a water-based heat pipe for molten steel applications in a tundish · 2018-12-03 · a water-based...
Post on 13-Apr-2020
9 Views
Preview:
TRANSCRIPT
A Water-Based Heat Pipe for Molten
Steel Applications in a Tundish
by
Nogol Madani
Department of Mining and Materials Engineering
McGill University
Montreal, Canada
July 2010
A Thesis Submitted to the Faculty of Graduate Studies and Research
In Partial Fulfillment of the Requirements
For the Degree of Master of Engineering
Nogol Madani
2010
II
Abstract
In the continuous casting of steel shapes, a tundish feeds the molten steel alloy into
oscillating moulds. Therefore, a better knowledge of some of the parameters of the steel
in this phase of the process will lead to better quality with more productivity and in a
safer environment. There are many control parameters to be measured in the tundish,
especially on a continuous basis, yet in this project we are focused on the following 3
main ones that are the basics of tundish metallurgy: the temperature of the steel, the
thickness of the slag, and the composition of the steel.
A viewing tube capable of withstanding the intense environment of the steel tundish
enables one to obtain information about temperature, slag thickness and composition of
the melt on a continuous basis. This probe based on heat pipe technology has been
designed and fabricated and tested in two different environments. One was in gaseous
environment in a natural gas filled furnace. The other environment in which the heat pipe
has been tested was in a molten aluminum bath. The results of the first experiments
showed that the heat pipe is capable of extracting 23 kW at a heat flux of 0.5 MW/m2
while in the next set of tests the heat pipe could not overcome the film boiling limitation.
This was a result of the internal design of the pipe, thus, some design modifications are
suggested to solve the problem.
III
Résumé
Dans la coulée en continu pour produire des formes en acier, un panier de coulée alimente
des moules basculants en acier en fusion. Par conséquence, une meilleure connaissance de
certains des paramètres de l'acier dans cette phase du processus permettra d’améliorer la
qualité ainsi que la productivité dans un environnement plus sûr. Il y a beaucoup de
paramètres de contrôle qui peuvent être mesurés dans le panier de coulée, surtout sur une
base continue, cependant cette étude concerne les trois principaux qui représentent la
base fondamentale métallurgique du panier de coulée : la température de l'acier,
l'épaisseur des scories, et la composition de l'acier.
Un tube capable de résister à l'environnement intense du panier de coulée et permettant
l’observation in-situ a été utilisé pour obtenir l'un rend capable d'obtenir des informations
sur la température, l'épaisseur de scories et la composition du fond sur une base continue.
Cette sonde fondée sur la technologie des caloducs a été conçue, fabriquée et essayée
dans deux environnements différents. L'un était dans un environnement gazeux dans une
fournaise alimentée en gaz naturel. L'autre environnement dans lequel le caloduc a été
essayé était dans un bain d'aluminium en fusion. Les résultats des premières expériences
ont montré que le caloduc est capable d'extraire 23 kW à un flux de chaleur de 0,5
MW/m pendant que dans la prochaine série de tests le caloduc ne pouvait pas surmonter
la limite imposée par le film qui bout. Ceci était un résultat de la conception interne du
caloduc, ainsi, quelques modifications de conception sont suggérées pour résoudre le
problème.
IV
Acknowledgments
I would like to express my sincere gratitude to my thesis supervisor, Professor Frank
Mucciardi and co-supervisor, Dr. André Moreau for their guidance, encouragement and
support through the whole project. Their scientific approach and hard work are examples
to emulate in the future.
Special thanks are due to Salar Tavakoli, Csaba Szalacsi, Roberto Tariello, Jean-Guy
Allard and Antoine Pelletier for their enormous help in the design and the fabrication of
the heat pipe. I would also like to thank CMQ, Centre de métallurgie du Québec for
sharing their space and facilities with us for the aluminum tests.
The pleasant and friendly international environment created by my fellow graduate
students along with their knowledge and skills are deeply appreciated. I gratefully
acknowledge McGill University especially the Mining and Materials Engineering
department for giving me this opportunity and IMI-NRC for the financial support.
Last but not least, I would like to thank my parent, sister, brother and my best friend
Saadi Daftari for their understanding, patience and continuous support.
V
Table of Contents
Abstract -------------------------------------------------------------------------------------- II
Acknowledgments--------------------------------------------------------------------------- IV
Table of Contents--------------------------------------------------------------------------- V
List of Figures------------------------------------------------------------------------------- VIII
List of Tables ------------------------------------------------------------------------------- X
Chapter 1 Introduction-------------------------------------------------------------------- 1
1.1 CONTINUOUS STEEL TUNDISH MEASUREMENTS----------------------- 1
1.2 OBJECTIVES-------------------------------------------------------------------------- 1
1.3 THESIS ORGANIZATION---------------------------------------------------------- 3
Chapter 2 Literature Survey------------------------------------------------------------- 4
2.1 TUNDISH MEASUREMENTS IN CONTINUOUS STEEL CASTING----- 4
2.1.1 Various Methods for Measuring Temperature in the Steel Tundish--------- 5
2.1.2 Slag Thickness Measurements in Steel Tundish------------------------------- 7
VI
2.1.3 Analysis of the Composition of the Melt in the Steel Tundish--------------- 9
2.2 STRATEGY---------------------------------------------------------------------------- 10
2.3 HEAT PIPE TECHNOLOGY-------------------------------------------------------- 14
2.3.1 Heat Pipe History------------------------------------------------------------------ 14
2.3.2 Classical Heat Pipes--------------------------------------------------------------- 15
2.3.2.1 Classical Heat Pipe Applications--------------------------------------------- 17
2.3.2.2 Classical Heat Pipe Limitations---------------------------------------------- 18
2.3.3 McGill Heat Pipe------------------------------------------------------------------- 22
2.3.3.1 Development-------------------------------------------------------------------- 22
2.3.3.2 Design and Operating Principles of the McGill Heat Pipe---------------- 23
2.3.3.3 Choking Limitation in the McGill Heat Pipe------------------------------- 26
2.3.3.4 McGill Heat Pipe Applications----------------------------------------------- 27
2.3.4 Heat Pipe Probe for a Tundish--------------------------------------------------- 31
Chapter 3 Experimental Procedure---------------------------------------------------- 34
3.1 THE DESIGN AND FABRICATION OF THE HEAT PIPE------------------- 34
3.1.1 Evaporator--------------------------------------------------------------------------- 34
3.1.2 Condenser--------------------------------------------------------------------------- 39
3.2 EXPERIMENTAL SETUP----------------------------------------------------------- 42
3.2.1 Sensors------------------------------------------------------------------------------- 45
Chapter 4 Results and Discussions------------------------------------------------------ 47
PART 1: GAS FURNACE TESTS-------------------------------------------------------- 47
VII
4.1 REACHING STEADY STATE------------------------------------------------------ 48
4.2 MEASURING THE EXTRACTED HEAT----------------------------------------- 52
4.3 COMPENSATING FOR HEAT LOSSES------------------------------------------ 53
4.4 CHOKING LIMITATION (DRY-OUT)-------------------------------------------- 61
PART 2: MOLTEN ALUMINUM TESTS----------------------------------------------- 63
4.5 MODIFYING THE INTERNAL DESIGN OF THE EVAPORATOR-------- 66
Chapter 5 Conclusions--------------------------------------------------------------------- 67
References------------------------------------------------------------------------------------- 69
VIII
List of Figures
Figure 1.1- A schematic of the probe in a steel tundish-------------------------------- 2
Figure 2.1- Continuous steel casting------------------------------------------------------ 4
Figure 2.2- Schematic of the LIBS-------------------------------------------------------- 10
Figure 2.3- The emissivity of liquid steel and molten slag---------------------------- 12
Figure 2.4- Pouring sequences of (a) steel and (b) slag-------------------------------- 13
Figure 2.5- Classical heat pipe------------------------------------------------------------ 16
Figure 2.6- Schematic of loop heat pipe------------------------------------------------- 18
Figure 2.7- Natural and forced convection boiling curves----------------------------- 20
Figure 2.8- An example of a flow modifier---------------------------------------------- 24
Figure 2.9- McGill heat pipe after stable and unstable operations------------------- 27
Figure 2.10- Conventional and heat pipe cooled copper blocks---------------------- 28
Figure 2.11- Copper taphole block under heat loading from a gas burner---------- 29
Figure 2.12- Slag launder (copper)------------------------------------------------------- 30
Figure 2.13- Copper launder under intense heat load---------------------------------- 31
Figure 2.14- Temperature mapping of launder------------------------------------------ 31
Figure 2.15- The views of water based heat pipe immersed in aluminum---------- 32
Figure 3.1- Flow modifier used in McGill heat pipe----------------------------------- 35
Figure 3.2- Solid Works drawing of the evaporator------------------------------------ 36
IX
Figure 3.3- The top and side view of the evaporator----------------------------------- 37
Figure 3.4- Nozzle in the evaporator from different angles--------------------------- 38
Figure 3.5- Different views of fabricated nozzle---------------------------------------- 38
Figure 3.6- Assembled evaporator ------------------------------------------------------- 39
Figure 3.7- (a) Schematic of the condenser, (b) Fabricated condenser-------------- 41
Figure 3.8- Heat pipe during an experiment in gas furnace--------------------------- 42
Figure 3.9- Equipments for heat measurement------------------------------------------ 45
Figure 4.1- Heat pipe experiment in gas furnace level 1------------------------------- 49
Figure 4.2- Heat pipe experiment in gas furnace level 2------------------------------- 51
Figure 4.3- Heat pipe experiment in gas furnace level 3------------------------------- 52
Figure 4.4- The curve of extracted heat vs. temperature in setup level 1----------- 55
Figure 4.5- The curve of extracted heat vs. temperature in setup level 2----------- 57
Figure 4.6- The curve of extracted heat vs. temperature in setup level 3----------- 59
Figure 4.7- The dry-out phenomenon in an experiment in gas furnace level 3----- 62
Figure 4.8- The evaporator after gas furnace experiments---------------------------- 62
Figure 4.9- The behaviour of the heat pipe and the condenser in aluminum bath-- 64
Figure 4.10- The thin frozen aluminum layer on the evaporator’s surface---------- 65
Figure 4.11- The view of corroded area on the evaporator---------------------------- 65
X
List of Tables
Table 4.1- A set of data captured in furnace setup level 1---------------------------- 54
Table 4.2- A set of data captured in furnace setup level 2---------------------------- 58
Table 4.3- A set of data captured in furnace setup level 3---------------------------- 59
[1]
Chapter 1
Introduction
1.1 CONTINUOUS STEEL TUNDISH MEASURMENTS
Several technologies have been coupled in the development of an innovative probe for
use in the tundish. This probe can be used as a viewing tube to measure the three most
important parameters of temperature, slag thickness and composition on a continuous
basis. The corrosive environment of the steel tundish along with its high temperature
makes this target difficult to achieve. Even though, there are some continuous methods
for temperature and slag thickness measurements in continuous steel casting, this industry
still lacks a method with which one can obtain a composition analysis in the steel tundish
on a continuous basis. Among the methods that have been developed recently for
temperature and slag thickness measurements, there is no single method which meets all
criteria in this industry such as cost, safety, accuracy and sufficient life time ensemble.
