Design and Optimization of Thermal Systems Episode 3 Part 7 doc

622 Design and Optimization of Thermal Systems
in the operating conditions, the user, ambient conditions, raw materials, and other
unpredictable parameters are used to leave an acceptable margin of safety. For
instance, a dimension obtained as 2 cm from the feasible or optimal design may
be increased to 3 cm for safety, giving a factor of safety of 1.5, particularly if this
dimension refers to the outer wall of a system and may fail. Higher safety factors
are used if damage to an operator may result from leakage of radiation, hot uid,
high-pressure gases, etc. This is particularly true in the design of nuclear reac-
tors, boilers, furnaces, and many other such systems where high temperatures and
pressures arise. Therefore, the design obtained on technical grounds is adjusted to
account for safety of the user as well as of the materials and the system.
Similarly, environmental concerns can affect the design. Constraints are
placed on the types of materials that can be used, on temperature and concen-
tration of discharges into the environment, on maximum ow rates that may be
used, on types and amount of solid waste, and so on. For instance, chlorouoro-
carbons (CFCs) are not allowed because of their effect on the ozone layer. The
temperature of the cooling water (from the condensers of a power plant) that is
discharged into a lake or a river is constrained to, say, within 10nC of the ambient
water temperature by federal regulations. This imposes an important constraint
on the design of the system. Similarly, the concentration of pollutants and the
ow rate of the uid discharged into air are constrained by regulations to ensure
a clean environment. Again, these aspects arise as additional constraints on the
design and must be taken into account.
The legal issues are often related to federal and state regulations on waste dis-
posal, zoning, procurement, transportation, security, permits, etc., and can generally
be considered in terms of the cost incurred, since these involve additional expen-
ditures by a company. Similarly, social issues can often be translated into costs
for the company. For instance, providing housing, health, and other benets to the
employees results in expenses for the rm. Therefore, these aspects lead to nancial
constraints that may be quantied. However, many aspects are not easy to account
for, such as the social effects of layoffs resulting from a particular industrial develop-

ment. There may be substantial effect on the local level. However, these aspects are
largely considered at higher levels of management and affect the design only as far
as the viability of the overall project is concerned. National considerations, such as
those pertaining to defense, sometimes override the technical and economic aspects.
Items and materials that have to be imported may be substituted by those that are
available in the domestic market. Thus, constraints may be placed on the choice of
materials, components, and subsystems to take these concerns into account.
These additional constraints are usually considered after a feasible design
has been obtained. At this stage, the constraints due to the costs, safety, environ-
ment, and other aspects are considered to ensure that these are not violated by
the design. However, some of these constraints may also be more effectively used
at earlier stages of the design process. The temperature and pressure limitations
arising from safety concerns, for instance, may be built into the design and opti-
mization process. The nal design that is communicated to the management and
fabrication facilities must satisfy all such additional constraints.
Knowledge-Based Design and Additional Considerations 623
11.3 PROFESSIONAL ETHICS
A topic that has come under considerable scrutiny in recent years is that of profes-
sional ethics. This is due to the large number of cases that have been uncovered
indicating lack of proper behavior, and resulting in damage to life and property.
Ethics refers to the principles that govern the conduct of an individual or groups
of people involved in a profession. Every profession sets down its own code of
ethics to provide rules of behavior that are proper, fair, and morally correct. It
is important for individuals involved in a profession that is self-regulated, such
as engineering, to know the ethical behavior that is expected of them and to pre-
serve their integrity under various circumstances. Though many recent cases
have involved the legal and medical professions, mainly because of their direct
impact on people, engineers are also concerned with many decisions that involve
professional ethics. Certainly, design is an area that has far-reaching implications
for the profession and must, therefore, be carried out with strict adherence to

ethical standards.
Ethics has its roots in moral philosophy, and the basic features are simply
the moral values that govern personal behavior. Some of the important aspects of
proper conduct can be listed as follows:
1. Be fair to others.
2. Respect the rights of others.
3. Do not do anything illegal.
4. Do not break contracts.
5. Help others and avoid harming anyone.
6. Do not cheat, steal ideas, or lie.
On the basis of such a set of moral values, many professional societies such as
the Institute of Electrical and Electronics Engineers (IEEE) and American Board
of Engineering Accreditation (ABET) have adopted appropriate codes of ethics
that are expected to guide the members as well as companies if ethical problems
arise.
Most routine engineering activities do not involve ethical conicts. Since
prot and growth are major driving forces in the engineering profession, most
engineers are involved with pursuing these goals for advancement. However, if
their actions start infringing on the rights of others and may lead to harm to
others, through damage to property or to individuals, it is important to balance
self-interest against ethics and to follow the appropriate course of action. In some
cases, the choice is clear, but the individual may have to sacrice his or her self-
interest in order to do the proper thing. An example of this is an engineer who
designs a system, but nds that it has safety problems during testing. It is obvious
that the system must be redesigned to avoid these problems. However, there may
be deadlines to be met, the job of the engineer may be on the line, others involved
in the design may try to downplay the problems and want to go ahead, and so
on. Therefore, even a relatively straightforward problem like this one may lead
624 Design and Optimization of Thermal Systems
to a dilemma for the individuals concerned. However, the correct procedure is to

report the problems and redesign the system.
Many such ethical questions are faced by engineers, particularly those
involved in the development of new processes, systems, and products. Some of
the common ethical situations that arise are
1. Preserving condentiality
2. Giving proper credit to appropriate groups or individuals
3. Reporting data correctly
4. Meeting obligations to the public versus the employer
5. Acting in accord with personal conscience
6. Responding to inappropriate behavior by others in the company
Many readers may have already experienced some of the preceding situa-
tions. Giving proper credit to different people involved in a project is always a
touchy issue, but it is important to be fair to all. Similarly, proprietary informa-
tion must be maintained as condential because access to such information or
data is allowed only to a few individuals under the condition of condentiality.
A subcontractor who is designing a component for a system manufactured by
a different organization is often provided with essential details on the system.
These details are for use in the design process and are not to be revealed to other
parties who are not involved. Correct reporting of data has been in the news a
lot in the last few years, as some researchers have been found to falsify experi-
mental results to downplay the side effects of the product and to claim important
advances. Several drugs and their manufacturers have been in the news lately due
to allegations of such wrongdoing. When uncovered, these actions have led to
large nancial settlements and public ridicule.
The issue of inappropriate behavior by others is a very difcult one and
requires considerable care. Suppose an engineer nds out that his company is
violating federal regulations on dumping of chemical hazardous wastes. Does he
or she simply ignore the problem? According to the code of ethics, this matter has
to be reported to the proper authorities. This is known as whistle blowing and it
cannot be taken lightly. It is necessary to be sure that the activity is illegal and is

