Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
I N T E R P R E T I N G A MODEL
Structure Page 'No.
3.1 Introduction
Objectives
3.2 Solution Of A Formulated Problem
Motion Of A Simple Pe~idulum
Phytoplankton Growth
3.3 Interpretation Of The Solution
3.4 Summary
3.5 Solutions/Answers
3.6 Appendix: Table Of Elliptic Integrals
3.1 INTRODUCTION
- -
We learnt in Unit 2 that we can choose more than one type'of formulation
for the same problem. The choice of a mathematical model to be developed
must depend on the purpose for which the model is required. In See. 2.3
Example 1 of Unit 2 we saw that if the purpose of studying the nlovements
of a simple pendulum is to find its period of oscillation, a quick solution
based on dimensional analysis will serve our purpose. But if the objective of
the study is to have a deeper insight into the problem we have to use a
different model. In this case a model based on Newton's law by resolving the
forces acting on the bob of the pendulum will serve the purpose. Similarly,
we have shown in Exan~ple2 of 9ec.2.3 of Unit 2 that if we want to have a I
In the last two units we discussed the problem of finding the period of
oscillation, tension in the string and the position of the bob at any time of a
simple pendulum. There we formulated the problem using Newton's law
(Example 1, Formulation 2, Unit 2). We shall now discuss how we find the
solution.
I In the last unit we saw that the formulation resulted in two differential
1 equations (see Eqns(8) and Eqn.(9) of Unit 2), which we give below:
I m-d2x = -Tsind
dt2 (1)
I m d2y
p = TcosB-mg (2)
42 *
1
-
where the tension T and the amplitude 0 are not known. Solution and
Our objective is to find the position of t h e pendulum and t h e tension Interpretation
in t h e s t r i n g at any i n s t a n t of time. This is possible if either we know
the position (x, y) of the bob a t that instant or the angle 8 the string makes
with the vertical at that instant (see Fig.1)
You know that x, y and 8 are connected'by the relation
x = t s i n O , y = !(l -cosO),
being the length of the pendulum. Let us now solve 8 by eliminating the
terms x and y.
d2x d2y
To eliminate the terms m~ and m--2 from Eqn.(l) and Eqn.(2), we apply
dt dt
the chain rule
-
dt
_
dx - -- dxde
de d t '
By repeated application of the chain rule, we get
Have we improved the situation? Eqns. (5) and (6) still look quite
formidable! What if we multiply (5) by cos 0 and (6) by sine and add the
two equations? Well, we get
d28
m t (cos2 8.f sin2 0) - = -mg sin 8 (7)
d t2
or
d2e
i.e., m !-
dt2
+ mg sin 8 = 0
Thus, we have found the equation, in terms of 0 alone as a function oft.
We try to find a similar formuh for the tension T also, For this, we multiply
(5) by sine, (6) by cos 8 and then take the difference. We then find that
[1- (g)1'
mt (sin2 8 + cos2 8) = mg cos 0,
We can- further simplify Eqn.(l2) by using the argument that when 8 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)
Isn't this an interesting result? Even for swings of the pendulum u p to f30°,
the tension is a constant.
Let us now go back to Eqn.(ll). Do you recognize it? It is nothing but the
classical simple harmonic equation with w21icl1 you were familiar even at high
school.
Eqn. (11) is a simple second order ordinary differential equation with
constant coefficients. From y o u knowledge of ordinary differential equations
you know that
e = eocos f i t (lfi)
Iristead of Eqn. (15-a) arid Eqn. (16-b), suppose we assume that
0 = 0, at t = O
do
and - = w , at t = O
dt
This riieans that a t t = 0, the initial amplitucle of the motion is 0 i.e. the
bob is at the equilibrium position and the initial speed is w. Now we leave it
as an exercise for you to clicclc that the solution in this case is given by
e= W/6 sin ( f i t )
Therefore,
Can you guess what physical situation the solutiori will correspond to? This
corresponds to the pendulum being released with an initial velocity w from a
point with angular distance 00 . We shall discuss the interpretation of the
solution in detail in Sec.3.3.