Therefore, researchers and experts in this field are still struggling to find a solution for a
more appropriate continuous measurement technique.
1.2 OBJECTIVES
In this project we intend to develop a probe that is based on heat pipe technology, and
which will allow electronic devices of various types to view the liquid steel from above
the tundish in order to make strategic measurements of the melt (Figure 1.1). We have
[2]
formulated four test programs to pursue the ultimate target of a steel environment. They
are:
a) Gas furnace
b) Molten aluminum
c) Molten cast iron
d) Molten steel
The purpose of the research presented in this thesis is the following:
1) To design a heat pipe capable of withstanding the steel tundish environment
2) To evaluate the functionality of the design in addressing the heat extraction,
durability, safety and cost in the first two test programs listed above.
Figure 1.1- A schematic of the probe in a steel tundish
[3]
1.3 THESIS ORGANIZATION
Chapter 2 reviews the literature pertaining to this thesis. A summary of the current state
of measurements technologies in the field of steel continuous casting is given. In addition,
the strategy of this project along with the history of heat pipe technology is described in
detail. Finally, the McGill heat pipe with its new features and applications are presented.
Chapter 3 describes the design of the new heat pipe and the process of its fabrication.
Moreover, this chapter includes the experimental setups for the first two phases of the
project meaning gas furnace and molten aluminum tests.
In Chapter 4 we describe the results which have been obtained from the experiments. In
each setup the capability of the system for maximum heat extraction along with the
behaviour of the heat pipe towards the limitations of film boiling and choking is verified.
In addition, some ideas for further experiments in the future are presented. Chapter 5
summarizes the conclusions of this work.
[4]
Chapter 2
Literature Survey
2.1 TUNDISH MEASUREMENTS IN CONTINUOUS STEEL
CASTING
Several technologies have been developed in the search of an innovative probe to be used
in the steel tundish. Steel represents 93% of all metal production out of which 96% is
continuously cast. In continuous casting, the molten steel is poured from the ladle into the
tundish where it is fed to oscillating moulds for casting (Figure 2.1). The phenomena that
occur in the steel ladle affect those taking place in the tundish directly and later, what
goes on in the continuous casting mould and downstream will be reflected in the
intermediate and semi-finished quality of the cast products.
Figure 2.1- Continuous steel casting [1]
[5]
There are several parameters that can be of use if they can be measured in the tundish
environment, especially, on a continuous basis. Among them, there are three main ones
which we are focused on in this project. They are the temperature of the steel, the
thickness of the slag and the composition of the melt. More knowledge of these
parameters will lead to greater quality and higher productivity in a safe environment in
the production of continuously cast steel products.
2.1.1 Various Methods for Measuring Temperature in the Steel Tundish
In steel continuous casting, both high and low superheats can have negative effects. High
superheat above the liquidus temperature in the tundish can increase central segregation
and affect grain size which would finally interrupt the continuous casting sequence. In
addition, high superheats also lead to reduced productivity. On the other hand, low
superheat in the tundish can clog tundish nozzles, entrap macro inclusions which affect
flux powder melting and consequently increase the probability of mould sticker formation
[2]. Therefore, the benefits of continuous temperature measurement are better steel
quality, and an increase in machine productivity in a safer environment.
Steel industry uses the instruments such as radiation pyrometers, optical pyrometers, and
thermocouples, etc. to measure the temperature of molten steel. For a batch measurement,
thermocouples are most preferable. Most commonly used thermocouples are consumable
ones having a heat sensing portion which comprises platinum-rhodium-based
thermocouple and a quartz glass tube to protect the tip of the thermocouple. When this
thermocouple is immersed in molten steel, the heat sensing leading end is destroyed in a
[6]
very short time (10 to 20 seconds) so that the temperature measurement operation is
carried out in a short time frame. In addition, the heat sensing portion must be replaced
for each measurement. Therefore, we can conclude that this kind of thermocouple cannot
continuously measure a temperature for a long interval, and it is expensive to measure
temperatures frequently. Moreover, the characteristics of each thermocouple differ from
one another, and it is difficult to maintain the accuracy of temperature measurement [3].
Nowadays, the adopted method is the optical infra-red pyrometer, whose operating
principal is based on the light emitted by a certain material. The light is captured by a
sensor and transmitted by optical fibre up to a signal converter. The readings are
processed by the electronic signal converter by using a mathematical equation that
calculates and displays the temperature. This system consists of a measuring device,
articulated manipulator, signal conversion device and a control system. The measuring
tube is a ceramic tube that is immersed into the molten steel and works as a viewing port
for the optical sensor [4].
From the economic point of view, the continuous method has proved to be more
advantageous than the batch measurements, yet it still encounters certain technical and
operating problems such as high temperature in the sensor body, the cooling of the sensor
and optical fibre without interfering with the measurement, the loss in accuracy arising
from the distance between the sensor and the bottom of the ceramic tube and the effect of
the steel bath level changing during the measurements [4].
Therefore, in 2008, a measuring device featuring a cooled jacket has been developed to
protect the optical infra-red sensor. The infra-red sensor which is located inside a ceramic
tube made of pressed graphite alumina enables the system to measure the temperature
accurately while the steel is in the tundish.
[7]
Using this measuring device has some advantages such as exposing the operator to high
temperature for less time which leads to enhanced operating safety, lower operating cost,
reduced maintenance cost and higher measurement reliability. Yet this method has a
major problem and that is the replacement of the ceramic tube every 15 hours of
operation [4].
2.1.2 Slag Thickness Measurements in Steel Tundish
In steelmaking, after the basic oxygen process, some of the slag floating on top of the
molten steel follows the steel when it is poured into ladles and a fraction of that flows
from the ladle into the tundish. While some slag is required to absorb impurities and
inclusions, excessive quantities of slag can lead to serious processing scenarios which can
impact the quality of the cast product and create dangerous working conditions for the
personnel. Therefore, it is necessary to know the amount of the slag that resides on the
steel in the tundish.
There are several methods to measure the thickness of the slag in the steel tundish.
Richard E. Kracich and Kenneth Goodson, in 1996, described a slag depth measuring
system including measuring the lower level of slag by a "slag/steel interface electronics"
coil, a probe which must penetrate the slag layer and beyond, where the induction effects
of the coil are changed by the presence of the molten steel [5].
The other kind of probe used in the past is a simple steel bar that is inserted until its
extremity touches the bottom of the molten steel. The thickness of the slag is the length of
the bar that has been less heated. Because of the much higher thermal conductivity of the
steel, the portion that resides in the steel appears to be hotter than that which was in the
[8]
slag. Sometimes, the upper crust of the slag does not heat the bar enough and hence the
measurement may be compromised. Nevertheless, such devices can often be relied upon
at least for determination of the upper level of the molten steel that can be measured
reliably at a single point.
Microwave radar which has been used to measure the levels of various substances is
another option for the slag thickness measurements. These devices that have recently
become more popular for their cost, size and accuracy work with a basic principle and
that is the distance from the radar sender/receiver which is directly related to the elapsed
time between sending and receiving the radar signal [6]. However, despite the advantages
that these systems have, Kracich and Goodson disapproved using of a microwave unit for
measuring slag thickness due to time constraint, cost and durability in the harsh
environment [5].
But in 2000, a new radar unit has been developed which can measure the slag thickness
with high accuracy. This device is based on the enormous difference between the
electrical conductivities of slag and steel. The electrically conductive materials such as
steel with a high electrical conductivity of λ=7140 Ω-1
at steelmaking temperatures have
good reflection characteristics for reliable measurements. On the other hand, slag with
electrical conductivities λ of about λ=0.4 to 0.7 Ω-1
is desirable for its low reflection. The
difference in conductivity, a factor of about 7000, between slag and molten steel can be
used in the radar to determine the position of the interface of the slag and the molten steel.
Despite other microwave radars which report solely the upper level of the slag, this
invention recognizes the distance from both the upper and lower surfaces of the slag and
consequently the difference which is the thickness [6].
[9]
The device appears to work well but it has been described as expensive and delicate.
2.1.3 Analysis of the Composition of the Melt in the Steel Tundish
At this time, in the steelmaking industry, a technology that enables operators to have an
on-line and in-situ analysis of the composition of the melt does not exist. Some methods
of in-situ analysis have been proposed with more or less success. Attempts have been
made by some researchers to generate an aerosol powder from a melt either using a probe
containing an atomization die or using laser ablation to evaporate a sample from the
molten metal. Thereafter, the removed material is transported by an inert gas line to an
inductively-coupled plasma torch remote from the probe, where the metal powder is
finally analyzed. However, the transport of the aerosol may induce subtle errors in the
results due to selective evaporation of the volatile elements in the particles [7, 8].
Moreover, some in-situ analysis has been done using LIBS (laser-induced breakdown
spectroscopy) in lab scales [9]. In the LIBS technique, a laser pulse is sent to a melt
surface and this creates a plasma on the surface of the melt. As the plasma cools down, it
emits light which is focused by collector lenses that direct the light to a fibre optic as
shown in Figure 2.2 [10]. The light is carried by the fibre to a spectrometer which
performs the chemical analysis. In lab scale, the plasma is created on the molten steel
surface and hence one can analyse its composition by installing the device near to the
steel crucible while in the industrial scale, the molten steel is covered by slag and
therefore we need a viewing tube to be immersed into the tundish deep enough that the
technique would be able to analyse the composition of steel and not the slag.
[10]
Figure 2.2- Schematic of the LIBS [10]
Many have tried to make a variety of such tubes for prolonged immersion in steel. Given
the temperatures that the tube is exposed to, and given the corrosive slag it is in contact
with, and given that it may become frozen in the top layer of the slag, finding a viable
tube (and configuration) to use has been a formidable task that has yet to be resolved.
Therefore, laboratory analysis on a batch basis is the only feasible method to determine
the composition of the melt for now.
2.2 STRATEGY
Our research work is tied into providing the other groups of researchers a clear view of
the molten steel from above the tundish. With the use of a viewing tube based on heat
pipe technology, one can measure the temperature of the molten steel on a continuous
basis with a radiation pyrometer. This technique that has been used for a number of
decades uses the emissivity of clean steel surface to measure its temperature. However, in
many cases, the temperature of the steel is measured when the steel is tapped.
