MATH 311 - Mathematical Modelling
MATH 311 - Mathematical Modelling
COURSE MATERIAL
FOR
1
ACKNOWLEDGEMENT
We acknowledge the use of the Courseware of the National Open University of Nigeria
(NOUN) as the primary resource. Internal reviewers in the Ahmadu Bello University
have also been duly listed.
2
COPYRIGHT PAGE
© 2018 Ahmadu Bello University (ABU) Zaria, Nigeria
All rights reserved. No part of this publication may be reproduced in any form or by any
means, electronic, mechanical, photocopying, recording or otherwise without the prior
permission of the Ahmadu Bello University, Zaria, Nigeria.
ISBN:
Tel: +234
E-mail:
3
COURSE WRITERS/DEVELOPMENT TEAM
Editor
Prof. M.I Sule
Language Reviewer
Ene Ojokojo
Instructional Designers/Graphics
Nasiru Tanko / Ibrahim Otukoya
ODL Expert
Dr. Abdulkarim Muhammad
4
CONTENTS
Title Page. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
Acknowledgement Page. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2
Copyright Page. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
Course Writers/Development Team. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4
Table of Content. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5
COURSE STUDY GUIDE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
i. Course Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6
ii. Course Introduction and Description. . . . . . . . . . . . . . . . . . . . . . . .6
iii. Course Prerequisites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7
iv. Course Learning Resources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7
v. Course Aim/Objectives and Outcomes. . . . . . . . . . . . . . . . . . . . . . .8
vi. Activities to Meet Course Aim/Objectives. . . . . . . . . . . . . . . . . . . . 8
vii. Time (To complete Syllabus/Course). . . . . . . . . . . . . . . . . . . . . . . . 8
viii. Grading Criteria and Scale. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
ix. OER Resources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
x. ABU DLC Academic Calendar. . . . . . . . . . . . . . . . . . . . . . . . . . . . .13
xi. Course Structure and Outline. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
xii. STUDY MODULES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17
MODULE 1: Basic Concept of Mathematical modelling. . . . . . . .17
Study Session 1: Methodology of model building. . . . . . . . . . . . . . . .17
Study Session 2: Identification and Formulation. . . . . . . . . . . . . . . . .41
MODULE 2: Solutions and Interpretations. . . . . . . . . . . . . . . . . . 71
Study Session 1: Solution of problems and Cause-effect diagrams. . .71
5
COURSE STUDY GUIDE
i. COURSE INFORMATION
Course Code: MATH311
Course Title: Mathematical Modelling
Credit Units: 3 Credit Units
Year of Study: 300 Level
Semester: First Semester
Description:
The status of this course is two units. It is packed into two Modules i.e. Module 1 and
Module 2 respectively. Module 1 is grouped into two study sessions, while module 2 has
6
only one study session. Therefore, the course can be summarized as having 2 modules
and 3 study sessions in all.
This Course Guide gives a brief summary of the contents of the course material:
methodology of the model building, identification of a model, solution of problems,
course-effect diagrams, equation types, algebraic, ordinary differential, partial differential,
differential integral and functional equations.
7
Quantitative Analysis in Management by C.N. Lomoba.
v. COURSE OUTCOMES
After you are done studying this course, you should be able to:
1. Identify different types of modelling.
2. Convert a worked problem into its equivalent mathematical formulation
3. Identify many of the formulae already familiar with as the
mathematical models of the real situation.
8
viii. GRADING CRITERIA AND SCALE
Grading Criteria
A. Formative assessment
Grades will be based on the following:
Individual assignments/test (CA 1,2 etc) 20
Group assignments (GCA 1, 2 etc) 10
Discussions/Quizzes/Out of class engagements etc 10
C. Grading Scale:
A = 70-100
B = 60 – 69
C = 50 - 59
D = 45-49
F = 0-44
D. Feedback
Courseware based:
1. In-text questions and answers (answers preceding references)
2. Self-assessment questions and answers (answers preceding references)
9
Tutor based:
1. Discussion Forum tutor input
2. Graded Continuous assessments
Student based:
1. Online programme assessment (administration, learning resource, deployment, and
assessment).
10
OER Commons: over 40,000 open educational resources from elementary school through
to higher education; many of the elementary, middle, and high school resources are
aligned to the Common Core State Standards
Open Content: a blog, definition, and game of open source as well as a friendly search
engine for open educational resources from MIT, Stanford, and other universities with
subject and description listings
Academic Earth: over 1,500 video lectures from MIT, Stanford, Berkeley, Harvard,
Princeton, and Yale
JISC: Joint Information Systems Committee works on behalf of UK higher education and
is involved in many open resources and open projects including digitising British
newspapers from 1620-1900!
11
Global Voices (http://globalvoicesonline.org/) is an international community of bloggers
who report on blogs and citizen media from around the world, including on open source
and open educational resources
Individuals (which include OERs)
1) Librarian Chick: everything from books to quizzes and videos here, includes directories
on open source and open educational resources
2) K-12 Tech Tools: OERs, from art to special education
3) Web 2.0: Cool Tools for Schools: audio and video tools
4) Web 2.0 Guru: animation and various collections of free open source software
5) Livebinders: search, create, or organise digital information binders by age, grade, or
subject (why re-invent the wheel?)
12
X. ABU DLC ACADEMIC CALENDAR/PLANNER
PERIOD
Semester Semester 1 Semester 2 Semester 3
Activity JAN FEB MAR APR MAY JUN JUL AUG SEPT OCT NOV DEC
Registration
Resumption
Late Registn.
Facilitation
Revision/
Consolidation
Semester
Examination
13
XI. COURSE STRUCTURE AND OUTLINE
Course Structure
WEEK MODULE STUDY SESSION ACTIVITY
14
Week 13 REVISION/TUTORIALS (On Campus or Online) &
CONSOLIDATION WEEK
15
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Course Outline
MODULE 1: Basic Concept of Mathematical modelling
Study Session 1: Methodology of model building
Study Session 2: Identification and Formulation
16
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
STUDY SESSION 1
Methodology of the Model Building
Section and Subsection Headings:
Introduction
1.0 Learning Outcomes
2.0 Main Content
2.1 Mathematical Modelling – What and Why?
2.2 Types of Modelling
2.3- Limitations of a Mathematical Models
3.0 Tutor Marked Assignments (Individual or Group assignments)
4.0 Study Session Summary and Conclusion
5.0 Self-Assessment Questions and Answer
6.0 Additional Activities (Videos, Animations & Out of Class activities)
7.0 References/Further Readings
Introduction:
You are welcome to study module 1 of this course, in this session, I shall introduce
you to basic concepts of mathematical modelling.
In the introduction, we indicated the need for mathematical modelling i.e., the use of
mathematics to solve our real-life problems. Our main aim is to develop the process
of mathematical modelling in which a physical system or a real-life problem is
17
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
translated into a mathematical problem. The examples given are taken from contexts
you are already familiar with e.g., motion of a simple pendulum, ratio active decay,
population growth etc.,
The need for modelling is illustrated with the help of real-life problems by beginning
with a non-trivial life problem a problem described in terms of words, about the
world around us. The different approaches to modelling a particular problem will be
discussed.
Simple exercises based on real life problems are inserted at various places so that you
can convert the life problems into abstract form by selecting a particular type of
modelling. At the end of the study session we have given an appendix where we have
discussed the method of dimensional analysis. In case you are not familiar with the
method, this would help you in understanding various examples wherever we have
used it for modelling various physical situations.
18
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
19
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
industrial production of a product would like to make calculations either to avoid the
unrealistically high cost of real scale experiments or to estimate some future situation.
It is in this context a mathematical model of a life problem gains enormous
significance.
You should understand that the concept of mathematical modelling is not a new one.
The Chinese, the ancient Egyptians, Indians, Babylonians and Greeks indulged in
understanding and predicting the natural phenomena through their knowledge of
mathematics. The architects, artisans and craftsmen based many of their works of art
on geometric principles.
20
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
effect on the process. For example, in studying the motion of a falling body, its
colour is usually of little interest.
iii) Mathematical Description: Each important quantity should be represented by
a suitable mathematical entity e.g. a variable, a function, a geometric figure etc.