Try this exercise now.
dt - f c o s ~ =a constant
5 (dR)2
j
i
If the initial condition is such that the pendulum is started s t rest from an
arbitrary angle 80, then at. t = 0 .
dB
Therefore if-we put 0 = do and ;;-;= 0 in Eqn.(20), we get that the constant
I 2
Substituting Eqn.(21) in Eqn.112) we get the value of the tension T in terms
of 0. You can check that the expression for T'is
,<P'
,,,drtd/p,,,,
J3'------- c e l J ~
Eqn.(22) has t o be integrated t o find 8, the position of the pendulum as a
swings
, of
of t.-go
functionfrom
interval 0 5 t 5
to +Bo
What
2
would
andbeback
the limits
again. of
Suppose
integration?
we denote
The as Tothe period
pendulum
the pendulum, during the period. A quarter period would be the time
say, from 0 = Bo to B = 0 (see Pig.2). Thus Eqn.(22) can
Fig.2: The time taken be integrated as follows: of the total period.
to go from A t o
1
0 is -
4
4 = sin -
Let sill - 00 sin$. Differentiating both sides, we get
2 2
4
cos -d$ = sin - 00 cos $ d.1)
2 2
Then
00 0
2 sin - cos ,$ 2 sin - cos $
dQ = 2
COS -4
2 = Ji- sinl
- d11,
3sin' +
Substituting for dr,b in tlle integral on the R.H.S. of Eqn.(23), we get
n- 00
2 sin - cos d~/)
2
+
sin
sin -
2
00
2 sir) -cos 7 ) d$
2
As you may recognize it, the integral in the R.H.S. of Eqn.(24) is a definite
integral which gjvcs you To as a function of 00 say f(Oo). The iiltegral is
called an elliptic integral. Tables are available to find the values of elliptic
integrals. We have given one sucll table in the appendix.
We shall now illustrate: through an example, how we find Tn for a given 1
and g, using the table.
E x a m p l e 1: Firid To if Bo = 20°1 given that i? = 20cin and g = (380cm/sec2
Solution: St~bstitul;ingfor On and 1 in Eqn.(24) we, get,
You compare the integral on the R.H.S. with the integral give11 in the
appendix, [look at the colurn~lof' 00" and row of 10"]. The11 we get
n-
e 1.58284 sec.
4
Introduction .: To = -7 x 1,58284 sec.
FZ 1,00448 sec.
Try these exercises now.
E2) Show.frorn Eqn.(l7) that the bob of the simple pendulum achieves its
maximum angular velocity at 8 = 0. Why is this physically reasonable?
Show that your results are applicable to both linear and nonlinear
problem.
E3) Using the non-linear model of the pendulum, find the period of
oscillation for Bo = 12 sec. and l = 4;
Let us now find the solution to the other problem we had mentioned at the
beginning of this section.
3.2.2 P h y t o p l a n k t o n Growth
The left hand side of the above equation is purely a function of Y and the
right hand side is purely ct function of X. So, we equal them to a constant,
say K. 'The constant has t o be negative, otherwise the model will predict an
exponential growth of phytoplanktons which will not be realistic. Therefore
for convenience, we take K = -A2, where X is a constant. You may again
recall from your PDE course that, when we take the constant as -A2, then
non-trivial solution exists. We now get, two equations for determining X(x)
and Y (t) as
Solution and
Interpretation
we get; that
,
B,= t-
2
1L n ~ x
f(x)sin -dx
L
Yoil can notice that for given value of f(x), we (:a11always evaluate the
integral on the right-hand side which gives Ihe value of 13:. Thus, the
resulting solution for C as given by Eqn.(27)is
C(x, t ) =
OC)
!1=1
13: sin ( y ) (a
exp - T-
where B: is defined by Eqn.(30)
We shall focus 011 the first term, i.e. the term corresponding to n = 1, of the
series on the right hand side of Eqn.(31). 'The argument of the exponen1;ial
fii~ictionis tiinodependent, being given by ((a - %): yT,
t. If a <
2
the
exponent becomes negative and hence the exponential furlction approadies
zero as t increases i.e. as time progresses. The second and higher order terms
in the series approach zero even fastsereIn the real situation this indicates
that the population of phytoplankton will then be uliable to maintain itself .
against diflusion, and the patch will disappear. On the other band, if
a> s, at least the first tern1 will increase indefinitely with time.