[11]
It is not possible to measure the steel temperature accurately if it is covered by slag. Thus,
in a tundish where the steel pool is covered by a slag layer, the use of a radiation
pyrometer to directly measure the steel temperature is not possible by simply pointing the
pyrometer at the slag surface. Now, if one inserts a tube of suitable diameter in the melt
such that it penetrates through the slag and into the steel, one then has a clear view of the
steel melt through the tube. To keep the slag or steel from entering the tube, an inert gas
such as argon is continuously bubbled through the tube. With a clear view of the steel
melt, one can then use a pyrometer to measure the temperature of the steel [11].
For a given wavelength and temperature, the amount of thermal radiation emitted by an
object is directly proportional to the spectral emissivity of the object's surface. This is
summarized in Planck's radiation equation (equation 2-1):
2
5
1( , )
1C
T
CI T
e
(2-1)
Where ε is the spectral emissivity, the wavelength, T the temperature, I (, T) is the
spectral emissive power and c1 and c2 are constants. Thus, it can be seen that the
temperature of the surface can be determined by measuring the spectral emissive power
of the surface. This is the basic principle most pyrometers are based on [11].
To measure the location of the liquid steel/liquid slag interface in the tundish, one can
design a technique that is underpinned by the fact that the emissivity of steel and slag are
substantially different for most wavelengths. This variation between the emissivity of
steel and slag is the basic principle behind a device that was developed by Tata Steel to
detect the flow of slag during the tapping of a furnace. Normally, as one taps a furnace,
[12]
the stream comprises steel. Near the end of the tapping, the stream will also carry slag.
The device from Tata Steel detects the appearance of slag in the stream and alerts the
operator [11].
Liquid slag has an emissivity that is about 0.8 while liquid steel has a relatively low
emissivity of about 0.3 (except at the very small wavelengths). This is shown in Figure
2.3 [12]. Photos of the steel tapping stream and the slag tapping stream are shown in
Figure 2.4 [12]. One notes that the steel stream shown in Fig. 5a appears dull as viewed
by the camera while the slag stream in Fig. 5b appears substantially brighter. The
difference between the 2 streams is the emissivity. Because steel has a lower emissivity
(about 0.4 at a wavelength of 2 μm), its heat flux is ½ that of slag at the same
temperature.
Figure2.3- The emissivity of liquid steel and molten slag [12]
[13]
(a) (b)
Figure 2.4- Pouring sequences of: (a) steel and (b) slag [12]
Our strategy for measuring the steel/slag interface requires that the tube through which
the steel is sighted for temperature measurements be moved in increments through the
steel into the slag phase. As the tube is moved in the vertical direction, temperature
measurements are taken with the pyrometer. When the front opening of the tube crosses
from the steel phase into the slag phase, the pyrometer will show a relatively large jump
in temperature (in the order of several hundred oC). This occurs not because the
temperature is higher but rather because the emissivity of the slag is higher. The
pyrometer picks up more infrared radiation and translates this to a sharp increase in
temperature. In this way, this device can be used to measure how much slag there is in the
tundish at any given time [11].
For measuring the third parameter, composition of the molten steel, with a clear pathway
to the molten steel, one can use the LIBS technique that has been described earlier to
generate an in-situ and online analysis of the composition of the molten steel.
[14]
The next section describes the basics of heat pipe technology, its applications and
limitations following some of the significant technological advances that the McGill heat
pipe group have made in recent years and with particular emphasis on those factors that
will help us achieve the objectives.
2.3 HEAT PIPE TECHNOLOGY
2.3.1 Heat Pipe History
Peterson in his book “An Introduction to Heat Pipe Modeling, Testing and Applications”
explains the initial concept of the heat pipe which goes back to mid 1800‟s in the patents
of A. M. Perkins and J. Perkins. Perkins tube which has been described in these patents
used either single or two-phase processor to transfer heat form a furnace to a boiler. Later,
Gay used the idea of the Perkins tube and added number of vertical tubes that were
arranged with the evaporator placed below the condenser. All these devices which are
categorized as thermosiphons generated the idea for the later development of what we
today call the heat pipe [13]. The modern concept for a capillary driven heat pipe was
first suggested by R.S. Gaugler of General Motors in 1944 who patented the idea [14].
Early 60s‟interest in heat pipes followed its independent invention by Grover in 1962 [15]
as a solution to heat transfer problems encountered in nuclear-powered thermionic
generators for use as space power supplies. During the late 1960s NASA played a large
role in heat pipe development by funding a significant amount of research on their
applications and reliability in space flight following from Grover's suggestion.
[15]
The first application of heat pipes in the space program was in thermal equilibration of
satellite transponders. As satellites orbit one side is exposed to the direct radiation of the
sun while the opposite side is completely dark and exposed to the deep cold of outer
space. This causes severe discrepancies in the temperature (and thus reliability and
accuracy) of the transponders. The heat pipe cooling system designed for this purpose
managed the high heat fluxes and demonstrated flawless operation with and without the
influence of gravity. The developed cooling system was the first description and usage of
variable conductance heat pipes to actively regulate heat flow or evaporator temperature
[16].
2.3.2 Classical Heat Pipes
A heat pipe is a sealed container of any shape that is evacuated of any substance other
than an appropriate working fluid. Therefore, the working fluid exists both as a liquid and
as a gas within a container, and is constantly at the saturation pressure and temperature, in
the other word the working fluid is constantly at its boiling point (Figure 2.5). The basic
construction of a classical heat pipe consists of the following:
a) An evaporator section, where an external heat source is applied
b) A condenser section, where external cooling is applied
c) And a capillary structure (wick)
Working fluid inside the device is boiled in the evaporator, is then transported to the
condenser as a vapour via pressure-driven flow, and is then subsequently condensed. The
condensed working fluid is returned to the evaporator primarily by capillary forces
supplied by the wick [17].
[16]
Figure 2.5- Classical heat pipe
Heat energy is stored in the working fluid as a phase change. In the evaporator, the
following process occurs:
Liquid + heat from heat source vapour
And in the condenser, the following occurs:
Vapour liquid + heat to external cooling circuit
Therefore, by adding the above equations, the overall process at steady-state is:
Heat from heat source heat to external cooling circuit
Heat pipe is, in effect, a „superconductor‟ of heat energy. Tests have shown that a heat
pipe can be as effective in transporting energy as 1,000 times the equivalent quantity of
copper under similar heat transfer conditions.
[17]
2.3.2.1 Classical Heat Pipe Applications
Grover and his colleagues were working on cooling systems for nuclear power cells for
space craft, where extreme thermal conditions are found. Heat pipes have since been used
extensively in spacecraft as a means for managing internal temperature conditions [15].
Present applications of classical heat pipes include electronics, plastic mould injection,
cryogenic medicine, cutting tools and etc. [18]. Nowadays, they are being used in many
modern computer systems, where increased power requirements and subsequent increases
in heat emission have resulted in greater demands on cooling systems. Heat pipes are
typically used to move heat away from components such as CPUs (central processing
unit) and GPUs (graphics processing unit) to heat sinks where thermal energy may be
dissipated into the environment.
One of the other applications of the heat pipe is in the Trans-Alaska pipeline where heat
pipes function as thermal diodes and are used to stabilize the permafrost foundation of the
pipeline [19].
In 1972, Loop Heat Pipes (LHPs) were developed when the first such device with a
length of 1.2 m and a capacity of about 1 kW, with water as a working fluid was created
and tested successfully by the Russian scientists Grasimov and Maydanik from the Ural
Polytechnical Institute. The appearance of LHPs was a response to the challenge
connected with the acute demand of aerospace technology for highly efficient heat-
transfer devices with all the main advantages of conventional heat pipes, but at the same
time much less sensitive to the change of orientation in the gravity field.
Loop heat pipes (LHPs) are two-phase heat-transfer devices with capillary pumping of a
working fluid (Figure 2.6). They possess all the main advantages of conventional heat
[18]
pipes, but owing to the original design and special properties of the capillary structure are
capable of transferring heat efficiently for distances up to several meters at any
orientation in the gravity field, or for several tens of meters in a horizontal position.
Besides, the LHP conception allows a wide variety of different design embodiments,
which essentially extends the sphere of functional possibilities and practical application
of these devices [20].
Figure 2.6- Schematic of loop heat pipe [21]
2.3.2.2 Classical Heat Pipe Limitations
There exist several limitations which can diminish heat transfer in classical heat pipes.
Some of the less significant limitations include the vapour pressure limit, the sonic limit,
the flooding limit and the dry-out limit. But there are two main limitations: film boiling
and liquid entrainment.
[19]
Film Boiling Limitation
To better understand the film boiling phenomenon, a brief introduction about boiling heat
transfer is necessary. Boiling or convection with a phase change is defined as evaporation
occurring at a solid-liquid interface. Utilizing the latent heat of vaporization as a means to
store and transfer heat can greatly enhance heat transfer. For instance, one can store 405
Joules in a single gram of water as sensible heat by increasing its temperature from 0°C to
100°C while by boiling the same gram of water at 100°C, one can store 2260 Joules.
Nukiyama identified the existence of several boiling regimes, namely natural convection
boiling, nucleate boiling and film boiling [22]. His experiment consisted of passing an
electrical current through a platinum wire immersed in water at atmospheric conditions.
By measuring the power through the wire, the temperature of the wire (T w ) and the
temperature of the water which was maintained at the saturation temperature (T sat ), the
classical boiling curve shown in Figure 2.7 has been constructed. This excess temperature
differential (T w -T sat ) represents the driving force for the boiling. If the wall temperature
is less than the saturation temperature, boiling will not occur. However, if the wall
temperature exceeds the saturation of the liquid, boiling heat transfer will be dominant.
This curve delineated by points A to E describes pool boiling. The curve between A and
B represents natural convention boiling, where bubbles are being formed on isolated spots
and fluid mixing occurs principally by natural convection. The curve between points B
and C shows nucleate boiling, where the surface becomes densely populated with
bubbles, and bubble separation induces considerable mixing. Point C is the critical heat
flux (CHF), beyond which film boiling is initiated. From point C to point D, partial film
boiling occurs, whereas the section between point D and point E represents fully
[20]
developed film boiling. The reason why heat transfer increases in this region is related to
the fact that radiation across the vapour film dominates. Given that radiative heat transfer
increases by the fourth power of temperature, heat flux from the wall increases in the film
boiling regime.
Figure 2.7- Natural and forced convection boiling curves [23]
Film boiling occurs when the vapour bubble density at the solid/fluid interface is large
enough to cause coalescence. This forms a stable vapour film, which encompasses a large
thermal resistance and temporarily reduces heat transfer. During this time, the applied
heat load is not being completely removed from the solid. Equation 2-2 can be used to
predict the critical heat flux for nucleate pool boiling [24, 25, 26].