Each relationship should be represented by an equation, inequality, or other
suitable mathematical assumption.
B) Finding the Solution: The mathematical formulation rarely gives us answer
directly. We usually have to do some operations and findings. This may involve a
calculation, solving an equation, proving a theorem etc.
START
FORMULATION
SOLUTION
EVALUATION
NO YES
SATISFIED STOP
21
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Before going further into the details of modelling, it is best we consider some of the
mathematical models or representations you are already familiar with.
i) Any interval of time can be modelled by an algebraic variable t, 0 < t<∞. The
numerical values of t must be obtained from the reading s of suitable clocks-a
starting time t1 and the current time t2 so that t = t2 – t1.
ii) A distance is modelled by a positive algebraic variable, say „d‟ (0 ≤ d <∞).
This distance „d‟ between two points is assigned a numerical value based on
the measurement using a rigid measuring rod (e.g., a metric rule).
iii) The modelling of space is more interesting and has led to the development of
many different axioms and theorems in geometry which in turn have played a
larger role in the application of mathematics (or more precisely in civil
engineering) to the construction of buildings, dams etc. To start with, space
can be thought of as a collection of points. This basic model can be
supplemented by further ideas such as direction and distance. We can further
supplement these by different results e.g.
(a) There is a unique circle passing through three distinct noncollinear points.
(b) The medians of a triangle are concurrent.
You are familiar with the representation of the points of space using coordinate
system e.g., the Cartesian system (x, y, z). This representation introduces you
automatically to the important features of space: (i) its three dimensionalities, (ii) its
infiniteness (if x, y, z is allowed to take all real values) (iii) its continuity in the three
directions.
Various axioms and geometrical proofs you are familiar with follow from these basic
models of distance and space.
Example 1: How would you model speed and velocity?
Solution: From their definition, speed/velocity is the rate of change of distance
travelled. Since speed is a scalar, we model it as L/T, where L is the distance
travelled and T is the time required to travel (Refer Appendix). While modelling
22
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
velocity, the direction too should be specified and hence, the model for velocity is v =
L/T where the vector notation is used additionally. Using Calculus, the model can be
further improved by writing the elementary distance as ds = (dx, dy, dz), so that v=
Only those hypotheses that have withstood large amounts of fact that the proposed
model agrees well with a small amount of data does not suffice since the
agreement could be just coincidental. It should be tested against a large amount of
data before accepting it as a law. This aspect should be clear from the fact that
nearly half a century elapsed between the works of Galileo and Newton.
23
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
24
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
large quantities etc. a mathematical model is very inexpensive and we know how to
represent a real problem in terms of appropriate equations and to solve them.
Moreover, in many situations like finding the mass of the earth or predicting the
Nigerian population in the year 2500 A.D., mathematical modelling is the only
recourse.
The mathematical approach has a number of advantages which can be illustrated by
considering the following specific examples:
i) What is the corrosive effect of the discharge of the Kaduna Refinery on the
mosaic of Arewa House?
For safety and cost reasons it would be undesirable to carry out the experiments on
the Arewa House itself without first knowing the outcome. A scaled physical model
could be used to obtain the desired information, but this would require facilities and
will not be cost effective. What do we do then? For this kind of study, a mathematical
approach is preferred.
ii) What will be the growth in the number of tourists to a historic city like Kaduna
over the next five years?
Information of this nature is frequently needed for planning purposes e.g., building
more hotels or arranging tourists attractions etc. There is really no scientific
alternative to a mathematical treatment for problems of this kind. There are more
situations like these where mathematical treatment of the problem becomes
necessary. Can you think of any?
25
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
(1)
when we take negative sign on the right hand side of Equation (1) i.e.,
, then equation models the radioactive decay. Where we assume that the rate of
decay of a radioactive atom is proportional to the number of radioactive atoms
present and λ > 0 is a decay constant. For a positive sign on the right hand side
Equation (1) represent linear models being linear differential equation.
You know from your knowledge of MATH 202 that it is very easy to handle this
equation. Its solution can be written as
N=N0 (2)
Where N0 in the case of decay denotes the original number of radioactive atoms at t =
0. This model, though very simple, agrees excellently with experimental results. In
the case of population growth would be the initial population.
We are not as lucky always. Most of the real life problems are not amenable to such
simple mathematical treatment. Many a time, the resulting equation is non-linear or
highly non-linear but still you are able to solve it. Without going into the details here,
you are given an example of the population growth model, better than the one given
by Eqn. (1),
26
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
(3)
where N is the size of the population and λ and B are the constants of proportionality.
This is a non-linear model but it is still easy to find the solution as
(4)
Where is an arbitrary constant. There are numerous experimental growth data,
say, that of bacteria with which the model agrees extremely well with. Why we call
this model better would become clear to you when we discuss the biological models
in details in section 3.
ii) Static or Dynamic
In static systems, time does not play any part, and hence the variables and
relationships describing the system are time-independent. In contrast, in dynamic
systems, time plays a very important role with the variables and/or relationships
describing the system changing with time. Consider for instance a fluid flowing
through a rigid diverging tube (see fig. 1.1.2)
A2
A1
v1
v2
P1
P2
Fig. 1.1.2: Instance of fluid flowing through a rigid diverging tube
Note: the point in a fluid flow at which the flow is directed radially outwards
symmetrically in all directions is a source. The fluid enters the system at this point.
The point at which the fluid leaves the system is a sink. The flow is directed radially
in-wards at this point in a symmetrical manner.
27
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Let the velocity of the fluid be V1 at the point P1 at which the area of cross-section of
the tube is A1. Let V2 be the velocity at the point P2 at which the area of cross-section
of the tube is A2. The principle of conservation of mass states that the rate of flow in
at P1 is equal to the rate of flow out at P2 since the tube is rigid and no at P1 extra fluid
is produced inside or nothing is taken out. In other words, there are no sources or
sinks inside or surrounding the tube.
Now, the rate of mass entering the tube at P1 = area x velocity = A1V1 rate of mass
leaving the tube at P2 = A2V2.
Conservation law can be written therefore in the form of an equation.
A1V1 = A2V2 (5)
(Rate of mass entering the tube at P1 = Rate of mass leaving the tube at P2).
Eqn. (5) is the conservation equation corresponding to the steady state i.e., all
variables are independent of time. Such a system is a static system.
In the dynamic formulations, the equations describing the model involve derivatives
of the dependent variables with respect to time.
Most of the real life problems e.g., the population growth (Eqn. (3)), the bacterial
growth, simple harmonic oscillator, rocket launch are time dependent and come under
the category of dynamic systems.
iii) Discrete or Continuous
Mathematical model may be discrete or continuous as the variables involved are
discrete or continuous. In a discrete model, the dependent variable assumes a range
of values and is characterized by discrete values of the independent variable e.g.,
suppose a population of cells divides synchronously, with each member producing a
daughter cell. Let us now define the number of cells in each generation with a
subscript, that is M1, M2, . . ., Mn are respectively the number of cells in the first,
second, …. nth generations. The number of generations, the independent variable, is
the discrete variable here. A simple equation relating successive generations is the
difference equation
28
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
29
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Such systems are called either probabilistic or stochastic system. For instance, if
you drop a rubber ball from a given height and measures the height of the bounce
with sufficient accuracy it will be found that if the same process is repeated many
times, the height of bounces is not the same every time. Even if all the maintained,
the results show lot of variability. In such cases, the system must be viewed as a
stochastic system.
Very often, when you go to a big shop what strikes you is the long queue in front of
the cash desk. The question “Why can‟t this popular shop have more than one cash
counter?” what comes to you mind.? How many counters the shop needs will depend
on the number of customers and their arrival rate, their departure rate, service time,
peak periods etc. if the arrival rate is same as departure rate the queue length will
remain the same. If the departure rate is more than the arrival rate, the queue will
disappear after some time.
If the departure rate is less than the arrival rate, then the queue will grow indefinitely
and it is this situation that requires more cash counters. Here in this situation, the
arrival time, departure time and the service time of a customer are not deterministic.