adversely affecting public safety or welfare. Proper documentation is needed to
support the accusation, and it must also be conrmed that management is aware
of the activities. If the problem lies within the company and is a consequence of
oversight, a simpler solution may be possible. Otherwise, appropriate regulatory
bodies may be contacted. Whistle blowing obviously requires great moral cour-
age, and the present position as well as the future advancement of anyone who
undertakes this effort are at stake. Such instances are rare, but they have been
increasing and they test the professional code of ethics.
In most cases where ethical conicts need to be resolved, internal appeal
processes are adequate. These start with the immediate supervisor, or with some-
one outside the region of conict, and follow the internal chain of command. Satis-
factory documentation is crucial in such appeals. Conicts arise, for instance, due
Knowledge-Based Design and Additional Considerations 625
to unfair treatment, improper credit being given, falsication of data, improper
use of funds, etc. Only if the internal appeal process is unsuccessful does one
need to resort to external options. These include contacting professional societies,
the media, regulatory agencies, personal legal counsel, and other external sources
for settling the conict. Many case studies are given by Ertas and Jones (1997)
and the IEEE code of ethics is given by Dieter (2000) to indicate the basic issues
involved and the types of ethical conicts encountered in practice.
11.4 SOURCES OF INFORMATION
An important ingredient in design is availability of relevant information on a
variety of topics that are needed for developing a successful design. Informa-
tion is needed to provide accurate data for the modeling and simulation of the
system and to use past experience and results for help with the design process.
Without adequate information on previous design efforts and existing systems,
we could repeat past mistakes, spend time on obtaining information that is read-
ily available, or generate designs that do not meet appropriate regulations and
standards. Therefore, it is crucial that we spend extensive effort on gathering
the most accurate and up-to-date information available on all facets of the given

design problem.
The types of information needed for design are obviously functions of the
system under consideration. However, the information sought for the design of
thermal systems is generally in the following main areas:
1. Material property data, including cost and availability
2. Design and operation of existing or similar systems and processes
3. Availability and cost of different types of components
4. Available computer software
5. Available empirical results, including heat transfer correlations, rel-
evant technical data, and characteristics of equipment
6. Federal, state, and local regulations on safety and environment
7. Standards and specications set by appropriate professional bodies
8. Current nancial parameters, including rates of interest and ination,
different costs, and market trends
We have seen in earlier chapters that all this information is needed at vari-
ous stages of the design process to obtain the desired inputs as well as to evalu-
ate the design. Property data are particularly important because the accuracy of
the simulation results, which are crucial to the design process, is determined by
the data provided to the model. Economic data are needed to evaluate costs, make
economic decisions during design, and determine if a project is nancially viable.
The design must meet the standards and regulations specied for the given pro-
cess or application and, therefore, it is necessary to obtain accurate information
on these. Finally, information on existing systems, software, and technical results
helps in minimizing the effort and avoiding duplication of work.
626 Design and Optimization of Thermal Systems
Throughout this book, references have been made to books, papers, and other
publications that deal in depth with a particular topic or area. Certainly, textbooks
in the areas of heat transfer, thermodynamics, and uid mechanics are a good
starting point for the design of thermal systems. Similarly, text and reference
books on numerical methods, engineering economics, optimization, and design

are important sources of relevant information. These can be used to provide the
basic technical background needed for the various aspects that are involved in
design. Different methods for analyzing various processes and systems, includ-
ing mathematical, numerical, and experimental techniques, are given in detail in
such books, along with characteristic results. Some information is also available
on common materials; components of interest in thermal systems such as pumps,
compressors, and heat exchangers; and general features of standard systems, like
air conditioners and diesel engines. The references quoted in these books and
papers may be further used to expand the source base. However, for detailed
information on material properties, applicable regulations, current economic
parameters, existing systems, characteristics of available equipment, and current
trends in industry, other sources of information are needed.
There are two main types of sources of information for design. These are
1. Public sources: Libraries, universities, research organizations, depart-
ments and agencies of the federal, state, and local government
2. Private sources: Manufacturing and supplying companies, banks, pro-
fessional societies, consultants, individuals, computer software compa-
nies, and membership and trade associations
Public sources are extensively used because the information obtained is free
or relatively inexpensive. Government reports and publications from departments
such as commerce, defense, energy, and labor are very valuable because these pro-
vide detailed information on results obtained, requirements and regulations, meth-
ods used, material property data, and various other guidelines. Patents issued also
provide an excellent source of information on existing systems. These are available
from public sources, through libraries and the Internet. Private sources are usually
expensive, but a lot of information can be obtained through company brochures for
advertising their products. The specications, cost, maintenance, servicing, and
performance of available equipment can be obtained from the supplier. In many
cases, additional details on the materials used, tests performed, basic design of the
component, and even samples can be obtained if one is interested in a particular

item. Though complete information on the component is generally condential,
enough information can be obtained to decide if a given item is appropriate for
the system being designed. Similar considerations apply to commercially avail-
able software. Extensive catalogs of manufacturers and suppliers are available
from listings such as Thomas Register of American Manufacturers. The Internet
is another important and expanding source of such information.
Most engineering companies maintain their own libraries that contain books,
technical magazines, journals, reports, information on their own products and
systems, listing of suppliers, and so on. In the present information age, they also
Knowledge-Based Design and Additional Considerations 627
have access to information available in the public domain, through the Internet,
library exchanges, and agreements with other research or professional establish-
ments. Such a collection may be large or small, depending on the size of the
company itself, but the available methods of literature search and procurement
make it a very important source of information. In addition, results from earlier
efforts, information on existing systems, properties of materials used in the past,
information on relevant regulatory and legal aspects, and so on, as applicable to
the given industry, are probably best stored here.
Therefore, there are many important sources of information on materials,
detailed technical data, existing processes and systems, items available in the
market, economic data, and on other inputs needed for design. These may be
listed as follows:
1. Handbooks
2. Encyclopedias
3. Monographs and books
4. Journals: technical and professional
5. Catalogs
6. Indexes and abstracts
7. Technical reports
8. Internet