Therefore, the critical size LC,is determined from the condition a = i.e.
You may recall, at this stage, the expression for LC we derive,d in Uiiit 2
based on dimensional analysis. It is I
E5) In the iast unit you must have formulated the model for the problem in
E6. Recall that the situation is that the interchange between thp
internal growth and loss of phytoplanktons hiis been going or^ for a long
I
time, and a steady state has been reached (Steady state means, the Solution and
organism concentration C does not depend on time t.) Interpretation
Find a solution of the mathematical formulation you obtained.
In the next section we shall see what the significances of the solution
obtained is, in the context of the real-problem.
Fig.3
Tlie solution corresponding to a. different set of boundary conditions, i.e.,
3
0 = 0, = w at t = 0 was given by Eqn.(17), Sec.3.2.1. In this case, the
--
amplitude is (~4:). The graph of this equation is given below.(see. Fig.4)
Fig.4
The solutions which are cosine/sine function imply that the oscillation of a
pendulum will carry on for ever and the travelling wave will travel to infinity
without reduction of its amplitude. This is where the interpretation of the
results leads to all contradiction - i.e., a result which contradicts the
observations. All real oscillations die out, unless forced to continue by
additional external forces. This is because there always are other forces
present which damp the oscillations. These forces result from frictional or
viscous action and you may recall no provision was made in either of our two
formulations to include these damping forces. We shall be talking about the
damped si~npleharmonic motion and forced oscillations in Unit 4 of the next
block. Thus interpretation of our results highlights the shortcomings of the
model and leads to other factors which will modify the model:
In a similar way we can make observations regarding the problem in ecology
also.
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 we know the diffusivity of the planktons and the rate of
growth, we can measure the planktons patches in the area of interest and
of
from there calculate the value C. (Incidentally, 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 a t the ' ,
expression.
Next let us consider Formulation (2). Eqn. (31) in Unit 3 gives a solution of Solution and
this problem. There we have shown that, if or < s, then the exponential
function in Eqn.(31) approaches zero as time increases. (See the paragraph
Interpretation
preceding Eqn. (31) of ~ec.3.2.2.)We also lcnow that LC= y. Therefore the
condition a < can be replaced by L < LC. Therefore, we get that for any
L < LC,no sustainable growt;h of Phytoplanktons occur. You can also note
from Eqn.(32) that as the growth a 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 carlnot survive outside a particular
region. Again, we make no mention of the wind driven displacement of the
planktons. Including many more factors will enhance tlie 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, rnore mathematical complexities.
Why don't you see if you have understood what has been discussed in this
section.
As the two examples, we formulated alid solved in detail, show, there is lot of
room for improvement of the nod el. The major limitation in Example 1 was
the absence of a term representing air resistance whereas in Example 2 it was
the absence of details surrounding the patch - the velocity of the stream etc.
So~lletimesit may happen that when we interpret the solution to fit the
real-life situation, we find that these is vast difference between the
theoretical model that has been creatcd (with d l thc 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.
With this we come to the end of this unit. Let us now summarise what we
have discussed in this unit.
3.4 SUMMARY
wegetw = Bfi
W
therefore B = --
fi
8
Hence we get the solution as 9 == w -sin
(fit).
- = ucos
dt
p
i) In the linear case from (17), we can write tlie velocity of the bob as
d9
E3) Here
To = 4/&xli d$
Jl-dn26m2rU
- -
- 7$ x 1.57511
--
Substituting for in Eqn.(35), we get Solution and
dB' 2. d26' dB Interpretat ion
-rnlsin~(,) +m.!? coso -- =--TsinB-Ccos19 -
dt2 d t;
Now rnake the approxi~natio:~~thi~tsin0 z d and cos 0 z 1 and neglect
the term invoivirlg
2
(if) (%l2
, since < 1. This gives
where SIQ canti 1112 are roots 'of the auxiliary equation of Eqn.(35).