They follow certain probability distributions with mean rate of arrival, departure and
service time. Arrival and departure times satisfy Poisson distribution whereas service
time obeys exponential distributions. Models based on fitting these probability
distributions to the arrival, arrive every five minutes (given time interval) then from
Poisson z 2 distribution e-z, ze-z, e-z etc. give the probability that 0, 1, 2 etc people will
join the queue within that time. We shall not go into the details of these models here.
We shall take up such models in study session 14, when we discuss probabilistic
models.
Every real system must be considered to be subject to randomness of one type or
another, all of which are ignored in the formulation of a deterministic model. Hence,
deterministic models generally present few mathematical difficulties but can only be
considered to describe system behaviour in some average sense. Stochastic models
30
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Example 2: Which type of modelling will you use for the launching of a
rocket/satellite for meteorological purposes?
Solution:
Modelling used will be dynamic, continuous and deterministic.
It is dynamic and continuous because the flight velocity will continuously depend on
time. It is deterministic because equations describing the flight can be set up based
on established laws and the path of the satellite/rocket can be predicted with certainty.
31
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
life problem and that the two are not the same. In fact, lack of distinction between
models and reality has often slowed down the progress in modelling. It is paradoxical
that some models which were very successful initially in understanding the problem,
have become stumbling-blocks to progress. The reason is we get used to a model and
continue to use it even after it is discredited. For instance, consider the solar system.
Till 16th century, it was believed that earth was the centre of the universe and all the
other planets and sun moved around the earth. Because of this theory the model used
to study the solar systems were circular paths with earth as the centre. It was called a
Geocentric model. This model was successful in explaining night, day, seasons etc.
But there were many observations, the model could not explain.
Later in 16th century Copernicus proposed another theory called Heliocentric theory
which describes that the sun is the centre of the universe, and that all planets moved
around the sun in elliptical paths. So, in this case the models used is an elliptical path
with sun as the centre. This model successfully explained most of the problems
connected with solar system, but people simply refused to accept this model, initially.
One of the reasons for this is that the geocentric model put the earth as the centre of
the universe and people were unwilling to discard such a favourite notion.
The model is only as good as the assumptions made while constructing it and any
extrapolation which violates the assumptions may be dangerous.
Consider for instance, Eqn. (1) viz.,
it does not give good results when used for modelling the population growth. This is
∞. This means population grows exponentially without any bound. Whereas, solution
32
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Both these outcomes are extreme and are not found to occur in the nature. In this
sense, the model has severe limitations. Thus, there is a need to modify this model.
ANSWER: Mathematical models are used in the natural sciences (such as physics, biology,
chemistry) and engineering disciplines (such as computer engineering, electrical engineering) as
well as in the social sciences (such as economics, psychology). A model may help to explain a system
and to study the effect of different component and to make predictions about behaviour.
factors you need here before you apply the Newton‟s laws of motion?
4 Give two situations where mathematical treatment of problem is necessary to get
33
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
surgery etc.). helping the hospital administrators reduce the cost of operating
hospitals.
v) Annual plant produces seeds at the end of summer. A fraction of these seeds
survives the winter, and some of these germinate at the beginning of the season
(say May), giving rise to the new generation of plants. The process depends on
the age of the seeds understanding this process.
The successful use of dynamic models is based on the understanding of three closely
related concepts – dimensionality, units and scaling. Natural laws, when properly
written in mathematical form are equally valid whatever system of scientific unit is
used to express them. To say that the universal laws should be independent of the
system of units is another way of saying that they should be dimensionally consistent.
One should not equate trains with match boxes, nor can one add cabbages to kings. If
you ran 25 kilometres and earned 20 naira, would it be right to say that you ran as
much as you earned? No, because the equation.
N. 20 = 20 kilometres (1)
does not make sense. Distance is measured in kilometres and no amount of naira can
ever equal a kilometre. Technically speaking, we say that distance has the dimension
of length i.e.,
(2)
[distance] = L
34
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
This equation is dimensionally correct, because [naira per kilometre] = V/L. Thus,
the right hand side of Eqn. (4) has the dimensions of L x V/ L = V, agreeing with the
left hand side.
Units are either Fundamental (or primary) or derived (or secondary). The nature of
the fundamental units is somewhat arbitrary. They are independent of one another. If
certain of the measurable properties of physical quantities are chosen as fundamental,
then the units of measurement of all the remaining quantities can be expressed in
terms of these fundamental quantities. Hence the latter units are called derived units.
We can express all the mechanical quantities in terms of units of mass m, length 1,
and time t. But when we consider problems involving heat, we have to introduce a
new fundamental unit namely the absolute temperature θ. This is necessary as the
thermometric scale is independent of the definitions of mechanical units. Modelling
in Sociology or Economics needs an additional dimension, namely the value of a
product or income. Thus, mass m, length 1, time t, absolute temperature θ and the
value V are the five fundamental units. All the physical quantities can be expressed
in terms of these fundamental units. We shall denote the dimensions of these
fundamental units of mass, length, time, temperature and value by [M], [L], [T], [θ],
[V] respectively.
Formulation of the dimensional formulas: Dimensional formulas for velocity,
acceleration, force, work-done, pressure, power etc. can be obtained from their
definitions directly.
Velocity = time rate of displacement
= distance/time
= [LT-1]
Acceleration = time rate of change of velocity = velocity/time
= [LT-2]
Force = mass x acceleration
= [MLT-2]
35
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
36
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
iv) The dimensions of an entity are entirely independent of its magnitude. Hence dx
must have the same dimension as x, even though the differential dx is
infinitesimally small.
For example, consider the equation for a radioactive decay where the quantity
disappearing at a given time t is proportional to the quantity
Q(t), present at that time, i.e.,
= KQ (5)
with solution
Q = Q0e-kt (6)
Magnitude of Units
We have not used any numerical magnitudes of the fundamental units in the above
discussion related to dimensional analysis. After a quantity‟s dimensionality has
been settled, the number that determines its actual value will still depend upon the
units in which those basic dimensions are measured. For example, velocity has the
dimensions of length per units of time. Thus, if length is measured in kilometre and
37
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
time in hours then a car travelling at 50 km/h., will travel at nearly 14 metre per
second.
Two frequent choices for the basic dimensions of mass, length and time are
kilogramme, metre and second (System Internationale, SI) and gram, centimetre and
second (CGS system).
In the SI system, the units of length, ass and time are primary/fundamental. But the
unit of force is a derived one: it is the Newton (N) which is defined as the force which
when acting on a mass of 1 kg. Produces an acceleration of 1m/s2 i.e., IN = 1 kg m/s2.
Similarly, the derived unit of work is the Joule (J) which is defined as the work done
by a force of 1 N in moving a distance of 1m in the direction of the force.
i.e., 1J = 1Nm = 1kg m2/s2 (9)
The derived unit of power is the watt (W) which is defined as the rate of doing work
= 1J/s.
38
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
39
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
40
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
STUDY SESSION 2
Identifying and Formulating A Model
Section and Subsection Headings:
Introduction
1.0 Learning Outcomes
2.0 Main Content
2.1 - Identifying the Essentials of a Problem
2.2 - Mathematical Formulation
3.0 Tutor Marked Assignments (Individual or Group assignments)
4.0 Study Session Summary and Conclusion
5.0 Self-Assessment Questions and Answer
6.0 Additional Activities (Videos, Animations & Out of Class activities)
7.0 References/Further Readings
Introduction:
You are welcome to another study session, in the previous session you were introduce
to the concept of mathematical modelling. We discussed the necessity and advantages
of studying a real life problem through mathematical modelling.
This time around we shall take on four problems. From mechanics, biology and
economics and tried to relate them to the new concept of mathematical modelling.
You might be familiar with some of these problems during your Ordinary levels.
We shall proceed with the next step in modelling- i.e., given a real life problem, how
do you convert it to model abstraction leading to a mathematical equation? We shall
herein discuss, through some simple examples, how to
i) identify the problem with all its complexities
ii) identify the essential characteristics of the problems which have to be
incorporated into the model
41
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
iii) simplify the model by neglecting features which are of secondary or lesser
importance
iv) write the basic equations based on the basic laws of nature or intuitive logic,
which retain the essential characteristics of the model.
As in study session 1, we shall deal with examples you are already familiar with so
that your attention is focused more on the modelling aspect.