Handbooks and encyclopedias are very useful in obtaining relevant and
detailed technical information for a given process or system. Encyclopedias are
available on physics, materials science, chemistry, uid mechanics, and so on.
McGraw-Hill’s Encyclopedia of Science and Technology is an example of such
reference books. Similarly, handbooks are available on pumps, air condition-
ing, heat exchangers, material properties, industrial engineering, manufacturing,
etc. Marks’ Standard Handbook for Mechanical Engineering, also published by
McGraw-Hill, is an example. A substantial amount of relevant and focused infor-
mation is generally available in such sources. Listings of indexes and abstracts
allow one to search for the appropriate source rapidly, particularly if the informa-
tion exists in journals, translations, and published reports. Technical and profes-
sional journals are also good sources, though these are often either too detailed
or too sketchy. Nevertheless, these can be used effectively to narrow the search to
specic and relevant sources of information. The Internet is certainly one of the
most important sources of information today.
Thus, information that is a crucial factor in design may be obtained from a
wide variety of sources. Many of these are free or quite inexpensive, while others
may be expensive. However, the current techniques available for literature searches
range from CD-ROM databases to the Internet, making it much less time consum-
ing than before to cover much of the available eld and determine what information
is available and at what cost. It is desirable to obtain as much relevant information
as possible on the given application or system so that the design process is carried
out efciently and without repeating existing data or past mistakes.
628 Design and Optimization of Thermal Systems
11.5 AN OVERVIEW OF DESIGN OF THERMAL SYSTEMS
Basic aspects. In this book, we have considered the design and optimization of
systems in which thermal transport involving heat and mass transfer, uid ow,
and thermodynamics play a dominant role. Many different types of thermal sys-
tems, ranging from refrigeration, heating, and power systems to manufacturing
and electronic equipment cooling systems, are employed as examples. The com-

plicated nature of these systems, particularly their typically nonlinear, transient,
three-dimensional, geometrically complex, and combined-mode characteristics,
leads to coupled systems of governing partial differential equations. These are
simplied through modeling to obtain algebraic equations and ordinary differ-
ential equations in many cases. Such models are combined with experimental
results, material property data, and other available information, often using curve
tting, to obtain a complete model for the system. This model is then used to
simulate the system and obtain detailed results, which can be used for the design
and optimization of the system. Thus, modeling and simulation form the core of
the design effort for thermal systems, and a successful completion of the project is
closely linked with the accuracy and validity of the model. This aspect of design
is stressed throughout the book.
The design of a thermal system starts with a close look at the problem. This
involves determining what is given or xed in the problem, what can be varied
to obtain an acceptable design, what the main requirements are, and what con-
straints or limitations must be satised by the design. This consideration leads to
the problem formulation in terms of given quantities, design variables, require-
ments, and constraints. It also denes the feasible domain for the design. The
next step is to obtain a conceptual design to achieve the desired goals. Different
ideas for the system are considered based on what is presently in use, and innova-
tive concepts are employed as well. A particular conceptual design is chosen by
employing simple estimates and contrasting different ideas.
With the design problem formulated and a conceptual design chosen, we can
now proceed to a detailed design process, the main steps of which are
1. Characterization of the physical system
2. Modeling
3. Simulation
4. Evaluation of different designs
5. Iteration to obtain an acceptable design
6. Optimization

7. Automation and control
8. Communication of the nal design
Workable design. Though all the preceding steps are involved in the devel-
opment of a successful design, modeling and simulation are particularly crucial
because most of the relevant inputs for design and optimization are obtained
from a simulation of the system using analytical, numerical, and experimental
Knowledge-Based Design and Additional Considerations 629
approaches to study the model developed for the system. Different types of mod-
els and simulation strategies are available to generate the inputs needed for design.
Even though the basic approach to modeling can be established and applied to
common thermal processes and systems, modeling remains a very elusive and
difcult element in the design process. It involves simplication, approximation,
and idealization to obtain a model that may be used to study the behavior of the
actual physical system. Creativity and experience are important ingredients in the
development of a model. A good understanding of the physical characteristics of
the system and the basic processes that govern its behavior is essential in decid-
ing what to neglect or how to approximate. The governing equations are obtained
based on the conservation principles for mass, momentum, and energy, taking
these simplications into account.
A very important question that must be answered for a model is if it accu-
rately and correctly predicts the characteristics of the actual system. This requires
a detailed validation of the model and estimation of the accuracy of the results
obtained. In many cases, simple models are rst obtained by neglecting many
effects that complicate the analysis. These models are then improved by includ-
ing additional effects and features to bring them closer to the real system. This
ne-tuning of the model is generally based on the simulation, experimental, and
prototype testing results. The model is used to study the system response and
behavior under a wide variety of conditions, including those that go beyond the
expected region of operation to establish safety limits and ensure satisfactory
operation in real life. Simulation results are also used to determine if a particular

design, specied in terms of the design variables, satises the requirements and
the constraints.
All these ideas concerning problem formulation, conceptual design, model-
ing, simulation, and design evaluation may then be put together for obtaining
acceptable designs. Different areas such as manufacturing, energy, transporta-
tion, and air conditioning, where thermal systems are of interest, are considered
to apply the various steps in the design process. It can again be seen that mod-
eling and simulation are at the very core of the design effort since the results
obtained are used to choose the appropriate design variables, evaluate different
designs, and obtain a domain of acceptable designs. Many additional aspects,
such as safety and environmental issues, may also be considered at this stage.
A unique solution is generally not obtained and several acceptable designs are
often generated. Thus, the systematic progression from problem formulation and
conceptual design to an acceptable design is highlighted in the rst ve chapters.
Though relatively simple thermal systems are considered in many cases to present
the methodology, much more complicated problems that typically arise in actual
practice can be treated in a similar way. Some actual industrial systems are con-
sidered to demonstrate the use of this methodology.
Optimization. Economic considerations and optimization need to be consid-
ered to complete the design and optimization process. Economic aspects are obvi-
ously very important in most problems of practical interest and are of particular
signicance in optimization since minimization of costs and maximization of
630 Design and Optimization of Thermal Systems
prot are important criteria for an optimal design. Economic considerations, such
as cost, return, payment, investment, depreciation, ination, and time value of
money, in the design process are discussed. Examples are given to show how
economic considerations can affect decisions on design in areas such as material
selection, choice of components, energy source, and so on. The importance of
nancial aspects in design cannot be exaggerated since the viability and success
of the project itself is usually determined by the prot, return, stock price, etc.

