war, and the slave had a right to anything necessary to his
freedom. But said I, “suppose you succeed in running off a few
slaves, and thus impress the Virginia slaveholder with a sense of
insecurity in their slaves, the effect will be only to make them sell
their slaves further south.” “That,” said he, “will be first what I want to
do; then I would follow them up. If we could drive slavery out of one
county, it would be a great gain; it would weaken the system
throughout the state.” “But they would employ bloodhounds to hunt
you out of the mountains.” “That they might attempt,” said he, “but
the chances are, we should whip them, and when we should have
whipt one squad, they would be careful how they pursued.” “But you
might be surrounded and cut off from your provisions or means of
subsistence.” He thought that could not be done so they could not
cut their way out, but even if the worst came, he could but be killed,
and he had no better use for his life than to lay it down in the cause
of the slave. When I suggested that we might convert the
slaveholders, he became much excited, and said that could never
be, “he knew their proud hearts and that they would never be
induced to give up their slaves, until they felt a big stick about their
heads.” He observed that I might have noticed the simple manner in
which he lived, adding that he had adopted this method in order to
save money to carry out his purposes. This was said in no boastful
tone, for he felt that he had delayed already too long and had no
room to boast either his zeal or his self denial. Had some men made
such display of rigid virtue, I should have rejected it, as affected,
false and hypocritical, but in John Brown, I felt it to be real as iron or
granite. From this night spent with John Brown in Springfield, Mass.,
1847, while I continued to write and speak against slavery, I became
all the same less hopeful of its peaceful abolition. My utterances
became more and more tinged by the color of this man’s strong
impressions. Speaking at an anti-slavery convention in Salem, Ohio,
I expressed this apprehension that slavery could only be destroyed
by bloodshed, when I was suddenly and sharply interrupted by my
good old friend Sojourner Truth with the question, “Frederick, is God
dead?” “No.” I answered, and “because God is not dead slavery can
only end in blood.” My quaint old sister was of the Garrison school of
non-resistants, and was shocked at my sanguinary doctrine, but she
too became an advocate of the sword, when the war for the
maintenance of the Union was declared.
In 1848 it was my privilege to attend, and in some measure to
participate in the famous Free-Soil Convention held in Buffalo, New

Advances in Heat Transfer 52 1St Edition J P Abraham Full Chapter

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Advances in Heat Transfer 52 1St Edition J P Abraham Full Chapter

Uploaded by

Copyright:

Available Formats

Advances in heat transfer 52 1st Edition

First edition 2020

Copyright © 2020 Elsevier Inc. All rights reserved.

For information on all Academic Press publications

Publisher: Zoe Kruze

Advances in Heat Transfer, Volume 52 # 2020 Elsevier Inc. 1

originally developed for forced wall-bounded turbulence such as shear-

on local properties becomes meaningless when the buoyancy effect is

Fig. 1 Canonical configurations of buoyancy-driven convection. (A) Differentially

and DHVC. Analysis of turbulent quantities, especially those involved

2. Prediction of Nusselt number

Material ν Kinematic viscosity [L2/t]

def gβΘmp δ3 1 gβΔT ð2δÞ

mid-plane. Such a defined Grashof number is compatible with the friction

Thus, one-third power of Grashof number can be interpreted as the ratio of

Nu ¼ f ðPr, RaÞ: (6)

Historically, power-law is often used to correlate the Nusselt number with

Nu C Prn Ram : (7)

While experimental measurements of Nusselt number for DHVC are scant,

Table 2 Examples of correlations for Nusselt number in RBC.

the similarity theory to include a double boundary layer. One interesting

Fig. 4 Nusselt number in DHVC. (A) Nusselt number vs Rayleigh number.

Table 3 Correlations for Nusselt number in DHVC.

2012 Kiš and Herwig Nu ¼ 0.186(Pr Ra)1/3.2

2.1 Lessons from forced convection

Nu 0:023 Pr0:4 Re0:8

In Section 9, a new prediction of the Nusselt number in buoyancy-

Angle brackets h i denote the conceptual ensemble averaging. β is the ther-

z¼0: U ¼ 0, Θ ¼ 0, Rwu ¼ 0, Rwθ ¼ 0: (14a)

In the study of forced wall-bounded turbulence, two extremely impor-

3.2 Laminar DHVC solution

turbulent DHVC are presented in Fig. 5B at two Grashof numbers. In tur-

zU max ¼ ð1 31=2 Þδ 0:423 δ; (19a)

4. Dimensional analysis of buoyancy-driven convection

Table 4 Dimensional analysis of laminar DHVC.

Length scale lc ¼ δ Ψ(Pr, Gr) lc ¼ δ Ψ(Pr, Gr)

4.1 Dimensional analysis of laminar DHVC

channel mid-plane. In contrast, the two control parameters qt and qu in

4.2 Dimensional analysis of turbulent DHVC

Velocity scale ui ¼ uτΨ(Pr, Rii) uo ¼ uτΨ(Rio)

It is obvious that Θmp cannot be an appropriate control parameter in the

def gβqt ν gβθτ ν Gr Nu Gr f

def gβqt δ gβθτ δ Gr Nu Gr f

For a turbulent DHVC at a sufficiently high Rayleigh number, there is a

Therefore, the inner Richardson number can be interpreted as the ratio

Fundamentally, flow and heat transport in turbulent DHVC is controlled by

Table 6 Nondimensional numbers that can be used in the study of DHVC.

Rayleigh number def

Inner Richardson number def

Friction Reynolds number def

Having remained abroad nearly two years, and being about to

New Experiences—Painful Disagreement of Opinion with old Friends—Final

PREPARED as I was to meet with many trials and perplexities on

Portrait of John Brown.

My First Meeting with Capt. John Brown—The Free Soil Movement—Colored

ABOUT the time I began my enterprise in Rochester I chanced to

You might also like