42
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
43
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Models rarely replicate a system. Also, they are not a unique representation and so
can mean different things to different people. Consider how a business man and
biologist view a mango tree. (see fig. 1.2.2)
Their conceptual views of the same object are rather different since they are both
heavily influenced by their own environment, background and objectives. The same
is true when we come to the mathematical modelling of any system or process.
Thus, there is no hard and fast approach to developing a model. But, you need to
broadly follow the following steps in the beginning:
i) Establish a Main Purpose for the model. Real situations are quite complex.
If one wishes to develop a model which will explain and account for all aspects
of a phenomenon, such a model will most likely be difficult to develop, very
complex and unmanageable. On the other hand, a model with limited purpose
will be easy to handle and still many important conclusions related to the main
purpose can be drawn. Thus, before developing a model we must be clear
about the purpose of doing it.
For example, in the case of a problem concerned with simple pendulum, what
is our main purpose? It is to find the period of oscillation of the pendulum.
ii) Observe the Real Life Situation and understand what is going on. These
44
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
The Search for Essentials of the Problem is related to the main purpose of the
model. We may be dealing with the same system but the objective of our study
related to the system may be different in each study. For example, consider
modelling the blood flow in the circulatory system. The blood cells are of a diameter
approximately 10-6 cms and hence their individual motion or rotation may not
contribute much to the fluid mechanics of blood flows in large arteries whose
diameter range from 1 mm to 1 cm. But in small capillaries of diameter 1 micro
metre, the cell sizes are comparable to the area of cross-section of the important. In
other words, a mathematical model trying to depict the flow of blood in large arteries
45
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Answer: There is no hard and fast approach to developing a model. But, you need to broadly follow
the following steps in the beginning: Establishing the main purpose for the model, observe the real
life situation and understand what is going on, and sift the essentials from the non-essentials of the
problem.
46
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Remark: g, is in fact, the gravitational acceleration of the surface of the earth. The
value of g depends upon the precise location of its measurement, but it is nearly
constant. Dimension of g = [LT-2] and its value in the SI unit is 9.8m/s2.
We start with
T0 = T0( , g) (1)
i.e., T0 is a function of and g.
It is clear that if we leave out some important quantities, we shall be in error.
Similarly, if we have included some quantities, which are, in reality, irrelevant to the
problem we will not only make the problem unnecessarily complicated but also will
arrive at an unreal answer. Very clear understanding of the problem can only help us
in making a correct choice of these quantities.
Since T0 has the dimension of time, the right hand side should also have the same
dimension. Since the length dimension appears in a linear fashion in both and g, it
follows that
T0 = T0( /g) (2)
2
You may wonder why is it /g and not + g or + g2 etc. in Eqn. (2)?
This is because [T0] = time and [g] = . Now if we want that length
should not appear on right hand side also, then and g should appear as the ratio .
47
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
For = 275 cm, one measured valued of the period is 3.371 sc. With g = 9.8m/sec 2 or
980 cm/sec2, we can use data in Eqn. (3) to find the constant A
which is approximately . If we assume, from this similarly that the period of the
pendulum is in fact given by
T0 = 2 π √ ⁄ (5)
Then you can calculate periods for strings of lengths used in the experiment. Thus,
you have in a way established a formula. Let us calculate the period using this
formula (with π = ). The values are given in Table – 1.2.2.
48
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Table – 1.2.2: Periods obtained theoretically using Eqn. (5) for two different
pendulums.
225 cm 275 cm
T 3.04 sec 3.36 sec
The agreement with the measured values given in Table 1.2.2 is quite good. In fact
the difference between measured and calculated value (for both the masses) is less
than 1.5%. thus, dimensional analysis gives us a fairly good insight into the
pendulum behaviour. However, there is much more to know about the pendulum, so
we need to develop some more detailed analytical model. Let us now try to do that in
Formulation 2.
Formulation 2: Formulation 1 was helpful in finding the period of oscillation of a
simple pendulum. But, what if we want to know more about the pendulum, for
instance, the tension in its string? We find that Formulation 1 is not enough. Hence,
we need to formulate a model which will improve our understanding of the problem
beyond Eqn. (5)
In the present formulation, we take recourse to the Newton‟s laws of motion. Here
since we are concentrating on the tension in the string we shall assume that the string
has so little mass of its own that it can be neglected in the model. We shall also
assume that the air offers little resistance. Then the only forces acting on the mass are
the tension T in the string and the gravitational force mg. the tension in the string
must act along the line of the string, while the gravitational force acts vertically
downward along the y-axis where we have assumed that the y-axis is roughly
perpendicular to the earth‟s surface (see Fig. 1.2.3).
49
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
y
x
(x, y)
mg
O
Fig. 1.2.3: y-axis is roughly perpendicular to the earth’s surface
Newton‟s second law tells us that the net force on a particle cause that particle to
be accelerated in direct proportion to its mass. Here the forces acting on the
particle are its weight mg and the tension. T. if f denotes the total force acting on the
system then we would write
Fx = m , ∑ Fy = m (6)
where ∑ Fx , ∑ Fy are net forces acting on the mass in directions parallel to the x and
y axes and the terms and are the components of the acceleration of the mass
parallel to the axes.
What is the component of T acting in the x-direction? It is – T sin θ . (Note that the
negative sign is because T acts upwards and the resolved components falls in the
negative x-direction).
What is the components of T acting in the y-direction? It is T cos θ - mg.
It then follows that
∑ Fx = - T sin θ (7)
∑ Fy = T cos θ – mg (8)
also, note that
x= sin θ and y = (1 - cos θ) (9)
Combining Eqns. (6), (7) and (8) we can obtain the following pair of differential
equations:
50
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
(10)
m = - T sin θ
m = T cos θ - mg (11)
Eqns. (10) and (11) can be solved to obtain the values of x, y by eliminating T. We
shall solve them in study session 3 once solved, this formulation helps us not only to
find the period of oscillation and the tension in the string but also the position vector
of the bob at different time „t‟.
If you now compare the two formulations, you will find that the Formulation 1 based
on dimensional analysis is quick and gives you a first guess about the nature of the
solution or the main purpose of your study. But Formulation 2, though more length,
gives you a deeper insight into the problem. Thus, the choice of a formulation
depends on how far you want to go, how much details you want to gather in hand.
Given two or more different adequate models, the question that arises is the
following. Is one of them better than the rest in some sense? There can be two
factors that can be used to rank different models to indicate the best.
i) A model M1 is preferred to a model M2 if M1 has fewer parameters. Thus
models can be ranked in terms of the number of parameters in the model.
Estimation of the parameters and design of experiments are not only costly but
also very tedious and hence to be avoided.
ii) If a model response is highly sensitive to the parameters of the model, then the
model is of limited use for prediction purposes, as small errors in parameters
will result in large errors in the model response. Thus, the models can be
ranked in terms of the sensitivity of the response to changes in parameter
values.
Note Ecology is the study of the interrelation between living organism and their
environment.
51
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
As we are get more aware of our environment, the pollution caused by the industries,
the need to conserve our forest and to maintain an ecological balance. In this context,
understanding the role played by the plants and trees around us on the earth as well as
those in the lakes, rivers or the seas becomes relevant.
We now take up in Example 2 the modelling of a problem related to ecology.
We discuss simple formulations to understand the distribution of phytoplanktons.
Phytoplanktons, as you may know, are microscopic plants, which, under certain
conditions exhibit directed motion (metres per day) along gradients of light, density
or chemical concentration. They are the basis of marine food cycle supporting life
from shrimps and cod to blue whales and lastly man. They also contribute to the
global changes in atmospheric carbon. Thus, understanding the plankton population is
of major importance in predicting future fish harvests and in assessing the possible
consequences of global warming.
52
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Now, the loss of organisms due to diffusion takes place through the boundary of the
patch; hence its rate is proportional to the surface area of the patch. On the other
hand, reproduction takes place locally within the patch, and hence its state is
proportional to its volume.