Again, several relatively simple situations are discussed to illustrate the approach
for considering economic factors. However, these ideas can easily be extended
to more complicated circumstances where several different considerations may
arise and interact with each other.
Optimization is discussed in detail to stress the crucial need to optimize
thermal systems in today’s competitive international market. The design process
generally leads to a domain of acceptable designs with no unique solution. Optimi-
zation with respect to a chosen criterion or objective function narrows this domain
substantially, as desired for a given application, so that the nal design is chosen
from a small range of design variables, making it close to a unique solution. Many
different strategies are available for optimization. Search methods are particularly
useful for thermal systems since discrete values of the variables are often encoun-
tered and simulation is complicated and time consuming, making it necessary to
limit the number of runs. Methods such as Fibonacci and univariate searches are
efcient and easy to use. Hill climbing techniques can also be used with numeri-
cally determined derivatives. Calculus methods and geometric programming are
useful if the simulation results are curve tted to obtain continuous expressions to
represent system characteristics. Other methods such as linear and dynamic pro-
gramming have limited use for thermal systems. Efcient computer programs are
available for most of the techniques presented here and may be used when dealing
with the large and complicated systems encountered in actual practice.
The choice of the optimization method is guided by the form in which the
simulation results are available and how involved each simulation run is. Several
simple examples are employed here to illustrate the basic ideas and the tech-
niques. These may be extended easily to many of the more complicated problems
discussed in earlier chapters and examples of practical thermal systems given in
the book. Typical large systems are considered rst in terms of the components
and subsystems, with the overall model and simulation scheme obtained by cou-
pling the simpler submodels. The complete model is simulated to obtain the char-
acteristics and behavior of the system. These results are then used for developing

acceptable designs, followed by an optimal design. There are obviously cases
where an acceptable design is not obtained, with the given requirements and con-
straints, or where acceptable designs lie within a narrow range of variables, mak-
ing it unnecessary to optimize the system.
Concluding remarks. The design and optimization of thermal systems is an
important, though complicated, eld. Many different situations can arise in actual
practice and may require specialized treatment. However, the basic approach
given in this book provides the general framework under which the design and
Knowledge-Based Design and Additional Considerations 631
optimization of a thermal system may be undertaken. Some modications may
be necessary in a few cases and additional information pertaining to materials,
economic parameters, regulations, available components, and so on, may have to
be obtained for specic applications. Creativity and originality are also impor-
tant ingredients in design, particularly in the development of the concept and the
model. In engineering practice, commercially available software packages are
often used for system simulation and optimization. However, some of the software
may be developed or programs in the public domain may be employed to provide
the exibility and versatility needed in many cases. Additional aspects, such as
safety and environmental issues, are of concern in most problems and are built
into the decision-making. The nal design is communicated to the appropriate
groups such as the management and fabrication facilities. Depending on man-
agement decisions, this could lead to prototype development, testing, marketing,
and sales. All these aspects are expected to be considered in the design projects
included at the end of this chapter.
11.6 SUMMARY
This chapter concludes the presentation on the design and optimization of ther-
mal systems by considering some recent trends, particularly knowledge-based
design aids, and some additional considerations. Knowledge-based, or expert,
systems, which use the expertise of people who are procient in given areas, have
been used effectively in several elds, such as medicine, chemical analysis, and

mining. In traditional design, such expertise is traditionally brought in by the
designer, who uses his or her knowledge of the process, materials, and system to
make decisions throughout the design process in order to obtain a realistic and
practical design. Expert systems use computer software based on articial intel-
ligence (AI) techniques to bring such considerations into the design process. The
basic components — expert knowledge and databases — in knowledge-based
design methodology are discussed. The advantages of this approach over tradi-
tional design methods are indicated. Several examples of thermal systems are
taken to illustrate the use of this methodology. These efforts are still recent and
have not been employed to their full potential. However, there is growing interest
in these methods and many relevant techniques have been developed in recent
years to help convergence to a realistic acceptable or optimal design.
This chapter also discusses the important topics of professional ethics,
sources of information, and additional constraints. These aspects arise in most
engineering endeavors, but these are particularly signicant for design because
of the innovative and creative ideas that are often involved. Different sources of
information are discussed, indicating methods to locate these and extract the rel-
evant information from them. Professional ethics can play a very signicant role
in the development of the design, in the use of available information, in the imple-
mentation of the design, and in the progress of the entire project. A brief discus-
sion of the important issues is given. Additional constraints that could affect the
feasibility of a design are presented and discussed in the context of the material
632 Design and Optimization of Thermal Systems
covered in the book. These could be of crucial importance in many cases and
could determine whether it is worthwhile to proceed with the implementation of a
given design. Finally, the chapter integrates all the ideas presented in this book as
an overview of the design of thermal systems and includes a few design projects
as problems.
REFERENCES
Budd, T. (2002) An Introduction to Object-Oriented Programming, Addison-Wesley,