[21]
q″ CHF = 0.149Hfg
ρ v [σg(ρ l -ρ v )/ρ 2
v ] 4/1 (2-2)
where
q″ CHF is the critical heat flux (W/m2
)
Hfg
is the latent heat of vaporization of the fluid (J/kg)
ρ v and ρ l are the densities of the vapour and liquid, respectively (kg/m3)
σ is the surface tension (N/m)
g is the gravitational acceleration normal to the surface (m/s2
)
Assuming that ρ v is much smaller than ρ l , equation 2-2 implies that the critical heat flux
increases with:
a) fluids with larger values of Hfg
b) larger vapour densities (at larger pressures)
c) larger body forces normal to the solid surface
Heat exchangers which utilize boiling to transfer heat are operated in the nucleate boiling
regime. In the case where the heat flux is beyond the CHF of the heat exchanger, film
boiling results in surface temperature, T w , which leads to failure. Therefore, using boiling
as a heat transfer mechanism is a double-edged sword: although very large heat transfer
coefficients are attainable, we have to be careful not to attain the critical heat flux and
induce film boiling.
In a classical heat pipe, the wick is the primary means by which liquid working fluid
refluxes the evaporator. However, at elevated heat fluxes, vapour formed at the liquid/gas
[22]
interface can expel liquid working fluid from the wick. The porous nature of the wick can
effectively trap the vapour, stabilizing a vapour film, and enhancing film boiling.
The Entrainment Limitation
The flow of fluid within conventional heat pipes is counter-current meaning the vapour
travels from the evaporator to the condenser, and liquid travels from the condenser to the
evaporator. Therefore, a shear force exists at the free liquid/vapour interface. The
entrainment limitation occurs at large vapour velocities where a significant portion of the
returning liquid can be entrained into the vapour stream, causing evaporator dry-out.
The limitations of classical heat pipes have impeded the realization of boiling‟s full
potential as a heat transfer mechanism, and thus heat pipes have only been successful in
applications with relatively small heat fluxes and heat loads. These limitations have
effectively prevented the metallurgical industry from using heat pipes as heat exchangers.
2.3.3 McGill Heat Pipe
2.3.3.1 Development
Recent research at McGill University has led to the development of a newly designed heat
pipe, referred to in the literature as a “McGill heat pipe” [27, 28, 29, 30, 31]. This type of
heat pipe is capable of overcoming the traditional limitations encountered at large heat
loads and heat fluxes by incorporating two features in the design [32]. In this report, the
term “heat pipe” refers to the novel design developed at McGill University.
[23]
2.3.3.2 Design and Operating Principles of the McGill Heat Pipe
Similar to classical heat pipes, McGill heat pipes consist of three parts: evaporator,
condenser and a working fluid contained within. Yet by incorporating two features they
are distinguished from the conventional heat pipes. These two features are as follows:
1. Flow modifier
2. Return line
Flow Modifier
The flow modifier imposes a resistance on the rising fluid in the evaporator, causing it to
rotate (Figure 2.8). It is believed that the resulting helical flow propels the denser liquid
against the inner evaporator wall due to the centrifugal forces generated, which ensures
homogeneous wetting and stabilizes the annular two-phase flow regime. In classical heat
pipes, capillary wicks serve as a way to distribute liquid homogenously, hence by adding
this feature, the wick which would be considered as a resistance, can be eliminated from
the heat pipe design [33, 34]. Moreover, the flow modifier is able to break any vapour
film which may form and thus can increase heat transfer dramatically. Measurements
with a water based heat pipe have shown that the critical heat flux can be greatly
increased. If film boiling is established in a water based heat pipe, a representative value
of the rate of heat extraction is about 80 kW/m2 when the heat pipe is operated at about
100oC while a McGill heat pipe operated at 100
oC can extract more than 2,500 kW/m
2
and still not be subjected to film boiling. Thus, it is clear that a swirled flow is of major
importance [11].
[24]
Figure 2.8- An example of a flow modifier [11]
Return Line
By eliminating the capillary wick, the McGill heat pipe relies solely on gravity to return
the condensed working fluid to the evaporator. Therefore, the condenser section of this
heat pipe must be located above the evaporator section. The return of liquid working fluid
is done through a pipe referred to as the return line.
The role of the return line is to provide a separate path through which the liquid can travel
down to the bottom of the evaporator. This liquid can be returned to the evaporator
because of the pressure head between the reservoir and the return line discharge without
any entrainment from the rising vapor. It is important to mention that the reservoir is the
section at the bottom of the condenser where condensed working fluid collects [33].
Working Fluid
The working fluid used in conventional heat pipes is usually either water or sodium,
based on (1) their favourable properties suited for high heat fluxes, and (2) operating
temperatures which fall within those desired for metallurgical applications. In addition to
the mentioned changes of the design of a conventional heat pipe, the viable working
[25]
substances were expanded to include sulphur. Ever since the early days of heat pipe
development, researchers have sought to use sulphur as a working substance. The reason
for this was that there are a number of working substances that can operate at low
temperatures such as water with a normal boiling point of 100°C while for high
temperature applications we have the alkali metals, and, in particular, sodium with a
normal boiling point of 880°C. Therefore, there is quite a large gap between the two.
The element, sulphur, is a material that is ideally suited to fill this gap. It has a normal
boiling point of about 445oC. For 40 years researchers have tried to incorporate sulphur in
conventional heat pipes. However, success has been limited because of the unique
viscosity curve of liquid sulphur. Sulphur at its melting point of about 115oC is relatively
fluid, however, as the liquid is heated to about 200oC, the viscosity increases dramatically
to the point that it hardly flows. As the temperature is increased further, the viscosity
drops rapidly. When the liquid is at its normal boiling point of about 445oC, the viscosity
of the liquid is similar to the viscosity it had when it melted. Additional details about
sulphur in a McGill heat pipe have been detailed by Zhao [35].
The three working substances (water, sodium and sulphur) have been used extensively in
McGill heat pipes. In comparison, water is the preferred one, sulphur is the second and
sodium follows as the third choice. Its high boiling point (880oC) implies that a sodium
pipe must operate at temperatures in excess of 800oC and possibly 900
oC. The plant trials
done with McGill heat pipe [36] have clearly demonstrated that sodium is an excellent
heat pipe working substance; however those same trials showed that a heat pipe (316L
stainless steel) which operates at such elevated temperatures will eventually deteriorate
and crack. Sulphur has a very low thermal conductivity and a relatively high viscosity.
[26]
These properties are opposite to what is required. Thus, sulphur was ranked second. The
clear winner as the most viable working substance which can be used at high heat fluxes
is water. Numerous tests for a number of applications with the patented design have been
conducted and considering the results we have concluded that with this design, water is
the preferred working substance, especially at high heat fluxes.
2.3.3.3 Choking Limitation in the McGill Heat Pipe
Because it is very difficult to get visual observations or precise flow measurements within
a heat pipe, concepts such as choking (dry-out) are still not completely understood. It is
generally accepted that the condenser has the lowest pressure as it is cooled and vapour
condenses there. On the other hand, the evaporator has the highest pressure as vapour if
formed there. Therefore, there exists a pressure difference between the evaporator (high)
and the condenser (low). Flow of liquid through the return line is primarily due to the
head of liquid working fluid that fills the return line, form the point of discharge in the
bottom of the evaporator to the reservoir in the condenser. The driving force must
overcome the pressure difference between the evaporator and the condenser such that
sufficient condensed working fluid flows down the return line and is discharged into the
evaporator, cooling it by vaporizing. Under high heat loads and low operating
temperatures (causing low vapour densities), the vapour velocities within a heat pipe are
rather large. With large velocities, it is thought that the flow resistance of vapour as it
rises to the condenser can be quite high, as flow resistance is generally found to be
proportional to the fluid velocity squared. The high flow resistance causes a large
[27]
pressure gradient in the evaporator, resulting in a relatively high pressure at the bottom of
the evaporator. Therefore, at the onset of choking, the large flow resistance caused by the
high vapour velocities accounts for a pressure drop that exceeds the hydrostatic pressure
of the liquid column in the return line. This slows the flow of liquid down the return line
such that an insufficient flow rate of liquid working fluid is discharged into the evaporator
and as a result the evaporator dries out. This phenomenon was first observed by Yuan in a
laboratory test reported in 2002 (Figure 2.9) [37].
Figure 2.9- (A) McGill heat pipe after 2 hour test, stable operation, (B) McGill heat pipe after 1 minute
test, conditions for choking limitation [37]
2.3.3.4 McGill Heat Pipe Applications
There are three groups of applications using water as the working substance that will be
briefly reviewed here. The fourth application is the objective of our project which will be
highlighted after reviewing the other 3 applications.
The 3 applications are as follows:
1) Taphole cooling blocks
[28]
2) Slag launder
3) Cooling for blast furnace refractory
Taphole Cooling Block
The taphole of a lead blast furnace is simply a hole that runs through the copper block and
which is cooled extensively by water. In a conventional copper block, passages for water
are embedded within the block. In this case, a crack in the copper can lead to a water leak
and a subsequent explosion. For this reason, an alternative copper block that is cooled by
two symmetrical heat pipes was designed and tested to determine if it could be a viable
substitute for the conventional block [38]. Schematics of the conventional water cooled
block and the modified heat pipe cooled block are shown in Figures 2.10a and 2.10 b.
(a) (b)
Figure 2.10- (a) Conventional copper block with water cooled passages and (b) Heat pipe cooled copper
block
[29]
Figure 2.11- Copper taphole block under intense heat loading from an oxygen/natural gas burner
The specifications were such that the conventional block extracts about 300 kW and this
heat is carried away by the cooling water. McGill heat pipe using water as the working
substance was to accomplish this rate of heat extraction. Moreover, the condenser of each
heat pipe was also water cooled. The trials were successful and clearly demonstrated the
capabilities of the McGill heat pipe under extreme heat loading conditions that exceeded
2 MW/m2. Shown in Figure 2.11 is a photo (and corresponding schematic) of the copper
block under the intense heat loading generated by an oxygen/natural gas burner.
[30]
Figure 2.12-Slag launder (copper)
Slag Lauder
When the slag is tapped from a lead blast furnace it flows over a runner and into another
vessel for processing. The focus of the application to be described is the runner (Figure
2.12). Prior to our study, the runner was made by creating a steel trough which was lined
with refractory. While this application was not deemed critical as was the previous one
with the taphole, it was nonetheless carried out to study the viability of using three
horizontal pipes and of connecting multiple evaporators to one condenser. More details
about this application are reported in [38].
The runner was subjected to intense heating from an oxygen/natural gas burner for
extended periods of time (days) (Figure 2.13).The runner was modeled extensively with
Fluent software. The temperature distribution within the copper at steady state was
computed and is shown for a typical run in Figure 2.14. One sees that the intense heat
extraction creates sizeable temperature gradients within the copper.