We know that for a sphere of radius r
The surface area of a sphere = 4 π r2
(12)
This means that a larger sphere carries less surface area relative to its volume than a
smaller sphere. As the volume of water mass decreases, (i.e. r becomes smaller and
smaller), diffusion plays an important role, and eventually a limit is reached beyond
53
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
which reproduction can no longer compensate for the loss due to diffusion. We want
to estimate the critical increase of plankton population.
Let D be the diffusivity, α be the rate of growth and L the size of the water mass. It is
obvious that the critical size Lc can be determined by the two parameters Dd and α
i.e. Lc = f(D, α) (13)
dimensionally
D = [L2 T-1] (14)
(You can assume this now but will be obvious to you when we do formulation 2).
Also you know, α = [T-1], Lc = [L] (15)
Dimensional analysis leads us to
Lc = A* + (16)
54
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
* + * + (17)
X
0 L
Fig. 1.2.5: the region of bloom be limited to (O, L)
The diffusion equation of the organism concentration C, is then given by
(18)
55
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
This is the one dimensional diffusion equation (refer Eqn. (17) with constant
diffusivity D. The last term α C is added to the equation because, there is an
additional production in the region, α being the growth rate.
Our aim here is to find the solution of Eqn. (18) that vanishes at x = 0 and x = L and
corresponds to a given initial concentration C(x, 0) = f(x). We shall do so in Unit 3.
But, what is interesting to note here is that this formulation is more detailed than
formulation 1 in the sense that we can find not only the critical length Lc of the water
mass but also the distribution of the concentration of the phytoplanktons at different
times and different points x in the region 0 ≤ x ≤ L. Thus, this formulation allows us
to get more details about the population.
Let us now imagine that the interchange between internal growth and loss at the
boundary has been going on for a long time and now a steady state has reached i.e., C
does not explicitly depend on t. What will happen then? Can you formulate the
problem in that situation? You may try to do that.
There is a lot of scope for improving Formulation 2. for example,
i) You can relax the assumption of a one dimensional model and consider a three
dimensional model. But, you must realize that this will make the equations
more difficult to solve. Still, for getting better insight into the problems more
difficult to solve. Still, for getting better insight into the problem, you should
not mind the complications on the mathematical analysis. Even if you cannot
solve equations analytically using existing techniques, you may try to solve
them numerically.
ii) We have, so far, not discussed the movement of the water mass. The patchiness
of the phytoplankton growth is very often wind driven and hence the velocity
of the movement of the water mass can also be included in refining the model.
This refinement will also call for more mathematical difficulties since you will
have more equations to solve. You will have to solve for the three velocity
components of the water mass and use them to solve the modified diffusion
56
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Formulation 1: We first model the raindrop as a freely falling body. For freely
falling bodies you know that the distance x travelled by a particle in a time duration
„t‟ is given by,
x = ut + (19)
57
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
i) the weight of the raindrop makes an important difference in the time it takes to
fall.
ii) the fastest time (for the largest raindrop) is about 40 seconds. Nearly three and
a half times more than the one predicted by our simple model based on the
theory of falling bodies.
Where is the snag? Before applying any theorem or a rule it is important to remember
the conditions or restrictions on which the theorem or the formula rests. In the case
of the foregoing analysis, we have tried to use a formula which is only valid if the
object is subjected only to the force of gravity. On the contrary, in the case of the
raindrops, the force of gravity is opposed by a significant amount of air drag – a
blessing indeed for otherwise we might be killed by falling raindrops. You can test it
with a golf ball. Air drag is present because of the greater density of the golf ball and
the shorter distance of fall.
Since our finding that the raindrop takes 12 seconds does not tally with the
experimentally observed findings, there arises a need to improve the model by
understanding the essentials of the problem – air drag in this case – and including it in
the formulation. We now consider another formulation of this problem.
Formulation 2: Stoke‟s law states that for spherical droplets falling in motionless air
and having a diameter D < 0.762 cm, the acceleration due to gravity is opposed by an
amount proportional to the velocity of the raindrop, specifically by an amount equal
to (0.329 x ) . Thus, we can write the equation for the rain drop as,
-
(21)
This is a simple ordinary differential equation which can be easily solved but we shall
not do that here. What you must notice here is the improvement we have introduced
58
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
into the Formulation of the model. Incidentally, this formulation goes beyond the
objectives of Formulation 1. We can predict from this model the existence of a
terminal velocity –i.e., the velocity which is an upper round to how fast the body can
go at any time during its fall.
To make this statement clearer we explain as follows:
velocity as
(22)
If the droplet ever achieves this velocity, then the acceleration rate of change of
velocity is zero. In such a situation the body continues with the same velocity and we
are able to predict the terminal velocity.
Actually, although we shall not prove it, in practice, a droplet falling according to
Eqn. (19) never quite reaches its terminal velocity but gets closer and closer, to it.
Unless its fall is interrupted by hitting the ground, the velocity eventually becomes so
close to vterm that, for practical purposes we consider it equal to vterm.
Furthermore, clouds are sufficiently high and a water droplet gets close to its terminal
velocity quickly enough that it is not a bad assumption to suppose that the droplet
travels at its terminal velocity for its whole trip.
Formulation 2 was an improvement over formulation 1 in that it introduced a very
essential item – the air drag – into the model. But this formulation too has its
limitations. It was based on Stoke‟s law valid for very small droplets.
We therefore consider yet another formulation.
59
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Formulation 3: In formulation 3 we shall use the fact that for spherical raindrops
falling in still air and having diameter D > 0.12 cm, the acceleration due to gravity is
opposed by an amount proportional to the square of its velocity, specifically an
is given by:
- * +
(24)
This model too can give the terminal velocity which can be obtained as before by
So far, models considered in Examples 1-3 were all continuous models leading to
differential equations but you know from study session 1, this is not the case always.
Models may be discrete as well. To illustrate this, we shall now take up an example
from economics, in particular from market equilibrium analysis which leads to a
discrete model resulting in difference equations.
Before we go into the formulation of the model, I shall familiarize you with the
terminologies like demand, supply, equilibrium price, stability of equilibrium etc,
associated with the market behaviour. We shall talk about them in detail in study
session 11 of section 4 when we discuss modelling in economics.
Economists often divide goods into two categories: commodities and manufactured
items. Commodities are the primary products of the earth, such as oil, corn, lumber
60
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
and so on. In both of these categories there are year-to-year fluctuations in prices of
manufactured goods which usually follow fairly smooth trends whereas the prices of
commodities often fluctuate up and down sharply.
Where do these fluctuations come from? Economists look for the answer in the
concepts of supply and demand.
The supply of a commodity in a given time period is simply the amount available for
sale in that period. But how does the supply come to exist in exactly that amount? A
fundamental fact about commodities is that one must plan far in advance for their
production. To get more wheat in the autumn, you must plant more in the spring.
Therefore, in the period between planting and harvesting, there is little that can be
done to affect the supply. This is called the Production lag.
The demand for a commodity is the amount that will be bought at a given price.
When the price goes up, demand goes down, and vice versa. Further, the response of
demand to price changes is immediate, there is no lag.
The market forces which determine the price and the quantity sold can be regarded as
manifesting themselves through the aggregate demand (D(p)) and supply functions
(S(p)) where p denotes price of the commodity. In general, Pi stands for the price of
the commodity Qi and qi denotes the quantity of the commodity Qi. Demand
function of the commodity Qi is qi = Di (pi) and the supply function of the
commodity Qi is qi = Si (Pi). But here we shall confine our discussion to a single
commodity Q. Now for any commodity Q the quantity demanded must equal the
quantity supplied at the equilibrium, price pe such that
Note: D(p) – S(p) = 0 for some p = pe (see Fig. 1.2.6).
61
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
S(p)
There is no guarantee that the equilibrium price will be established if the market is not
in equilibrium when the contracting begins. Changes in consumer preference and
innovations both tend to disturb an established equilibrium situation. The change
defines a new equilibrium, but there is again no guarantee that is will be attained. In
general, a disturbance denotes a situation in which the actual price is different from
the equilibrium price. An equilibrium is stable if a disturbance results in a return
to equilibrium and unstable if it does not. A disturbance usually creates an
adjustment process in the market. For instance, if the actual price is less than the
equilibrium price, the adjustment may consist of some buyers raising their bids for the
commodity.