Boston, MA.
Clark, K.L. and McCabe, F.G. (1984) Micro-Prolog: Programming in Logic, Prentice-
Hall, Englewood Cliffs, NJ.
Cox, B.J. (1986)
Object Oriented Programming. An Evolutionary Approach, Addison-
Wesley, Reading, MA.
Dieter, G.E. (2000) Engineering Design: A Materials and Processing Approach, 3rd ed.,
McGraw-Hill, New York.
Dixon, J.R. (1986) Articial intelligence and design: a mechanical engineering view,
AAAI, 86:872–877.
Ertas, A. and Jones, J.C. (1997) The Engineering Design Process, 2nd ed., Wiley, New York.
Giarratano, J.C. and Riley, G.D. (2005) Expert Systems: Principles and Programming, 4th
ed., Thompson Course Technology, Boston, MA.
Ghosh, A. and Mallik, A.K. (1986) Manufacturing Science, Wiley, New York.
Jackson, P. (1999) Introduction to Expert Systems, 3rd ed., Addison Wesley, Boston, MA.
Jaluria, Y. and Lombardi, D. (1991) Use of expert systems in the design of thermal equip-
ment and processes, Res. Eng. Des., 2:239–253.
Jamalabad, V.R., Langrana, N.A., and Jaluria, Y. (1994) Rule-based design of a materials
processing component, Eng. Comp., 10:81–94.
Karwe, M.V. and Jaluria, Y. (1990) Numerical simulation of uid ow and heat transfer
in a single screw extruder for non-Newtonian uids, Numerical Heat Transfer,
17:167–190.
Ling, S.K.R., Steinberg, L., and Jaluria, Y. (1993) MSG: a computer system for automated
modeling of heat transfer, AI EDAM, 7:287–300.
Lombardi, D., Jaluria, Y., and Viswanath, R. (1992) Simulation of the transport processes
in a thermal manufacturing system using symbolic computation, Eng. Appl. Artif.
Intel., 5:155–166.
Luger, G.F. and Stubbleeld, W.A. (1989) Articial Intelligence and the Design of Expert
Systems, Benjamin/Cummings, New York.
Michaeli, W. (2004) Extrusion Dies for Plastics and Rubber: Design and Engineering

Computations, 3rd ed., Hanser Gardner, Cincinnati, OH.
Rauwendaal, C. (2002) Polymer Extrusion, 4th ed., Hanser Gardner, Cincinnati, OH.
Russo, M.F., Peskin, R.L., and Kowalski, A.D. (1987) A Prolog-based expert system for
modeling with partial differential equations, Simulation, 49:150–157.
Rychener, M.D., Ed. (1988) Expert Systems for Engineering Design, Academic Press,
Boston, MA.
Sriram, D. and Fenves, S.J., Eds. (1988) Knowledge-Based Expert Systems for Engineer-
ing, Computational Mechanics Publications, Ashurst Lod ge, Southampton, UK.
Suh, N.P. (1990) The Principles of Design, Oxford University Press, New York.
Knowledge-Based Design and Additional Considerations 633
Tong, C. and Sriram, D. (2007) Articial Intelligence in Engineering Design, Academic
Press, Boston, MA.
Viswanath, R. and Jaluria, Y. (1991) Knowledge-based system for the computer aided
design of ingot casting processes, Eng. Comp., 7:109–120.
Winston, P.H. and Horn, B.K.P. (1989)
LISP, 3rd ed., Addison-Wesley, Reading, MA.
PROBLEMS
11.1. For any thermal system mentioned in Chapter 1, draw a simple sketch
of the system considered. Give the main assumptions you made to
obtain an appropriate mathematical model. Justify these and discuss
changes that you would make in order to improve the accuracy of your
model. In addition, considering yourself an expert, what expert knowl-
edge would be worth including in the design process?
11.2. An expert system is to be developed for the solution of algebraic equa-
tions. Using the information presented in Chapter 4, give a suitable tree
structure for the storage and retrieval of the relevant expert knowledge.
Consider single and multiple equations, as well as linear and nonlinear
systems. Also, include the various methods available for solution.
11.3. List the important knowledge that may be used for developing an
expert system for the design of a building air conditioning system. The

relevant information may be obtained from reference books, as out-
lined in this chapter. Of particular interest are current practice, avail-
able equipment, and costs involved. Be brief.
11.4. For the design of a solar thermal energy storage system, using water
for storage, choose a conceptual design and give the design variables.
Then discuss the expert knowledge that could be employed, in con-
junction with the basic design process, to obtain an optimal, practical
design of the system.
11.5. Present the knowledge base for the design of a furnace for heat treat-
ment of steel rods as a tree structure. Consider different types of fur-
naces, different heat sources, and different congurations.
11.6. What would the material knowledge base for the design of an elec-
tronic cooling system consist of? Use the information given in earlier
chapters on such systems and a few relevant references.
11.7. An expert system is to be developed for solving ordinary differential
equations. List the expert knowledge needed to solve an ODE, analyti-
cally or numerically. Also, present the tree structure for storing this
knowledge. Outline the scheme employed by the expert system to solve
a given ODE.
11.8. You are working as an engineer in the testing division of a large com-
pany. The division is involved in testing and evaluating equipment
from various subcontractors and recommending the best ones to the
634 Design and Optimization of Thermal Systems
production department. You are asked to test two different heaters, A
and B, for durability and performance. You nd that heater B is supe-
rior and inform your boss of your recommendation. Then you nd
out that he has already recommended heater A because of time con-
straints and he just wanted you to conrm his cursory evaluation of
the heaters. Now he wants you to manipulate the data so that heater
A is shown to be better. What are the ethical questions involved and

how will you handle the situation? Your job is probably on the line
here.
11.9. After working as a design engineer in Company A for 4 years, you are
hired by Company B, which is a competitor of Company A. You are
asked to reveal various items concerning your previous employer, such
as the new products being developed and new facilities being acquired.
Discuss the ethical issues involved here, for you as well as for the new
employer. How would you handle this problem?
11.10. You and a colleague discussed the solution to a problem found to
arise in a manufacturing process. During the discussion, an interest-
ing solution is proposed. You do not remember who suggested it or
how it was brought in. Nevertheless, it is a promising idea and it is
decided that both of you should look into the proposed concept in
detail. However, you have to travel for a couple of days and when
you come back you nd that your colleague has already passed on
the idea to management as his own, with no acknowledgment of your
contribution. Initial tests have shown that the concept will work and
this would mean a patent for your colleague and possible advance-
ment. Obviously, you are hurt at this betrayal of trust. But, you have
no written proof of your discussions with your colleague. How will
you proceed with this situation?
11.11. For the design of the thermal systems corresponding to the following
applications, list the sources that may be consulted and the type of
information being sought:
(a) Residential air conditioning and heating
(b) Hot water storage and transportation in a large industrial building
(c) Manufacture of molded plastic parts
(d) Annealing of steel rods
(e) Air conditioning of aircraft cabins
11.12. Choose any thermal system on the basis of your experience. Discuss its