[31]
Figure 2.13- Copper launder under intense heat load Figure 2.14- Temperature mapping of
launder
Cooling for Blast Furnace Refractory
The iron blast furnace requires cooling of the tuyeres and the refractory to maintain a long
and stable operation. One of the current methods of cooling the refractory is to use what
are referred to as „cigar‟ coolers. These are copper pipes several inches in diameter and
about 2 feet in length which are embedded in the refractory at numerous locations around
the hearth of the furnace. Each of these is water cooled. In the event of failure, the
situation can be dangerous. In addition, it can happen that water leaks into the furnace and
freezes some of the melt. This can cause problems related to furnace productivity. As an
alternative to water cooled units, a new heat pipe based cooler has been designed, built
and tested successfully [11].
2.3.4 Heat Pipe Probe for a Tundish
Before discussing the use of heat pipes in the steel tundish, it is instructive to briefly
review the testing of a water based heat pipe in molten aluminum that has been carried
out at McGill University [37]. In Figure 2.15a the evaporator of the heat pipe which is
immersed into the molten aluminum is shown. Typically, the heat flux on the heat pipe
[32]
surface was about 2 MW/m2. Moreover, the uniformity of the heat flux is demonstrated
by examining the frozen Al shell on the pipe. Since the outer surface of the shell
represents an isotherm, a smooth uniform contour of the outer shell surface indicates
uniform heat extraction from the inner surface of the pipe (Figure 2.15b).
Figure 2.15- (a) water based heat pipe immersed in Aluminum, (b) the view of heat pipe after it was
removed from the Al melt.
The research with molten aluminum showed that the water based heat pipe can extract 2
MW/m2. To put this flux in perspective it should be noted that a typical DC casting mold
for aluminum extracts heat at a peak heat flux of about 1 MW/m2 at the top of the mold.
As stated, the application we are now focused on is the use of a water heat pipe in molten
steel. The most visible difference between the 2 configurations is the melt temperature. In
the aluminum tests the melt temperature was about 800oC, however, in the steel tests the
melt will be in excess of 1500oC. While we tend to focus on temperature, it is the heat
flux that the heat pipe is most concerned with. Our estimate based on classical heat
transfer shows that the anticipated heat flux in the steel system will be around 1 MW/m2.
[33]
This is about ½ of the rate the heat pipe has been operated at in the lab. In addition, one
should recall that the heat pipe in the taphole copper block extracted heat at peak fluxes of
about 2.5 MW/m2. Thus, we are confident that the water based heat pipe which is
immersed in a molten steel environment is not only viable but very close to being a
reality. Once the stability of McGill heat pipe in molten steel tundish is approved, it can
be used as the viewing port with which one can measure the tundish temperature, slag
thickness and composition of the melt using the methods explained earlier in the strategy
section (2.2).
[34]
Chapter 3
Experimental Procedure
This section consists of the following:
1) The design and fabrication of the heat pipe
2) Experimental setup
3.1 THE DESIGN AND FABRICATION OF THE HEAT PIPE
As explained in Chapter 2, the McGill heat pipe is equipped with two new features: flow
modifier and liquid return line. Therefore, this heat pipe which has been granted US
patent 7,115,227 does not have wicks that are used in conventional heat pipes and it only
relies on the flow modifier to distribute the liquid homogenously [40]. Consequently, this
design consists of two parts:
1) Evaporator
2) Condenser
3.1.1 Evaporator
The evaporator is an annular tube which has been fabricated at IMI-NRC in Boucherville,
Québec. It is made of stainless steel 316L due to its good strength and corrosion
resistance. The bigger pipe has a diameter of 7.6 cm with 6 mm thick wall for safety
reasons while the smaller pipe has a diameter of 2.5 cm with a 3 mm thick wall. The pipe
[35]
is designed to be annular to serve as a viewing port for various measuring devices.
Moreover, it is used to transfer an inert gas such as argon into the bath in order to prevent
the molten metal or slag from blocking the viewing tube. It is important to mention that
all the designs have been drawn using Solid Works software.
Flow Modifier
One of the additional features in the McGill heat pipe is the flow modifier which has been
sized based on past experiences (Figure 3.1). The outside diameter of this flow modifier
is 6mm and it has a 3.8 cm pitch.
Figure 3.1- Flow modifier used in McGill heat pipe
Return Line
The second new feature of the McGill heat pipe are the two return lines in the evaporator
with an outside diameter of 1.2 cm. They are approximately 77 cm long and at 10 cm
from the top of the evaporator they are connected to the related hoses to bring the liquid
from the condenser to the pipe. On each return line, a valve is installed in order to be able
Viewing tube Flow modifier
Turns: 22
Outside diameter:
6 mm
Pitch: 3.8 cm
[36]
to effectively turn off the heat pipe. By shutting off the valve, the flow of liquid to the
evaporator can be stopped and the evaporator eventually dries out. In the absence of
liquid working fluid in the evaporator, no cooling would take place.
Vapour Line
There are two vapour paths to bring the generated vapour to the condenser. These holes
which have the inside diameter of 1.2 cm are connected to the related hoses at 3 cm from
the top of the evaporator. The schematic of the evaporator is shown in Figure 3.2 and 3.3.
Figure 3.2- Solid Works drawing of the evaporator
Return Line
Vapour Path
Flow Modifier
Thermocouples
[37]
Figure 3.3- The top and side view of the evaporator
Bottom Part (Nozzle)
The most critical part of the evaporator is its 6cm bottom part which is inserted into the
molten metal. Precision in the design and fabrication of this nozzle which is made of
stainless steel 316L has been considered. If the heat pipe works properly, a layer of metal
will freeze on the nozzle. However, if the thickness of the frozen material exceeds a
certain amount, it may block the viewing port and interfere with the continuous
measurement process. Therefore, the nozzle is designed to be tapered as shown in Figures
3.4 and 3.5. In Figure 3.6 the assembled evaporator is shown.
[38]
Figure 3.4- Nozzle in the evaporator from different angles
(a) (b)
Figure 3.5- Different views of fabricated nozzle: (a) inside view, (b) outside view
[39]
Figure 3.6- Assembled Evaporator
3.1.2 Condenser
The condenser of the heat pipe which has been described in Chapter 2 has a lower heat
flux than the evaporator due to its distance from the heat source. Therefore, it is made of
stainless steel 304. It is a cylinder with 60 cm outside diameter and height which has been
fabricated along with the evaporator at IMI-NRC in Boucherville, Québec. The system is
water cooled, thus, the condenser is equipped with 4 copper coils to bring the cooling
water to the device. The coils are 50 cm long with 10 turns and approximately 1cm
outside diameter (3/8 inch). Moreover, there are 2 vapour pipes (30cm long and 1.25 cm
inside diameter) which bring the vapour from the evaporator to the hoses and into the
condenser (Figure 3.7). The vapour is condensed by the water coils and the liquid is
collected at the bottom of the condenser. There are two holes to bring the liquid to the
evaporator. Last but not least feature in the condenser is the vent which is used to
minimize the possibility of blocking the liquid return pipe by vapour.
[40]
Working Fluid
Water was chosen as the working fluid in this project. Previous work has shown that
water can dissipate a heat flux in excess of 1.5 MW/m2
without experiencing film boiling
[37]. As explained in Chapter 2, sodium could not be used due to its high operating
temperature (>800ºC). In addition, sodium is flammable and requires special startup
procedures as it is solid at room temperature.
The bottom of the condenser which is called the reservoir was charged with 5 litres of
water. This quantity of water is sufficient to completely fill the evaporator, return line,
and hoses linking the evaporator to the condenser. This ensures that the evaporator walls
are completely wetted when starting a test.
All the hoses used in the system are Swagelok products and they are made of stainless
steel inside and outside. On top of the condenser, a pressure relief valve was installed for
safety issues. Moreover, there is a pressure gauge transmitter which enables us to read
pressure at any time during the process.
The fabricated heat pipe was first tested in a gas furnace as demonstrated in Figure 3.8.
[41]
(a)
(b)
Figure 3.7- (a) Schematic of the condenser, (b) Fabricated condenser
Vent
Cooling Coil
Vapour Pipe
Reservoir
Liquid Path
Pressure Gauge Transmitter
Pressure Relief Valve
Vent
[42]
Figure 3.8- Heat pipe during an experiment in gas furnace
3.2 EXPERIMENTAL SETUP
The ultimate target of this project is to insert the newly designed heat pipe into molten
steel to serve as a viewing tube for continuous measurements of temperature, slag
thickness and composition of the melt. However, before inserting the stainless steel
device into the molten steel, we have to have a thorough knowledge of its functionality
and behaviour in different environments. Therefore, considering the three main criteria of
safety, temperature and heat flux, we have formulated a 4 step experimental program to
pursue this target. The test program comprises:
Condenser
Hoses
Evaporator
[43]
a) Gas furnace test
b) Aluminum test
c) Cast iron test
d) Steel test
From the safety aspect, inserting the heat pipe in the gas furnace is the safest process of
the 4. The maximum temperature of the gas furnace is around 1400°C, yet considering the
relatively low heat flux of the gas on the outer wall of the heat pipe, this step is the safest
one. The second safest procedure is the test in molten aluminum wherein the temperature
of the furnace is around 700°C while the third is the test in cast iron with a melting
temperature of approximately 1100°C. The least safe procedure is the steel test, the metal
of which the device is made and that has a high melting temperature of about 1500°C.
The last but not least criterion to be discussed is „heat flux‟. Heat flux is the flow of
energy per unit of area per unit of time. As described in Chapter 2, film boiling which is
considered as the main limitation in heat pipe technology plays an important role.
Referring to Figure 2.8, film boiling happens due to large vapour density at solid/liquid
interface which leads to heat flux decrease despite the increase in temperature differential
(T w -T sat ). The critical heat flux, the peak where film boiling occurs, depends on the heat
pipe configuration and design. In this project we are experimenting with a newly designed
heat pipe in which the critical heat flux cannot be determined except with experiments. If
film boiling takes place, heat extraction due to poor heat flux would decrease
substantially and the temperature of the heat pipe would increase. If the environment
where heat pipe is exposed to has high temperature as the steel furnace for instance, the
heat pipe temperature may exceed 1500°C and the device will melt. Therefore, finding
[44]
the critical heat flux for the current design before the molten steel test is a must. Our
estimate based on classical heat transfer shows that the anticipated heat flux in the steel
system will be around 1 MW/m2
. Among the mentioned steps, molten aluminum which
has the highest heat flux of 2 MW/m2 can confirm the durability of heat pipe in the
molten steel furnace. In all 3 steps (gas, molten aluminum and cast iron), we are in search
of critical heat flux in a safe and controlled environment. On the other hand in the steel
furnace we are expecting to extract 30 kW in the molten steel furnace and thus the
capability of the design to extract this amount of heat along with critical heat flux in each
step is being verified.