Static Analysis investigates at a particular time the adjustment process and considers
only the nature of the change, i.e., whether it is towards or away from, equilibrium.
Define E(p) = D(p) – S(p) as the excess demand at price p. the Walrasian stability
condition is based on the assumption that buyers tend to raise their bids if excess
demand is positive and sellers tend to lower their prices, if it is negative. Assuming
this, a market is stable (static) if a price rise diminishes excess demand, i.e. if
62
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
In this case, nothing is said about the time path of the adjustment one might not
expect instantaneous adjustments in the present model. If the initial price is not equal
to the equilibrium price, it changes, and reconstructing takes place. If the new price
is still different from the equilibrium price, it is again forced to change. The dynamic
nature of the reconstructing may be formalized in a model in which reconstructing
takes place during periods of fixed length; say one hour, with the auctioneer
announcing the new price at the beginning of each period. The analysis of dynamic
stability investigates the course of price over time, i.e., from period to period.
Equilibrium is stable in the dynamic sense if the price converges to (or approaches)
the equilibrium price over time and it is unstable if the price change is away from the
equilibrium.
Example 4: Let us now see how we can formulate the dynamic stability of market
equilibrium.
Formulation 1: Suppose the demand function Dt for the periods t are given as
follows:
Dt = apt + b (26)
St = Apt + B (27)
Where a, b, A and B are all consonants
Then the equilibrium price pe determined by setting Dt – St = 0 for pt = pe is given by
Pe = (28)
The assumption that a positive excess demand tends to raise price can be modelled in
many different ways. A commonly used mathematical model is given by
Pt –pt -1 = kE(pt-1) (29)
Where pt is the price in period t and k is a positive constant. It means that a positive
excess demand E(pt-1) includes buyers to bid a price pt = pt-1 + kE(pt-1) > pt – 1 in the
following period.
The excess demand of period (t – 1) is then given by
D(pt – 1) – S(pt – 1) = E(pt – 1) = (a – A)pt – 1 + (b – B) (30)
63
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
= kE(p)
and Eqn. (31) takes the form = k(a – A)p + k(b – B) with solution
64
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Production plans are made after the harvest. The output corresponding to these
production plans appears in the market a year later. We thus give another formulation
of the dynamic stability but with lagged adjustment this time.
The quantity demanded in any period depends upon the price in that period, but the
quantity supplied depends upon the price in the previous period. It is assumed that
the quantity supplied in period t is always equal to the quantity demanded in that
period i.e. Dt – St = 0 This gives
Pt = pt− 1+
(35)
Eqn. (35) is a difference equation and its solution is given by
Geometrically, this happens if the slope ( ) of the demand curve has smaller
65
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
equilibrium and the unstable equilibrium are as shown in Fig. 7 (a) and 7 (b)
respectively.
p
p
po
S(p)
po
D(p)
D(p)
S(p)
q0 q q0 q
66
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
quotations from ten different vendors. The problem which you face is to choose
one of those ten vendors whose quotation maximizes your profit.
2) As in (E1), propose at least two real life problems and list the essentials and
non-essentials in the problems.
3) Consider the free fall of a body in a vacuum. The fall must be related to the
gravitational acceleration g and the height h from which the body is released.
Use dimensional analysis to show that the velocity V of the falling body is
determined by the dimensional equations V/ gh = constant.
4) A string of length is connected to a fixed point at one end and to a stick of mass
m at the other. The stick is whirling in a circle at constant velocity v. Use
dimensional analysis to show that the force in the string is determined from the
dimensionless equation
= constant.
on the pressure drop per unit length , the diameter d, and the viscosity µ.
Show that only one dimensionless equation can be formed, from which it
* +
6) Formulate Example -2 in the case of a steady state i.e., when C does not
explicitly depend on the time t. What type of equation you obtain in this case.
67
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
4.0 Conclusion/Summary
In this study session we have covered the following:
1. Real-life problem may be analysed to sift the most essential characteristics of the
problem from details of minor importance. Examples of simple pendulum, rate of
growth of world population or blood flow in the circulatory system are discussed
to make you think what is of foremost importance in the problem that needs to be
included in the model.
2. Once the essential characteristics of the model are listed according to their
priority, there can be more than one way to approach the problem. In other words,
the conversion of the real life problem into a mathematical description in terms of
the equations can be done in different ways. The particular familiar examples- (i)
Motion of a simple pendulum, (ii) Growth of phytoplankton population were
formulated in two different ways through (a) dimensional analysis and (b)
68
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
2) List two ways/methods to convert real life problem into mathematical equations.
Answer: (i) Dimension Analysis
(ii)Deterministic Methods
69
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
70
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
STUDY MODULE 2
MODULE 2: Solutions and Interpretations
Contents:
Study session 1: Solution of problems and course-effect diagrams
STUDY SESSION 1
Solution of Problems Course-Effect Diagrams, Equation Types, Algebraic,
Ordinary Differential, Partial Differential, Difference Integral and Functional
Equations
Section and Subsection Headings:
Introduction
1.0 Learning Outcomes
2.0 Main Content
2.1 - Solution of Formulated Problems
2.1.1 Motion of a Simple Pendulum (Ordinary
differential Equations) Non-Linear Model (Integral and Functional
Equation)
2.1.2 Phytoplankton Growth (Partial Differential Equation)
2.2 Interpretation of the Solution
5.0 Tutor Marked Assignments (Individual or Group assignments)
4.0 Study Session Summary and Conclusion
5.0 Self-Assessment Questions and Answers
6.0 Additional Activities (Videos, Animations & Out of Class activities)
7.0 References/Further Readings
Introduction
In the two previous study sessions, I introduced you to some of the aspects of
mathematical modelling. Study session 1, was prepared to help you to understand
71
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
definitions in mathematics, physics and biology with which you may already be
familiar. But these very familiar concepts were presented to you as a part of
mathematical modelling.
In module 1 study session 1, we have discussed different types of modelling. In study
session 2 of module 1, we dealt in detail with the first important stage of
mathematical modelling, viz., identifying essential characteristics of the problem at
hand, bringing out the most important features of the problem and formulating
mathematical representations are algebraic / differential / difference equations and
their combinations.
In study session 2 you have seen that some equations are easy to handle with such
that we can deduce the relevant information from the equation, directly as in the case
Formulation 1 in Example 1. In the other cases we need to go further and solve these
equations depending upon what type of equations they are. Here we shall discuss the
second important stage of modelling, viz. solving the formulated equations and the
interpretation of the solutions. We shall solve some of the problems formulated in
study session 2. We have given some of them as an exercise for you to try. In Sec
2.1 we have mainly discussed two problems – one related to motion of a simple
pendulum, and the other related to the growth of phytoplankton.
You will find that the different techniques you have learnt in differential equations
will be extremely useful in solving these problems.
A model is complete only when we interpret the mathematical solution of the model.
In Sec. 2.2 we shall discuss this aspect of mathematical modelling namely
interpreting/evaluating the solution. We shall explain how we interpret the solutions
obtained for the problem of simple pendulum and that of phytoplankton. You will
see that the interpretation helps us to gauge how effective the model is. In this
section we shall also talk briefly about limitation/shortcoming of a model.
You may notice that in this study session we have used only the techniques in
differential equations to obtain a solution because the resulting equations were
72
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
differential equations. But this is not the case always. In the later blocks you will see
that there are other techniques like techniques in probability and linear algebra which
are used in obtaining a solution.
Do try the exercises given in this study session sincerely. This will help you to gauge
whether you have followed concepts and the techniques we have explained.
73
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
We discuss in this section the solutions of the deeper problems formulated in study
session 2.
We shall first consider the second formulation related to the movement of simple
pendulum.
m = - T sin θ (1)
m =Tcosθ-mg (2)
A x
θ
cos θ
sin θ
(x, y)
(1 - cos θ )
mg
74
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Fig. 2.1.1: position (x, y) of the bob at that instant or the angle θ the string makes with
the vertical at that instant
to eliminate the terms m and m from Eqn. (1) and Eqn. (2), we apply the
chain rule
m cos θ [ * + ] (5)
m sinθ [ * + ] (6)
Have we improved the situation? Eqn. (5) and (6) still look quite formidable! What if
we multiply (5) by cos θ and (6) by sin θ and add the two equation? Well, we get
75
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
or
This is another equation of motion, in a direction along the string, from which the
tension can be calculated once θ (t) has been determined from Eqn. (8).