basic features and formulate the design and optimization problem for
the system. Develop a simple mathematical model and the correspond-
ing simulation scheme. Outline how you would proceed to an accept-
able design and then to an optimal one. Do you expect to nd any use
for the knowledge base available on similar systems?
Knowledge-Based Design and Additional Considerations 635
DESIGN PROJECTS
Give the following information on the various design projects:
1. Problem statement: Fixed quantities, design variables, requirements,
and constraints.
2. Conceptual design: Sketch various possible systems, evaluate each,
and select one for design.
3. Mathematical modeling of the chosen design: Obtain governing equa-
tions, constraints, and requirements.
4. Computer simulation of the system: Off-design conditions, effect of
various parameters.
5. Economic analysis and optimization of the system: Discuss the initial
cost and the operating costs. Also, outline the control and safety of your
design.
6. Final report: Description of design and operation of the system,
numerical results, drawings, references, specications of components.
The design report should be concise, but complete and detailed.
Complete the design and optimization of the following problems, making
appropriate assumptions and obtaining the relevant inputs:
1. A plastic reclamation system employs the extrusion of heated plastic
to obtain cylindrical rods from scrap plastic. Design a system to heat
and extrude the plastic at the rate of 10 kg/h, assuming the maximum
allowable temperature to be 250nC and the minimum temperature for
extrusion to be 100nC.
2. In a solar energy collection and storage system, hot water is available

at 90nC, when the ambient temperature is 20nC. If a total ow rate of
500 kg/s is possible, design a system for power generation, employing
a low-boiling uid like Freon for the turbines.
3. A storage room 3 m × 3 m × 3 m is to be maintained at –5nC, when
the ambient temperature is 30nC in Arizona, for food storage. Design a
vapor refrigeration system for the purpose.
4. Steel is to be heat-treated in a furnace, which contains an inert nitrogen
gas atmosphere and which is heated electrically. The process involves
steel rods of length 1 m and diameter 20 cm. These are placed on a
conveyor, which takes the rods through the furnace. The rods must be
heated up to 700nC, with a maximum allowable temperature of 850nC,
followed by water-spray cooling after the furnace. Design the furnace
and the conveyor for the system.
5. A three-bedroom house in Boston is to be maintained at 25nC during
windy winter days when the outside temperature is –15nC. Design a
636 Design and Optimization of Thermal Systems
heat pump for heating, assuming a one-story house that has no other
heat input.
6. Design a at-plate solar collector system for heating applications. It is
to be used in Denver, Colorado, in winter and we wish to obtain water
at 60nC at the rate of 200 kg/h. Assume solar ux incident on a normal
surface to be 200 W/m
2
.
7. In a manufacturing process, only the surface of spherical steel balls of
diameter 10 cm is to be heat-treated. Interest lies in heating the mate-
rial to a depth of at least 1 mm, with a maximum of 2 mm, and then
cooling it in air. The desired temperature for heat treatment is 550nC,
with a maximum of 650nC. Design a system to accomplish this at the
rate of one ball per second.

8. Design the heat exchanger system to remove the heat rejected by a
power plant of 500 MW capacity and 33% efciency. Assume that the
heat exchangers are cooled by cold water at 10nC from a neighboring
lake.
9. An industrial heat rejection system supplies 20 MW of thermal energy
in the form of hot water at 60 K above the ambient temperature. Design
a system to recover as much as possible of this waste energy and to
provide it as electricity.
10. Design an internal combustion engine, using the Otto cycle, to obtain
a power of 150 kW. The maximum temperature and pressure in the
system must not exceed 2000 K and 2.5 MPa, respectively. A heat
loss of 10 to 15% of the total energy input may be assumed to the
surroundings.
11. A piece of electronic equipment dissipates a total of 400 W. Its base
dimensions must not exceed 40 cm × 30 cm and the height must be less
than 15 cm. A maximum of six boards can be employed to mount the
components. Design a system to restrict the temperature anywhere in
the system to less than 120nC, using an appropriate cooling method.
12. Design a hot water storage and supply system to t in a cubic region of
side 1 m. An electric heater located in the water tank is to be used for
the energy input, and the inow of water is at 20nC. The hot water must
be supplied at 70nC o 5nC, at a ow rate of 0 to 5 kg/s. The system must
be capable of handling both steady-ow and transient situations.
13. Design a continuous thermal annealing system in which spherical alu-
minum pieces of diameter 2 cm are be heat treated by moving them
on a conveyor belt and heating them by overhead radiation lamps. The
pieces must undergo fast heating to a given temperature of 350nC, held
at this temperature for a given time period of 30 seconds, and then
cooled slowly for annealing at a rate of less than 1.5nC/min. The sys-
tem is to be designed to achieve the given temperature cycle, within an

allowable tolerance of o5%. The conveyor, the heating conguration,
and the enclosure have to be designed and fabricated.
Knowledge-Based Design and Additional Considerations 637
14. Design an experimental system for the visualization of ow over 2 to 5
isolated heat sources, representing electronic components, located on a
at horizontal plate in a wind tunnel. The system must be able to move,
rotate, and position the sources, as well as the board, and then visualize
the ow by means of smoke and a schlieren optical system. The heat
input by each source is 100 W and the airow velocity ranges from 4
to 15 m/s. The sources may be taken to be 1 cm wide and 0.5 cm high,
being large in the transverse direction so that a two-dimensional ow
may be assumed.
15. Design an applicator for polymer coating of optical bers. The manu-
facture of optical bers involves a process where a thin coat of polymer
coating is applied on the ber. To maintain strength and integrity, coat-
ings must be concentric, continuous, and free of bubbles. Many experi-
mental coating applicators have limitations of performance because
of their large size and difculty of alignment. The ber diameter is
200 μm and the speed is 5 to 15 m/s. The clearance between the ber
and applicator surfaces should not exceed 50 μm. Design an applicator
that will meet these requirements.
16. Model and design a carbon furnace for use in a vacuum chamber. A
cylindrical carbon furnace is used in the materials processing of spe-
cial reactive materials, and the heating elements, which are made of
carbon, break. The heating capacity of the furnace depends upon these
element sizes, and their arrangement. The design problem would be to
choose the right size and number of elements for uniform heating. Ease
of repair is also a design criterion. The maximum temperature should
be 2000 K o5%, and the height and length of the furnace are given as
0.4 m and 0.2 m, respectively.