In molten metal furnaces the heat pipe due to high heat extraction works at a temperature
lower than metal‟s freezing point which then leads to the formation of a frozen layer of
molten metal on the evaporator. This layer which has different thicknesses depending on
the heat conductivity of the metals would form up to the point where its outer surface
experiences the temperature more than metal‟s melting point. There, the liquid would no
longer solidify. This phenomenon is shown in Figure 2.15.
[45]
3.2.1 Sensors
In order to measure the amount of heat being extracted and to calculate the corresponding
heat flux in the gas furnace, two thermocouple wells are incorporated in the system: 1 in
the evaporator and 1 in the condenser. In the evaporator two thermocouples are inserted
at different depths, yielding the temperature of the bottom and 0.6 m from the bottom of
the evaporator (almost where the evaporator is inserted in the furnace). However, in the
condenser the two thermocouples report the temperatures of the bottom and the mid-point
of the condenser. In phase 2, in the aluminum furnace, another thermocouple is added on
the wall of the evaporator to report the temperature of the frozen aluminum formed on the
outside of the evaporator.
Using enthalpy balance formula, we can easily measure the amount of heat picked up by
the heat pipe (Figure 3.9, Equation 3-1)
(a) (b)
Figure 3.9- Equipments for heat measurement, (a) container for water flow rate measurement, (b) digital
thermometer
[46]
p (3-1)
where
is the extracted heat (W)
is the water flow rate (g/s)
p
is the specific heat capacity (J/(g·K) (4.2 for water)
is the temperature rise of the cooling water (°C)
However, when the heat pipe is running at high temperatures, heat is being lost from the
condenser‟s wall which is not insulated, from the vapour formed in plastic hoses along
with hot water which is neglected in measurements and etc. Therefore, the amount we are
measuring based on the cooling water is not the exact amount of heat that the heat pipe
has picked up from the furnace. There must be a solution to compensate for the heat
losses. All the solutions and results of the heat pipe experiments in gas furnace and
molten aluminum test will be presented in next chapter.
[47]
Chapter 4
Results and Discussions
This chapter consists of two parts: part 1 which describes the results of gas furnace tests
and part 2 where the results of molten aluminum tests are presented.
Part 1: GAS FURNACE TESTS
Newly designed heat pipe has been tested in a gas furnace at McGill University over a
number of days from January 10th
to April 10th
, 2010. The first experiments were
preliminary tests which explored the operability of the heat pipe while the rest aimed to
determine the maximum heat extraction capacity and to verify the film boiling
phenomenon. It is important to mention that LABVIEW software has been used to record
the behaviour of the heat pipe through the experiments.
The tests in the gas furnace were divided into the following 3 segments:
a) Gas furnace temperature at approximately 730 °C (Level 1)
b) Gas furnace temperature at approximately 1370 °C (Level 2 )
c) Gas furnace temperature at approximately 1480 °C (Level 3)
In each setup, there are 3 steps in order to calculate the total extracted heat. To have
reliable information, the measurements have to be taken when the experiment is
stabilized, and thus reaching steady state is the first step in each measurement. Second, as
[48]
described earlier, the amount of extracted heat should be measured. The third step is to
account for the heat losses from the condenser, etc...
4.1 REACHING STEADY STATE
When the furnace is turned on, the evaporator is exposed to the heat source generated in
the furnace. Consequently, its temperature increases and the water inside the pipe
evaporates. The generated vapour will travel through the hoses and enter the condenser
environment where at the beginning it is at the room temperature. As a result, the
temperature of the condenser starts to increase but with slower speed than the evaporator
due to its distance from the heat source. Therefore, there will be a larger temperature
differential between the condenser and the evaporator at the beginning of the test (Figure
4.1). In this phase which is called start-up, the change of liquid to vapour in the
evaporator and vapour to liquid in the condenser does not happen uniformly because of
the instability of the system. Figure 4.1 presents the temperature distribution of the heat
pipe system versus time through a 3 hour experiment of heat pipe in gas furnace level 1
meaning a furnace temperature around 700°C. As explained earlier there are 4
thermocouples inserted in the evaporator and the condenser. In the graphs evap. B
presents the thermocouple at the bottom of the evaporator while evap. M shows the
temperature at 0.6 m from the bottom. In a lower temperature range, cond. B stands for
the thermocouple at the bottom of the condenser and cond. M demonstrates the
temperature at 0.3 m from the bottom of the condenser. In the start-up phase, as it is seen
[49]
in Figure 4.1, the system is experiencing an unstable situation where the temperatures in
both evaporator and condenser change substantially in short periods of time.
Figure 4.1- Heat pipe experiment in gas furnace level 1, temperature vs. time
. The thermocouples at the bottom of the evaporator and condenser where the liquid water
is collected show more instability than the other two thermocouples. In the start-up phase,
in the evaporator, water evaporates and is replaced by the liquid coming from the
condenser irregularly. On the other hand, in the condenser, there is some liquid left in the
reservoir when the condensed vapour is added to this volume of liquid. This volume of
liquid has different temperature and that justifies the oscillation in green thermocouple.
Time wise, all these phenomena occur so fast that the system cannot have the expected
uniform cycle.
0
10
20
30
40
50
60
70
80
90
1001
50
1
10
01
15
01
20
01
25
01
30
01
35
01
40
01
45
01
50
01
55
01
60
01
65
01
70
01
75
01
80
01
85
01
90
01
95
01
10
00
1
10
50
1
evap B
evap M
Cond B
Cond M
Time
(s)
Tem
per
ature
(°C
)
Cooling Start-up Increase Cooling
[50]
Another important point to mention is the temperature in red and velvet thermocouples
which mostly report higher temperatures than the other two. In the evaporator, the liquid
at the bottom evaporates and the flow modifier rotates the formed vapour all the way up
the vapour path. At the bottom we have the combination of liquid coming from the
condenser with a lower temperature and vapour while at the position of evap. M, there is
only a large density of generated vapour. Because of some undercooling in the reservoir
pool in the condenser, the returning liquid can be at a temperature below the saturation
point. Thus, the temperature at evap B can be slightly below that at evap M because the
fluid will have reheated to saturation.
Similarly, in the condenser cond. M is hotter than the cond. B. This thermocouple has the
same elevation as the vapour pipes which bring the hot vapour into the condenser.
Comparing to the bottom of the heat exchanger where the liquid stays with lower
temperature, higher temperature in the mentioned location is predictable.
Figure 4.2 shows the heat pipe behaviour in the gas furnace operating at level 2 meaning
an environment temperature of 1350°C. In this setup, the temperature of both evaporator
and condenser increases faster due to the hotter environment which the evaporator is
exposed to. Consequently the cycle of phase change (liquid to vapour and vice versa)
occurs earlier in comparison to level 1. As a result, despite a larger instability that the heat
pipe experiences, this phase takes less time than level 1 and the system reaches steady
state earlier.
In steady state where all four thermocouples follow a very smooth and stable trend, we
modified the situation by adding some cooling water. As seen in Figure 4.2, by increasing
[51]
cooling water flow the temperature in both evaporator and condenser decreases in a
uniform manner. Surprisingly enough, the temperature in the system does not exceed
90°C while the evaporator is in a 1350°C environment.
Figure 4.2- Heat pipe experiment in gas furnace level 2, temperature vs. time
In level 3, the heat pipe is experiencing an environment temperature of 1500°C. Figure
4.3 shows the behaviour of heat pipe towards this environment. It is important to mention
that in these graphs, level 3 comparing to level 2 has a larger start-up phase. This is due to
the fact that the data in level 2 has been captured when the furnace was on for an hour.
Therefore, the heat pipe has taken less time to reach steady state in level 2 than in level 3.
However, in an equal condition level 3 needs less time than level 2 to reach steady state.
0
10
20
30
40
50
60
70
80
90
100
1
50
1
10
01
15
01
20
01
25
01
30
01
35
01
40
01
45
01
50
01
55
01
60
01
65
01
70
01
75
01
80
01
85
01
90
01
95
01
10
00
1
10
50
1
11
00
1
evap B
evap M
cond B
cond M
Time
(s)
Tem
per
ature
(°C
)
Cooling
Start-up
Increase more cooling Increase cooling
[52]
Figure 4.3- Heat pipe experiment in gas furnace level 3, temperature vs. time
4.2 MEASURING THE EXTRACTED HEAT
Each furnace setup allows the heat pipe to extract a certain amount of heat and
consequently heat flux. Our target in this step is to measure the heat that the heat pipe has
captured at each furnace level. To pursue this target, we have measured the extracted heat
with the process described in Chapter 3 for at least 5 points in each setup.
In each furnace level, despite the fact that all the measurements have been done in the
same furnace temperature, the results were different due to heat losses primarily from the
condenser. There are several sources of heat loss in this system. For instance, the wall of
the condenser is not insulated and is exposed to an ambient temperature of approximately
0
10
20
30
40
50
60
70
80
90
100
110
1
50
1
10
01
15
01
20
01
25
01
30
01
35
01
40
01
45
01
50
01
55
01
60
01
65
01
70
01
75
01
80
01
85
01
90
01
95
01
10
00
1
10
50
1
11
00
1
11
50
1
evap B
evap M
cond B
cond M
Time
(s)
Start-up
Tem
per
ature
(°C
)
Cooling
[53]
25°C and this can be considered as a source of heat loss. In addition, in high operating
temperatures, the air trapped in the hot water coming from the condenser would evaporate
and release some vapour which its energy is neglected in the measurements. Moreover,
there is some heat being lost in the hoses and plastic tubes. Therefore, we can conclude
that the measured heat does not present the heat which the system has picked up in each
furnace setup. Consequently, we should find a solution to compensate for the heat losses.
4.3 COMPENSATING FOR HEAT LOSSES
Figure 4.4 shows the curve of extracted heat versus the condenser temperature in level 1.
The dashed red line extrapolates the curve to 25°C which is defined as room temperature.
If the temperature difference between the heat pipe system and the environment where the
experiment is being done decreases, the heat losses would be minimized. Thus, we
assume that the condenser has the room temperature as the environment to eliminate the
heat losses from condenser, hoses and vapour formed in plastic tubes. However, in reality
this condition does not occur. With this assumption we can declare that the amount of
heat captured from this method can represent the exact amount of heat extracted from the
furnace by the evaporator.
Solving equation 4-1by inserting 25 for the x axis will lead to calculate the exact amount
of extracted heat, q , in level 1.
(4-1)
W
[54]
In furnace setup level 1with approximately 700°C environment, the heat pipe system is
capable of extracting 3940 W from the furnace. By calculating the active area of the
evaporator in the furnace, we can find out the correspondent heat flux, " (equation 4-2,
4-3).
active (4-2)
active=0.4 cm
"avg
=
kW/m
2 (4-3)
Knowledge of total extracted heat aids to obtain more information regarding volumetric
flow rate and vapour velocity which is presented in Table 4.1.