Now, have a close look at Eqn. (8) and (10). What type of differential equation are
they? You might have noticed that both (8) and (10) are
your previous knowledge of differential equations you must have observed that it is
not easy to find the solutions of non-linear differential equations, in general.
One way of getting a quick solution is to make some approximations which will
change the non-linear equations to a linear one and then solving the resulting linear
equations by known methods. But while doing so, you should always see that these
approximations do not omit the essential details of the problem.
I will now show you how both of these equations (8) and (10) can be simplified if we
are ready to make an approximation i.e. the oscillations are small. With this
approximation you will see that the equations become linear and the solutions are
easy to obtain.
76
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
T = mg [ * + ] (12)
We can further simplify Eqn. (12) by using the argument that when θ is
Small * + < 1 and hence the second term in the bracket is much smaller than the
first term. Therefore, we can neglect the second term. This would imply
T = mg (13)
77
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Isn‟t this an interesting result? Even for swings of the pendulum up to ± 30o, the
tension is a constant.
Let us now go back to Eqn. (11). For you recognize it? It is nothing but the classical
simple harmonic equation with which you were familiar even at high school.
Eqn. (11) is a simple second order ordinary differential equation with constant
coefficients. From your knowledge of ordinary differential equations, you know that
at t = 0 (15-b)
Eqn.(15-a) would imply that at t = 0, the amplitude of motion of the pendulum is θ0.
Eqn.(15-b) implies that the initial speed of the pendulum is zero. Thus, conditions
Eqn.(15-a) and Eqn.(15-b) correspond to initially holding the pendulum at rest at any
arbitrary angle θ 0 and then letting it go.
When we put t = 0 and apply Eqn. (15-a) in Eqn. (14), we get A=θ0.
Then we obtain from Eqn. (14) and apply the condition when t = 0 to
get
78
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
θ = θ0 cos √
(16)
instead of Eqn. (15-a) and Eqn.(15-b), suppose we assume that
θ = 0, at t = 0
and = ω, at t = 0
This means that at t = 0, the initial amplitude of the motion is 0 i.e. the bob is at the
equilibrium position and the initial speed is ω. Now we have it as an exercise for you
to check that the solution in this case is given by
θ = ω√ sin[√ ] (17)
Thus, individually Eqn. (16) and Eqn. (17) is also a solution of Eqn.
(11), being the solutions of a linear differential equation.
Therefore,
θ = θ0 cos[√ ] + ω√ sin[√ ]
(18)
= ω, at t = 0
Can you guess what physical situation the solution will correspond to? This
corresponds to the pendulum being release with an initial velocity ω from a point
79
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
with angular distance θ 0. We shall discuss the interpretation of the solution in detail
in Sec. 3.3
In the foregoing discussion, you have seen that the approximations we introduced
enable us to have a linear model which corresponded to the simple harmonic
oscillation. But you must keep in mind that any simplification we introduced in the
model will cost you something-in this case there had to be a restriction on the range
of values of the amplitude i.e. |θ | ≤ 30o
Suppose you have a problem in which you cannot assume the oscillation to be small.
Then in that case the non-linear characteristic of Eqn. (8) and Eqn. (10) has to be
maintained which means that the resulting model will be a non-linear model. Next,
we shall discuss the solution in this case.
* +
(19)
which we can also rewrite as
( [ ] )
* + =a constant (20)
80
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
If the initial condition is such that the pendulum is started at rest from an arbitrary
angle θ 0, then at t = 0
θ(t) = θ0, =0
Therefore, if we put θ = θ0 and = 0 in Eqn. (20), we get that the constant is (- cos
θ0)
i.e., * + - cos θ0
Therefore,
* + * + (21)
T = mg [ * +] (22)
√
√
Eqn. (22) has to be integrated to find θ, the position of the pendulum as a function of
t. What would be the limits of integration?
The pendulum swinging from - θ0 to θ0 and back again. Suppose we denote as
T0 the period of the pendulum, during the period. A quarter period would be the time
interval say, from θ = θ 0 to θ = 0 (see Fig. 2).
81
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
√ ∫ ∫
√
T0 = √ ∫
√
T0 = √ ∫
√
(23)
Then
82
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
dø = dψ = dψ
√
T0 = √ ∫
√ √
= √ ∫
√
= √ ∫
√
(24)
= √ ∫
√
As you may recognize it, the integral in the R.H.S. of Eqn. (24) is a definite integral
which gives you T0 as a function of θ 0 say f(θ 0). The integral is called an elliptic
integral. Tables are available to find the values of elliptic integrals. We have given
one such table in the appendix.
We shall now illustrate through an example, how we find T 0 for a and g, using the
table.
83
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
T0 = √ ∫ sec
√
You compare the integral on the R.H.S. with the integral given in the appendix, [look
at the column of 90o and row of 10o]. Then we get
∫ ≈ 1.58284 sec.
√
x 1.58284 sec.
≈ 1.00448 sec.
Let us now find the solution of the other problem we had mentioned at the beginning
of this section.
= +αC (25)
84
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
⇒D * +
the left-hand side of the above equation is purely function of Y and the right-hand
side is purely a function of X. so, we equal them to a constant, say K. The constant
has to be negative, otherwise the model will predict an exponential growth of
phytoplankton‟s which will not be realistic. Therefore, for convenience, we take K =
– λ2, where λ is a constant. You may again recall from your PDE course that, when
we take the constant as – λ2, then non-trivial solution exists. We now get, two
equations for determining X(x) and Y(t) as
= – λ2
D * + – λ2
85
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Zn = Bn sin x
86
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
∫ n sin n x sin m x d x =,
we get that
'
B'n = ∫ (x) sin dx (30)
You can notice that for given value of f(x), we can always evaluate the integral on the
right-hand side which gives the value of B'n. Thus, the resulting solution for C as
given by Eqn. (27) is
87
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
and hence the exponential function approaches zero as t increases i.e. as time
progresses. The second and higher order terms in the population of phytoplankton
will then be unable to maintain itself against diffusion, and the patch disappear. On
the other hand, if α < , at least the first term will increase indefinitely with time.
α< i.e.
Lc = π * + (32)
You may recall, at this stage, the expression for Lc we derived in study session 2
based on dimensional analysis. If is
Lc = A* +
(33)
Where A is a non-dimensional constant. After the present calculations, based on the
diffusion equation, we identify the constant of proportionality as π. Thus, a more
detailed model as the present one is more specific about the critical size, below which
no phytoplankton population is possible. This model also gives you the distribution of
plankton as a function of space and time i.e., C(x, t), (see Eqn. (31)).
This solution based on Formulation 2 in study session 2 is definitely more
informative than the solution using dimensional analysis based on Formulation 1.
But you must also realize that Formulation 1 was quick and served a, limited purpose
of getting preliminary information but the derivation involved more mathematical
tools, solving a second order partial differential equation in this case.
88
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
The discussion above tells us that each formulation of a model has some advantages
and disadvantages. In fact, we have to consider many other factors to evaluate the
effectiveness of a model. In the next section we shall talk about this in detail. Before
that why don‟t you try this exercise.
In-text Question: what are the techniques used in solving the formulated models?
Answer: the formulated models leads to linear equations, ordinary differential equations and
partial differential equations which can be solve using a suitable methods such as the method of
undetermined coefficient, variation of parameter method e.t.c. to obtain the exact or approximate
solution
In the next section we shall see what the significances of the solution obtained is, in
the context of the real-problem.
89
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
the interpretation of the graphs will demand a thorough knowledge about the
problems being modelled. We shall illustrate these facts using an example.
Example 1: Interpret the solution obtained for different formulations of the model of
a simple pendulum.