17. Design a warm air medication dispensing system. The employment of
aerosol as a method for drug dispensing is used widely with both liquid
and dry powder. What all devices that produce aerosolized drugs have
in common is that the inhaled aerosol is at, or below, room tempera-
ture. Preliminary testing with liquid aerosol has shown that a warm
aerosol is more comfortable to inhale and thus more effective than a
cold aerosol. The scope of this project is to design and build the hand-
held portion of an aerosol delivery system that delivers warm aerosol at
the mouthpiece. Ideal temperatures are around body temperature. Use
typical breathing air volume for calculating the ow rate. Assume uid
properties to be those of air.

639
Appendix A
Computer Programs
MATLAB
A.M.1: Common MATLAB commands for matrices
A.M.2: Common MATLAB commands for polynomials
A.M.3: Programs for root solving
A.M.4: Solutionofasystemoflinearequations
A.M.5: Interpolationandcurvetting
A.M.6: Solution of ordinary differential equations
FORTRAN
A.F.1: Gaussianeliminationmethodforatridiagonalsystemoflinear
equations
A.F.2: Gauss-Seideliterativemethodforsolvingasystemoflinearequations
A.F.3: Secantmethodforndingtherootsofanalgebraicequation
A.F.4: Newton-Raphsonmethodforndingtherealrootsofanalgebraic
equation
A.F.5: Successive over-relaxation method for the Laplace equation

A.F.6: Successivesubstitutionmethodforsolvingthesystemofnonlinear
algebraic equations in Example 4.6
A NOTE ON THE COMPUTER PROGRAMS
This appendix presents several computer programs that are commonly used for
the numerical modeling and simulation of thermal systems and processes. The
numerical methods covered here include those for curve tting, solution of a sys-
temoflinearequations,rootsolving,solutionofordinarydifferentialequations,
andsolutionofasystemofnonlinearequations.Detailsonthemethodsusedare
giveninChapter3andChapter4,aswellasinreferencescitedtherein.
TheprogramsarewrittenforMATLAB,whichhasbecomethemostfre-
quently used software for solving mathematical equations that arise in scientic
andengineeringproblems,andinFortran,whichwasprobablythemostcommon
programming language used in engineering applications in the past and which
continues to be important even today. The Fortran programs have been compiled
andexecutedonaUnix-basedcomputersystemandmayeasilybemodiedfor
other programming languages and computer systems. A few MATLAB commands
were discussed in Chapter 3 and Chapter 4, and several references for numerical
640 Design and Optimization of Thermal Systems
analysis and computer solutions were given. Additional MATLAB commands and
programsaregivenhere. Theresultsobtainedfromsomeoftheprogramsgivenin
this appendix are discussed in the text. The main purpose of presenting these com-
puterprogramsistoprovidereadyaccesstoafewimportantandsimpleprograms
that may be used for obtaining inputs necessary for the design and optimization
ofthermalsystems.Inaddition,theprogramspresentthebasicalgorithmandthe
logic employed for code development. This information may be used for develop-
ingadesirednumericalmodelorforlinkingwithavailablecodesinthepublicor
commercial domain to obtain the results needed for design.
A.M.1
For matrices a and b:
a.*b a./b a.\b element by element arithmetic; a and b must have

identical rows and columns
a*b a/b a\b matrix algebra; a and b must have appropriate
rows and columns to perform these operations
rand(n) generates random numbers between 0 and 1 for a n
x n matrix
b=26*rand(3)-10 generates 3 x 3 matrix of random numbers between
-10 and 16
max(a) gives maximum element in one-dimensional array a
min(a) gives minimum element in array a
max(max(a)) gives maximum element in matrix a
min(min(a)) gives minimum element in matrix a
[i,j]=find(a== gives row and column where maximum
max(max(a))) element is located
For system of equations Ax = b
inv(A) gives inverse A
-1
of the matrix A
AA
-1
= I identity matrix
x=A
-1
b yields the solution x; b is column vector
x = A\b also gives the solution x
Therefore,
x = inv(A)*b yields the solution vector x
Output:
>> a = 2.0;
>> b = 4.5;
>> s = [‘The number that is obtained is’, num2str(a)]

yields
The number that is obtained is 2.0
>> s = sprintf(‘The number %.5g is modified to %.5g.’,a,b)
yields
Appendix A 641
The number 2.0 is modified to 4.5.
Similarly, try other formats. %.0g gives integers. Try %8.3f for
floating point numbers, and so on.
Use of disp(s) suppresses the printing of s =
A.M.2
P
OLYNOMIALS
Roots
>>p=[1-47-62]
represents
x
4
-4x
3
+7x
2
-6x+2
>> r = roots(p)
gives the roots as:
1.00 + 1.00 i
1.00 - 1.00 i
1.00
1.00
>> pp = poly(r)
pp =

1.00 -4.00 7.00 +0.00 i -6.00 - 0.00 i 2.00 + 0.00 i
gives the polynomial with the array r as the roots.
Algebra
>>a=[1234];
>>b=[14916];
>> c = conv(a,b) % convolution (multiplication) of polynomials
c=
162050758464
>>d=a+b %addition
d=
261220
>>d=b-a %subtraction
d=
02612
>> [q,r] =deconv(c,b) % division
q=
1234 %quotient polynomial
r=
0000 %remainder polynomial
>>g=[162048697244];
>> h = polyder(g)
h=
6 30 80 144 138 72
>>x=[0.1.2.3.4.5.6.7.8.91.0];
>> y = [ 45 1.98 3.28 6.16 7.08 7.34 7.66 9.56 9.48 9.30 11.2];
642 Design and Optimization of Thermal Systems
Curve Fitting, Plotting
>>n=2;
>> p = polyfit(x,y,n)
p=