Evap. B
Temp.
(°C)
Evap.M
Temp.
(°C)
Cond.B
Temp.
(°C)
Cond.M
Temp.
(°C)
Water
Temp.
(°C)
Pressure
(atm)
Q water
(W)
Q lost
(W)
Volumetric
flow rate (1)
(m3
/s)
(2)
(m/s)
81
81.3
71.5
72.5
60.6
0.50
1422
2518
0.005
11.3
7 3
73.3
59.6
60.3
55
0.37
1864
2076
0.007
15.9
70
71
56
56
53.5
0.33
2205
1735
0.008
18.8
68
69
53.5
52.5
51.5
0.30
2802
1138
0.009
20.4
62
63.5
44
41.5
43.5
0.24
2845
1095
0.011
25.0
59.5
60.5
38.5
33.5
37
0.22
3348
592
0.012
27.2
Table 4.1- A set of data captured in furnace setup level 1
(1) Volumetric flow rate of water vapour
(2) Velocity of water vapour at the discharge
0.042 m2
[55]
Figure 4.4- The curve of extracted heat vs. condenser temperature in gas furnace setup level 1
The first four columns in Table 4.1 reflect the behaviour of 4 thermocouples used in the
system while the fifth column shows the temperature of the cooling water measured for
each data point using a digital thermometer. The fifth column demonstrates the pressure
in the condenser revealed on the pressure gauge screen. The pressure inside the
evaporator cannot be easily obtained due to the fact that the swirler inside the evaporator
causes vortex inside the environment which leads to a dynamic behaviour and therefore
we cannot detect its pressure directly from temperature as in equilibrium status.
Unfortunately, at this moment we do not have enough information to explain the
phenomenon in more detail.
By knowing the total extracted heat and the heat captured from the cooling water, one can
calculate the heat loss for each data point. As mentioned earlier, the heat loss decreases
when the temperature of the system decreases and this statement is justified in column 8
of Table 4.1.
The last column in table 4.1 shows the vapour velocity in the heat pipe discharge. The
heat pipe is a sealed system and this property challenges the researcher to discover all the
y = -53.908x + 5289
0
1000
2000
3000
4000
0 20 40 60 80
q w
ater
(W
)
Temperature
(°C)25
[56]
phenomena that occur in this technology. One of these phenomena is choking or dry-out
which has been explained earlier in Chapter 2 and is directly related to vapour velocity.
Although it is difficult to measure this term, yet using ideal gas law one can calculate the
vapour velocity in each furnace setup (equation 4-4)
(4-4)
Where
is the pressure of the system (Pa)
is the volumetric flow rate (m3/3)
is the rate of vapour formation per time (g/s)
is the gas constant (J/molK) (8.314 J/molK)
is the average temperature in the evaporator (K)
If we accept that the heat captured from the heat pipe vaporises the liquid inside the
evaporator and also consider that for 1 g of liquid water which is 1/18 mole, 2260 kJ/kg
as a latent heat of vaporization is needed, we can easily calculate the quantity of vapour
( being formed in each furnace setup.
Once we obtain the volumetric flow rate from equation 4-4, by solving equation 4-5 we
can calculate the vapour velocity.
* (4-5)
[57]
* is the area of the hose discharge and is 4
m2
in the design.
The results of the experiment in furnace setup level 2 where the environment has the
temperature of 1370° are demonstrated in Figure 4.5 and Table 4.2.
Figure 4.5- The curve of extracted heat vs. condenser temperature in gas furnace setup level 2
Equations 4-6 and 4-7 show that the heat pipe has extracted a substantially greater amount
of heat when compared to level 1.
(4-6)
W
And
"avg
= kW/m2
(4-7)
The vapour velocity in this setup exceeds 100 m/s meaning 360 km/hr which reflects the
intense environment that the heat pipe experiences. It is an accomplishment to extract
18.5 kW with the maximum operating temperature of 80°C. More important, in the first
two setups none of the limitations of film boiling and choking have been detected.
y = -172.15x + 22901
0
5000
10000
15000
20000
0 20 40 60 80
Temperature
(°C) 25
q w
ater
(W
)
[58]
Evap. B
Temp.
(°C)
Evap.M
Temp.
(°C)
Cond.B
Temp.
(°C)
Cond.M
Temp.
(°C)
Water
Temp.
(°C)
Pressure
(atm)
Q water
(W)
Q lost
(W)
Volumetric
flow rate (1)
(m3
/s)
(2)
(m/s)
79
80
70
69.5
69.3
0.48
10230
8370
0.027
61.3
79
79.7
68
67.5
66.5
0.47
11480
7120
0.029
65.9
75.7
76.4
62
61.5
61
0.42
13400
5200
0.031
70.4
74.9
75.7
62
61.5
61.4
0.41
11860
6740
0.032
72.7
71.1
71.8
54.6
54
54.2
0.35
12900
5700
0.037
84.0
70.7
71.4
52
51.5
50
0.34
15000
3600
0.038
86.3
69
69.7
48
47.5
44.5
0.31
15870
2730
0.041
93.1
65.9
66.7
43.5
42.5
37
0.28
14770
3830
0.045
102.2
65.7
66.8
41.5
40.5
27.3
0.28
16660
1940
0.045
102.2
Table 4.2- A set of data captured in furnace setup level 2
(1) Volumetric flow rate of water vapour
(2) Velocity of water vapour at the discharge
In furnace setup level 3, the heat pipe reached its maximum capacity for heat extraction in
the gas furnace experiments and that is 23500 kW which leads to a heat flux of 560
kW/m2 (equations 4-8 and 4-9).The results of this setup experiment is shown in Figure 4.6
and Table 4.3. It is important to mention that in molten steel furnace, we are expecting to
have a heat flux of 1000 kW/m2
. Comparing to level 3 in gas furnace, this heat flux
seems to be easily achievable unless we encounter the design limitation of film boiling. In
gas furnace experiments no film boiling has been detected otherwise, the heat extraction
due to the formation of vapour layer on the wall of the evaporator would have
substantially decreased. However, we pushed the system to the limit of dry-out in order to
[59]
verify the maximum vapour velocity which the heat pipe can handle in a safe and
controlled environment.
Evap. B
Temp.
(°C)
Evap.M
Temp.
(°C)
Cond.B
Temp.
(°C)
Cond.M
Temp.
(°C)
Water
Temp.
(°C)
Pressure
(atm)
Q water
(W)
Q lost
(W)
Volumetric
flow rate (1)
(m3
/s)
(2)
(m/s)
94.2
94.4
88.4
88.2
85.9
0.82
10850
12700
0.021
48
87.5
88
79
78.8
76.8
0.64
13500
10050
0.027
61
82.9
83.4
71.9
71.5
70.2
0.54
15330
8220
0.031
71
79.1
79.5
65.5
64.9
63.8
0.47
16615
6935
0.035
81
77.5
78.1
63.2
62.5
61.8
0.44
15550
8000
0.038
86
76.8
77.4
61.5
61
60.4
0.43
17400
6150
0.039
88
72.2
73
53.8
53.2
51.4
0.36
17950
5600
0.046
104
71.1
72.1
51.6
51
48
0.34
18310
5240
0.048
110
69.3
70.5
48.2
47.4
42.8
0.33
18960
4590
0.049
113
Table 4.3- A set of data captured in furnace setup level 3
(1) Volumetric flow rate of water vapour
(2) Velocity of water vapour at the discharge
Figure 4.6- The curve of extracted heat vs. Condenser temperature in gas furnace setup level 3
y = -189.71x + 28294
0
5000
10000
15000
20000
25000
0 10 20 30 40 50 60 70 80 90 1002
Temperature
(°C)
q w
ater
(W
)
[60]
(4-6)
W
And
"avg
= kW/m2
(4-7)
[61]
4.4 CHOKING LIMITATION (DRY-OUT)
By decreasing the operating temperature of the evaporator in the gas furnace setup level 3
which has a large heat load, the system choked as shown in Figure 4.7. As explained in
Chapter 2, the increased vapour velocities cause large pressure gradients to develop
across the evaporator and hoses. This pressure gradient opposes the pressure head
between the reservoir and return line ( ). It is the pressure head that causes the return
of liquid working fluid to the evaporator. In Figure 4.7 the thermocouple with the higher
elevation in the evaporator reports that the temperature increases substantially while
choking occurs in the system. This is because of the pressure drop which prevents liquid
from reaching the evaporator and therefore, the dry-out phenomenon starts from the top
of the evaporator while at the bottom due to the presence of some liquid water the
temperature is not changed dramatically. Figure 4.8 shows the evaporator being darker in
the middle where it dried out and silver at the bottom and this justifies the fact that the
bottom has not been hot when compared to the middle section. The pressure drop within
a pipe is generally found to be a function of the vapour velocity squared yet because of
the complexity (both geometric and mechanical) of the evaporator, very little theory is
available to predict the velocity field within it. However, using equations 4-4 and 4-5 one
can calculate the velocity that the system tolerates without drying out.
The maximum vapour velocity in this set of experiments is 103 m/s, which is equal to
about 1/3 Mach (the speed of sound). The fact that gases start to compress at this velocity
also justifies the statement that at this velocity the system chokes due to the pressure drop.
However, we have to keep in mind that this limitation is user controllable and by
[62]
decreasing cooling water flow and thus decreasing the vapour velocity we can easily
avoid the choking phenomenon.
Figure 4.7- The dry-out phenomenon in the heat pipe system in an experiment in gas furnace level 3
Figure 4.8- The evaporator after gas furnace experiments
0
50
100
150
200
250
1 101 201 301 401 501 601 701 801
evap B
evap M
cond B
cond M
Tem
per
atu
re
(°C
)
Time
(s)
[63]
Part 2: MOLTEN ALUMINUM TESTS
After studying the results of the first phase which were satisfying, we decided to insert the
McGill heat pipe into molten aluminum. The experiments took place in CMQ (Centre de
métallurgie du Québec), Trois-Rivières, in a furnace with 0.4 m diameter crucible filled
with molten aluminum 356 alloy. As described in Chapter 3, in this phase we were
expecting to experience a heat flux of 2 MW/m2
. We were investigating the capability of
the pipe to avoid the film boiling limitation. Compared to the choking limitation, film
boiling is a consequence of a design problem which cannot be controlled during an
experiment. By encountering film boiling, due to the vapour layer formed on the wall of
the evaporator, the system would not be able to extract the expected amount of heat and
thus the molten aluminum would not freeze on the wall as predicted.