Solution: We have already shown you in formulation 1 how the constant of
proportionality was derived as 2 π. This was done by relating our formula for period
of oscillation (Eqn. 3, study session 2) to the experiment results with pendulums of
different lengths and masses. After establishing the formula for the period of the
pendulum as
T0 = 2 π√ (34)
90
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
θ0 T0
T 0/2 t = T0
t=0
i.e., θ = 0 = ω at t = 0 was given by Eqn. (17), Sec. 3.2.1. In this case, the amplitude
ω √
t=0 T 0 T0
2
91
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Example 2: Discuss the solution obtained for the phytoplankton growth problem.
Solution: In formulation 1 given in Unit 2 we could find the constant „C‟ given by
Eqn. (13) of unit 2, if you have observational data about the planktons. Also, if you
know the diffusivity of the planktons and the rate of growth, you can also measure the
planktons patches in the area of interest and from there calculate the value of C.
(Incidentally, if it is worth knowing that the plankton patches in the open sea appear
to occur in the order of 10-100 km). Thus, we could interpret the solution by directly
looking at the expression.
Let us consider Formulation (2). Eqn. in study session 3 gives a solution of this
problem. There we have shown that if α < then the exponential function in Eqn.
(31) approaches zero as time increases.
(See the paragraph preceding Eqn. (31) of Sec. 3.2.2.) we also know that
Lc = . Therefore, the condition α < can be replaced by L < Lc. Therefore, you
get that for any L < Lc, no sustainable growth of Phytoplankton occurs. You can also
note from Eqn. (32) that as the growth α increases, the critical size Lc gets smaller;
whereas if the intensity of diffusion D increases, Lc also increases. Both the
conclusions are in keeping with what we expect.
Even this model leaves ample scope for modification. We have assumed in both the
formulations that the planktons cannot survive outside a particular region. Again, we
make no mention of the winded driven displacement of the planktons. Including
many more factors will enhance the model and take it closer to reality. But it should
be borne in mind that the sophistication in the model may bring, along with it, more
mathematical complexities.
As the two examples, we formulated and solved in detail, show, there is lot of room
for improvement of the model. The major limitation in Example 1 was the absence of
92
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
a term representing air resistance whereas in Example 2 it was the absence of details
surrounding the patch – the velocity of the steam etc.
Sometimes it may happen that when we interpret the solution to fit the real-life
situation, we find that there is vast difference between the theoretical model that has
been created (with all the assumption) and the real-life situation. In such a situation
as you have seen and will see, the model needs to be either scrapped or revised.
In-text Question: what are the relevant of interpreting the solutions of the formulated model?
Answer: The essential facts which the solution represents have to be understood through
interpretation which have great significance in the validation of the model.
1. Obtain the solution given in Eqn. (17) under the initial condition = ω, at θ =
0 at t = 0.
2. Show from hat the bob of the simple pendulum achieves its maximum angular
velocity at θ = 0. Why is this physically reasonable? Show that your results are
applicable to both linear and nonlinear problem.
3. Using the non-linear model of the pendulum, find the period of oscillation for θ
0 = 12 sec. and = 4.
4. How would you modify Formulation 2 in unit 2 by including an external force,
say, air resistance? Find the solution of the new model.
5. In the last study session, you must have formulated the model for the problem
in E6. Recall that the situation is that the interchange between the internal
growth and loss of phytoplanktons has been going on for a long time, and a
steady state has been reached (steady state means, the organism concentration
C does not depend on time t)
Find a solution of the mathematical formulation you obtained.
93
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
(Hint: )
Answer:
Where and are constant to be determined.
2)what are the two conditions used in finding unknowns of a given ODE.
Answer: initial condition and boundary conditions.
94
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
95
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
Glossary
Glossary
absolute value
angle
arc
asymptote
compose
composition
constant
coordinate system
cosine
cube root
domain
equation
exponent
exponential
96
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
exponential decay
expression
factor
factorization
Fibonacci sequence
a sequence of numbers in which each number equals the sum of the two preceding numbers
formula
function
graph
growth
input
intercept
interval
inverse function
a function obtained by expressing the dependent variable of one function as the independent variable of
another; f and g are inverse functions if f(x)=y and g(y)=x
invertible
linear
97
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
logarithm
logarithmic
maximum
midline
the median plane of the body (or some part of the body)
minimum
model
negative
notation
output
parameter
period
periodic
periodicity
polynomial
positive
98
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
quadrant
quadratic
radian
the unit of plane angle adopted under the Systeme International d'Unites; equal to the angle at the center of
a circle subtended by an arc equal in length to the radius (approximately 57.295 degrees)
range
rate
rational
recursive
characterized by repetition
sequence
sine
ratio of the length of the side opposite the given angle to the length of the hypotenuse of a right-angled
triangle
square root
subtend
symmetry
table
tangent
99
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
translate
trigonometric function
function of an angle expressed as a ratio of the length of the sides of right-angled triangle containing the
angle
zero
100
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
INDEX
101
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
αϕ 35O 40
O
45
O
50
O
60
O
65
O
102
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
ϕ
65o 70o 75o 80o 85o 90o
α
103
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
104
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
105
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
SECTION SUMMARY
In this section; we have looked at the main stages involved in the mathematical
modelling of a real-life situation. These are:
1)Formulation of the mathematical equivalent.
2)Obtained a mathematical solution.
3)Interpreting the solution in terms of the situation and validating the solution.
In study session 1, you learnt the basic objectives of modelling and the different types
of modelling.
In study session 2, you were introduced to the formulation of our problems
i) A very familiar problem of motion of a simple pendulum
ii) An ecological problem – Growth of phytoplankton in a water mass.
iii) Time taken for a raindrop to fall from the clouds and reach the ground.
iv) A problem related to economics from market equilibrium
While going through these examples, you would have realized that while formulating
a model you need to:
i) Understand the essentials of the problem.
ii) Have the objectives (limited (limited or detailed) clear in your mind.
iii) Have the appropriate mathematical skills.
In Example 3, we also indicated how mindless simplification of the problem without
the inclusion of the essentials can lead to wrong results.
In study session 3 the first two examples of Study session 2 (the third and the fourth
examples were solved in Study session 2 itself) were analysed using the solutions of
the resulting differential equations.
After reading this unit, you should have realized that merely getting the solutions of
the formulated problem is not enough. You need to relate solution to the real-life
problem concerned by interpreting it appropriately to see if it reflects the observed
phenomena. You may need to modify the model by relaxing the assumptions or
106
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
including some important characteristic (for e.g., air resistance in the case of simple
pendulum model).
As we have stressed throughout this section; and as you will see in the unit that
follow, the modelling of any problem must go through these stages broadly. The
following flow chart given at the end may help to give you a quick overview of the
whole process.
Note that some of the options/steps given in the flow chart are not discussed in detail
in the examples discussed in this section; But we have given them here for the sake of
completeness. You would realize their importance when we discuss more models in
the latter blocks.
While formulating a model for any real life situation use this flowchart to see if you
have followed the procedure shown in it. You may do this by asking for instance, in
the case of simple pendulum these questions can be of the following type.
1)Did we make any assumptions? (Yes, we ignored the air resistance)
2)Could we find an analytic solution?
(Yes, when we linearlised. But No, when we retained the non-linearity. We had to
solve the elliptic integral numerically).
We have not discussed computer simulation in this block but we included that
possibility in the flow chart for completeness.
i) Could we validate the result using experimental values? (Yes, as far as the
period of the pendulum was concerned. The answer is NO if we are verifying
the instantaneous position of the pendulum. Our result showed that the
pendulum oscillated indefinitely whereas in reality, it comes to rest. This aspect
takes you back to model abstraction, you will have to include air resistance and
do the whole process again).
Interpretation of the results definitely allows you to understand the physics of the
problem in a better way. It is economical too, since after establishing the formula for
period of oscillation, you do not have to repeat the experiments for different lengths
107
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
of the pendulum any more. The comparison with experimental results helped to fix
the unknown parameter – the constant of proportionality in T0 α /g - as 2 π. After the
solutions are tested and validated, no more experiments will be necessary and you can
predict the values using the derived formulae.
108
MATH 311 INTRODUCTION TO MATHEMATICAL MODELLING
FLOW CHART
REAL PROBLEM
IF POSSIBLE
RECONSIDER
MODEL ABSTRACTION
MODIFY
ASSUMPTIONS OR APPROXIMATIONS
TO BE INTRODUCED
YES
109