-9.81 20.13 -0.03
>> xi = linspace(0, 1, 100);
>> yi = polyval(p, xi);
>> plot (x,y,‘g*’,xi,yi,‘b-’)
>> xlabel(‘x’), ylabel(‘y = f(x)’)
>> title(‘Second Order Curve Fitting’)
A.M.3
% Bisection Method for Finding Roots of Equation f(x) = 0
%
format short g
%
% Enter limits of the domain
%
a = input(‘Enter lowest value of interval, a =’);
b = input(‘Enter highest value of interval, b =’);
%
% Apply Bisection method
%
for i = 1:20
fa = f(a);
fb = f(b);
c(i) = (a+b)/2;
fc = f(c(i))
%
% Check for convergence
%
if(abs(fc) <= 0.02)
disp(sprintf(‘Iteration converged’))
break
end

%
% Next iteration
%
if(fa*fc < 0)
b = c(i);
else
a = c(i);
end
end
c=c’
c=c’;
%
Secant Method for Root Solving
%
function [p1,err,k] = secant(f,p0,p1,delta,max1)
%
Appendix A 643
% Apply Secant method
%
for k=1:max1
p2=p1 - feval(f,p1)*(p1-p0)/(feval(f,p1)-feval(f,p0));
%
% Calculate error
%
err=abs(p2-p1);
%
% Update values
%
p0=p1;
p1=p2;

%
% Apply convergence condition
%
if (k>2)&(err<delta)
break
end
end
%
% Indicate convergence not achieved
%
if(k==max1)
disp(‘Max number of iterations reached’)
end
sprintf(‘The root is = %7.3f’,p1)
% Newton-Raphson Method for Root Solving
%
% Given equation: f(x) = exp(x) – x
2
=0
%
format short g
%
% Enter starting value of the root
%
x(1) = input(‘Enter the initial guess, x(1) = ’);
%
% Apply Newton-Raphson method
%
for i=1:20
x(i+1) = x(i) - f1(x(i))/(exp(x(i))-2*x(i));

%
% Check for convergence
%
if(abs(x(i+1)-x(i)) <= 0.01)
disp(sprintf(‘Iteration converged’))
break
end
end
x=x’
x=x’;
644 Design and Optimization of Thermal Systems
A.M.4
% Direct Solution of a System of Linear Equations
%
% Enter given coefficients and constants
%
a=[351;142;223];
b = [16;15;15];
%
% Obtain solution by matrix inversion
%
x = inv(a)*b;
x1 = a\b;
%
% Obtain solution by LU decomposition
%
[l,u,p] = lu(a);
y = l\(p*b);
x3 = u\y;
%

Gauss-Seidel Method for Linear Equations
%
% Enter Initial Guess
%
x=[0 0 0];
%
% Gauss Seidel Iteration
%
for k=1:100
%
% Store Old Values
%
xold=x;
%
% Calculate New Values
%
x(1)=(17-x(2)-2*x(3))/5;
x(2)=(8-x(1)-x(3))/3;
x(3)=(23-2*x(1)-x(2))/6;
%
% Check for Convergence
%
if abs(x-xold)<=0.001
sprintf(‘No. of iterations = %g x(1)= %7.3f x(2)= %7.3f x(3)= %7.3f’,k,x)
break
end
end
%
Gauss-Seidel Method for Finite Difference Solution of ODE
%

% Problem in Example 4.4
%
% Enter initial guess
Appendix A 645
%
for i=1:31
x(i)=0.0;
xold(i)=0.0;
end
%
% Enter boundary values
%
x(1)=100;
x(31)=100;
%
% Apply Gauss-Seidel method
%
for k=1:1000
xold(i)=x(i);
for i=2:30
x(i)=(x(i+1)+x(i-1))/2.00504;
end
%
% Check for convergence
%
if abs(x(i)-xold(i))<=0.001
sprintf(‘The Solution is:’)
%
% Print the results obtained
%

x
break
end
end
%
% Plot the results
%
j=1:31;
plot(j,x)
A.M.5
% Interpolation
%
% Enter data
%
x=[123456];
y=[106.4 57.79 32.9 19.52 12.03 7.67]’;
%
% Form Matrix for exact fit
%
c1=[x(1)^5 x(1)^4 x(1)^3 x(1)^2 x(1)^1 1];
c2=[x(2)^5 x(2)^4 x(2)^3 x(2)^2 x(2)^1 1];
c3=[x(3)^5 x(3)^4 x(3)^3 x(3)^2 x(3)^1 1];
c4=[x(4)^5 x(4)^4 x(4)^3 x(4)^2 x(4)^1 1];
c5=[x(5)^5 x(5)^4 x(5)^3 x(5)^2 x(5)^1 1];
c6=[x(6)^5 x(6)^4 x(6)^3 x(6)^2 x(6)^1 1];
c=[c1;c2;c3;c4;c5;c6];
646 Design and Optimization of Thermal Systems
%
% Find coefficients of polynomial
%

a=c\y
plot(x,y,‘-b’,x,y,‘*’)
%
% Find interpolated values at x = 2.5 and 3.6
%
x1=2.5;
y1=a(1)*x1^5+a(2)*x1^4+a(3)*x1^3+a(4)*x1^2+a(5)*x1+a(6)
x2=3.6;
y2=a(1)*x2^5+a(2)*x2^4+a(3)*x2^3+a(4)*x2^2+a(5)*x2+a(6)
%
Curve Fitting
%
% Enter data
%
x=[123456];
y=[106.4 57.79 32.9 19.52 12.03 7.67];
%
% Use MATLAB function “polyfit” for curve fitting
%
p1=polyfit(x,y,1)
p2=polyfit(x,y,2)
%
% Plot results
%
xi=linspace(1,6,100);
z1=polyval(p1,xi);
z2=polyval(p2,xi);
plot(x,y,‘*’,xi,z1,‘g’,xi,z2,‘b’)
% Exponential Expressions
%

% Enter data
x=[1251015202530];
y=[109.6 99.3 73.8 45.2 26.8 17.2 9.9 7];
%
% Define New Variables for Linearization
%
z=log(y);
%
% Use MATLAB function “polyfit” for curve fitting
%
p=polyfit(x,z,1)
%
% Obtain Constants A and a
%
a=p(1)
A=exp(p(2))
%
Plotting Polynomials
x=0:0.01:2.5;
p1=[1 -4 7 -6 2];