In the aluminum phase of the test program, the heat pipe was preheated by a burner
before its insertion into molten aluminum. The maximum heat extracted with the burner
was 10 kW. By immersing the evaporator into the molten aluminum at 800°C, the
thermocouple in the evaporator was expected to report a relatively higher temperature
than with the burner but, Figure 4.9 does not show the predicted trend because of what is
suspected to be film boiling in the system.
The evaporator was pulled out several times to investigate the formation of frozen
material. However, due to insufficient heat extraction the system was not able to freeze
enough aluminum (Figure 4.10). The frozen material can play a role in protecting the
evaporator and the lack of the frozen coating can conclude in partial corrosion of the
evaporator which is demonstrated in Figure 4.11. The corrosion can be a result of several
[64]
parameters such as Ar flow that enhances convective mass transfer. It is important to
mention that the maximum heat extraction in the aluminum test was 11 kW meaning half
of the heat extracted in gas furnace experiment and this can be only justified with the
existence of film boiling during the process. The next section proposes a solution to
overcome this limitation.
Figure 4.9- The behaviour of the heat pipe and the condenser in aluminum bath
0
10
20
30
40
50
60
70
80
90
1 501 1001 1501 2001 2501 3001 3501 4001 4501 5001 5501 6001
evap in
cond M
cond B
Time (s)
Tem
per
ature
(°C
)
[65]
Figure 4.10- The thin frozen aluminum layer on the evaporator‟s surface
Figure 4.11- The view of corroded area on the evaporator after 2 hour experiment in aluminum bath
Corroded area
[66]
4.5 MODIFYING THE INTERNAL DESIGN OF THE EVAPORATOR
We have reviewed the design and construction of the heat pipe and concluded that the
design can be improved. We do believe it is possible to attain 2MW/m2
heat flux because
it has been done with other pipes at McGill University [37]. Thus, by improving the
design, we can hit the target heat extraction rate.
Our suggestion for the next step is to build a new evaporator. This evaporator would have
3 hoses instead of 6 meaning the elimination of one liquid return and vapour path. This
configuration along with larger flow modifier diameter (7.6 cm instead of 3.14 cm) can
give the fluid in the evaporator more space and better swirl and decrease the chance of
vapour layer formation on the wall and therefore more heat extraction.
[67]
Chapter 5
Conclusions
The newly designed McGill heat pipe has been tested repeatedly in a gas furnace for a
time frame of 4 months. This set of experiments was conducted at McGill University and
they were the first phase among the four steps formulated for the project (gas furnace,
molten aluminum test, cast iron test and molten steel test).
In order to leave the heat pipe in a steel tundish for long period of time and capture
information on temperature, slag thickness and composition on a continuous basis, the
system has to be able to extract 30 kW heat and 1 MW/m2
heat flux. In gas furnace test
the maximum heat extraction of the heat pipe is 23.5 kW with 0.56 MW/m2
heat flux. No
film boiling was detected and the results are satisfying. The operating temperature of the
heat pipe is 70°C which proves the capability of the system to extract 30kW in steel
furnace with an operating temperature less than 100°C meaning a pressure less than 1
atm. This statement takes into account some of the safety issues.
Moreover, the choking limitation is identified and the vapour velocity of maximum 103
m/s is determined to be the vapour velocity limit in the system to prevent dry-out
phenomenon. However, this limitation can be controlled by changing the water cooling
rate in the heat pipe.
The second phase is immersing the heat pipe into molten aluminum bath. This experiment
was done in CMQ (Centre de métallurgie du Québec), Trois-Rivières. The heat pipe
[68]
which was preheated by a burner before its insertion into aluminum bath could only
extract 11 kW heat from molten aluminum. This is half of the heat extraction in the gas
furnace test. Therefore, the system is not able to extract 2 Mw/m2
as predicted. This is
due to film boiling that occurs during the process and the formation of the vapour layer
which leads to lower heat transfer and thus less heat extraction. To overcome this
limitation, some design modification is suggested. The flow modifier can be made larger
in diameter and a vapour path and liquid return line can be eliminated. The improvements
in the design would give the fluid inside the evaporator more space to rotate and break the
vapour films on the wall and as a result increase the chance that the heat pipe extracts
sufficient heat to survive in molten metal baths.
[69]
References
1) Met Soc website http://www.metsoc.org/virtualtour/processes/steel
2) Gastόn, A., Sánchez Sarmiento G., and Sylvestre Begnis, J.S., “Thermal Analysis
of a Continuous Casting Tundish by an Integrated FEM Code”, Latin American
Applied Research, Argentina, 2008, pp. 259-266.
3) Shia et al., “Thermocouple Temperature Sensor and a Method of Measuring
Temperature of Molten Iron”, United State Patent, No. 5181779, 1993.
4) Barbosa et. al, “Device for Continuous Measurements of Molten Steel in the
Tundish Using Optical Fibre and Infra-red Pyrometer”, United State Patent, No.
0205480-A1, 2008.
5) Kracich, R.E., and Goodson, K., “Ladle Slag Depth Measurement”, Steelmaking
Conference Proceedings, pp. 53-60, 1996.
6) Meszaros et al., “Measuring the Thickness of Hot Slag in Steelmaking”, United
States Patents, No. 6130637, 2000.
7) Kenney, G. B., U.S. Patent, 4,578,022.
8) Jowitt, R., and Abell, I.D., U.S. Patent, 4,598,577.
9) Gruber, J., Heitz, J., Strasser, H., Bauerle, D., and Ramaseder, N., “Rapid in-situ
analysis of liquid steel by laser-induced breakdown spectroscopy”,
Spectrochimica, 2001.
10) Ocean Optics web site http://www.libsresources.com/technology/index.asp
[70]
11) Mucciardi, F., Yuan, Z., and Madani, N., “Measuring the Steel Temperature and
Slag Thickness in a Tundish on a Continuous Basis”, 12 p, 2008.
12) Tata Steel website http://www.automationtatasteel.com/products.htm
13) Peterson, G., “An Introduction to Heat Pipe Modeling, Testing and Applications.”
chapter 1, pp 1-16.
14) Gaugler, Richard (1944), Heat Transfer Devices, Dayton, Ohio: U.S. Patent Office,
pp. 4, 2350348.
15) Grover G M, Cotter T P and Erikson G F 1964 J. appl. Phys. 35 1990-1.
16) http://www.lanl.gov/orgs/esa/epe/Heat_Pipe_Site/ancient.shtml (Los Alamos
National Laboratory website)
17) Dunn, P.D., Reay, D.A., “Heat Pipes”, 4th
Edition, Elsevier Science Ltd, Oxford,
England, 1994.
18) Koepfer, C., “Drilling Dry with a Heat Pipe”, Modern Machine Shop, pp. 56-57,
May 2003.
19) http://www.alyeskapipe.com/InTheNews/MonthlyNews/2004/December/dec2004_
featurestory.asp
20) Gerasimov, Yu.F., Maydanik, G.T. Shchogolev, et al., “Low-temperature heat
pipes with separate channels for vapor and liquid”, Eng.-Phys. J. 28 (6) (1975)
957–960 (in Russian).
21) China Heat Pipe net http://www.china-heatpipe.net
[71]
22) Nukiyama, S., “The Maximum and Minimum Values of the Heat Q Transmitted
from Metal to Boiling Water under Atmospheric Pressure”, Journal of the Japanese
Society of Mechanical Engineers, 37, pp.v367-374, 1934.
23) Mucciardi, F., Zheng, G., “Heat Transfer in the Boiling Regime”, CIM Conf. of
Metallurgists Ottawa, Canada, 2000.
24) Kutateladze, S.S, “On the Transition to Film Boiling under Natural Convection”,
Kotloturbostroenie, No. 3, pp. 10-12, 1948.
25) Zuber, N., “On the Stability of Boiling Heat Transfer”, Trans. ASME, Vol. 80,
pp. 711-720, 1958.
26) Lienhard, J. H., Dhir, V.K., and Riherd, D.M., “Peak Pool Boiling Heat Flux
Measurements on Finite Horizontal Flat Plates”, Journal of Heat Transfer, 95,477-
482, 1973.
27) Navarra, P., “Heat Pipe Cooling of Metallurgical Furnace Equipment”, Ph.D.
Thesis, McGill University, Montreal, Canada, 2006.
28) Mucciardi, and Jin, N., “Extending Lance Life in Top Blowing”, Copper 99,
TMS, Oct.1999, pp. 207-222.
29) Mucciardi, F., “Improving Injection Lances with Heat Pipe Technology”, The
Brimacombe Memorial Sym.-Poster Proceedings, MetSoc-CIM, Oct. 2000, pp.
205-218.
30) Mucciardi, F., and Jin, N., “Mathematical Model of a Heat Pipe Lance”, (Paper
presented at the Computer Applications in Metallurgy and Materials Processing
Sym., CIM, 1998.
[72]
31) Mucciardi, F., and Yuan, Z., “Waterless Oxygen Lance for Steelmaking and
Refining”, Sym. on Innovation Technologies for Steel and Other Materials,
MetSoc-CIM, 2001, pp. 159-170.
32) Mucciardi, F., et al., “Heat Pipe”, U.S. Patent Applications, Field on Feb. 25,
2002.
33) Zhang, C., Mucciardi, F., and Gruzleski, J.E., “Controlled Cooling of Permanent
Moulds in the Casting of Aluminum”, Light Metals Sym. Of MetSoc, CIM, Aug.
2001, pp. 431-441.
34) Mucciardi, F., and Zheng, G., “Heat Transfer in the Boiling Regime”,
Fundamentals of Metallurgical Processing- J.M. Toguri Sym., MetSoc-CIM, Aug.
2000, pp. 335-350.
35) Zhao, H., “Development of a Sulphur-Based McGill Heat Pipe.” , Ph.D. Thesis,
McGill University, Montreal, Canada, 2007.
36) Yaun, Z., and Mucciardi, F., “Waterless, Non-Consumable Thermopump
Lances”, Ecomaterials and Ecoprocesses, MetSoc of CIM, Vancouver, 2003, pp.
229-242.
37) Yuan, Z., and Mucciardi, F., “Heat Fluxes from Aluminum Melts to Isothermal
Surfaces”, MetSoc-CIM, 2002, pp. 335-349.
38) Navarra, P., Mucciardi, F., and Van Rompaey, T., “Heat Pipe Cooling of a Slag
Tapblock”, Converter and Fire Refining Processes, EPD Congress (TMS), 2005,
pp. 701-712.
[73]
39) Navarra, P., Zhao, H., and Mucciardi, F., “Improvement of Flow Modifiers in
McGill Hear Pipes”, Multiphase Phenomena in Materials Processing (TMS), 2005,
pp. 31-41.
40) Mucciardi, F., Gruzleski, J. E., Zheng, G., Zhang, C., and Yuan, Z. US Patent
7,115,227, “Heat Pipe”, Oct., 2006.
[74]
top related