YouSub
Untitled

en English

Untitled English subtitles

0 downloads · 407 lines · added Jul 28, 2015

Report a problem
What's wrong?
Advertisement

English transcript

Welcome to today’s class of non-linear vibration. Today’s class we are going to study about
this Lindstedt Poincare method for finding the solution of non-linear equation motion.
So in the previous two classes, we have studied about this graphical methods or this qualitative
analysis we did, and we know what is a solution or equilibrium point, types of equilibrium
points, and representation of the solution, how solution of a linear system differs from
the non-linear systems and how to determine the solutions. So, to determine the solutions,
we studied some qualitative approach and quantitative approach. In this quantitative approach, so
last class we have studied about the straight forward expansion method and today we are
going to study about this Lindstedt Poincare technique.
So, also I told you about these numerical methods. So, these are some of the non-linear
equations we are going to study or you, by this time know about these systems. So, first
one is the Duffing equation. So, in case of Duffing equation, we can take this term equal
to 0 for free vibration or for force vibration, we can this term equal to epsilon f cos omega
t. So, these equations may be strongly non-linear or weakly non-linear depending on the coefficients
of the non-linear terms.
For example, if I will take an equation, so let me take an equation in this form, that
is x double dot plus 4 x plus 0.1 x cube. So, this is equal to 0.1 cos omega t. So,
in this case, I can use the scaling parameter and the book keeping parameter to order this
equation. For example, I may use, let x equal to alpha y. So, if I will substitute this
equation, in this equation, so then this will becomes alpha y double dot plus 4 alpha y
plus 0.1 alpha cube y cube, so this will be equal to 0.1cos omega t. So now, I can write
this equation again in this form, y double dot plus 4 y plus 0.1 alpha square y cube
equal to 0.1 cos omega t. So, in this case, this if I want to make this strongly non-linear
or coefficient of u cube, if I want to increase, so I can put this 0.1 alpha square equal to
4 and I can find this alpha square equal to 4 by 0.1. So, this is equal to 40 or I can
find this alpha root over 40. So, which cannot be neglected.
In this equation, one can neglect this 0.1 with respect to this 4 and consider this equation
as a linear system. But if I will keep this term alpha equal to root over 40, so this
term cannot be neglected with respect to this 4. So, this way, one can order the system.
So, by ordering the system, so one can make a weekly non-linear system to that of a strongly
non-linear system and also one can use this book keeping parameter. For example, in this
equation, so if I will put a book keeping parameter epsilon, which is equal to 0.1.
Let me take epsilon equal to 0.1. So this equation, I can write in this form. So, x
double dot plus 4 x plus, I can put this is epsilon x cube equal to, so this for this
0.1, I can write this is equal to epsilon, so epsilon cos omega t. So, here the coefficient
of x cube, so will be equal to, if this epsilon I represent this is a small parameter and
this coefficient will be 1, so which is comparable to this one, this 4.
Here also, the coefficient of cos omega t, so I can write this epsilon into 1 into cos
omega t, so this will also be comparable to 1 or if I will take a term, which is less
than this, let me take epsilon equal to 0.01. So, if I will take epsilon equal to 0.01,
then the same equation I can write in this form, x double dot plus 4 x plus, so here,
if I will divide this thing by 0.1 by 0.01, so then it will be epsilon into 10 x cube
will be equal to, so epsilon into 10 cos omega t. So, one can change this value, that is
the coefficient by substituting different type of epsilon, different book keeping parameters
and one can write this equation in a different way. So, these coefficients can be varied
by taking different epsilon or the book keeping parameter or different scaling parameter like
this alpha. So, after getting a non-linear differential equation of motion, so we know,
by using this qualitative approach, so we can solve this equation.
For example, here we use this qualitative approach. So, in this qualitative approach,
if this equation is u double dot plus f u equal to 0, so one can write this velocity
or x dot will be equal to root over 2 into h minus f x, where f x is the potential function
and h is the total energy.
We have solved some example, that is this equation we have taken, x double dot plus
x minus 0.1 x cube equal to 0 and we got this.
So, this is the potential function and these are the phase portraits. So, corresponding
to this maximum value of the potential energy, so we have a saddle point and corresponding
to the minimum value, we have a center. So, corresponding to minimum potential energy,
we have a center and corresponding to this maximum potential energy, we have saddle point.
So, in this way we did this qualitative analysis. After carrying out the qualitative analysis,
also we have studied in the last class by using this Runge-Kutta 4th order method.
So, in which, the algorithm can be written in this form. So, given a first order differential
equation d y by d x equal to f x y, where the initial condition y a equal to y 0, so
a belongs to a b. So, in that case, the k plus 1th solution is related to the kth solution,
which is derived by this Taylor series by this equation. So, y k plus 1 equal to y k
plus k 1 plus 2 k 2 plus 2 k 3 plus k 4 by 6, where k 1 equal to h into the functional
value of f at x k and y k. So, k 2 also, k 2 equal to h into the functional value of
f at x k plus x by 2 plus y k plus k 1 by 2. Similarly, k 3 equal to h into functional
value of x k plus h by 2 and y k plus k 2 by 2 and k 4 equal to h into f x k plus h
k by 2 and y k plus k 3 by 2. So, one can write a program taking this value
k 1 k 2 k 3 k 4 and substituting in this equation and taking different time increment, so one
can find this y k plus 1 in terms of y k and one can iterate it till one obtains a steady
state solution.
So, to use this differential equation for our purpose, so for example, in our case,
our equation is, let it is x double dot plus x double dot plus omega 0 square x plus alpha
x cube equal to f sin omega t. So, in this case, first we have to reduce it to that of
a first order equation and after reducing this thing to a first order equation, and
then only one can apply this 4th order Runge-Kutta method to find the solution. So, one can write
the, to reduce it, one can write this, one can take this x dot. So, this is the first
equation.
One can take x equal to, let z 1 equal to x and one take z 2 equal to x dot, so thus
first equation one can write z 2 or d z 2 equal to d x dot. That is your x dot and then
this x double dot d z 1, so this is d z 1, so d z 1 equal to x dot and d z 2, that is
the differentiation of this x dot, that is x double dot can be written. So, this is d
z 2. That is, x double dot, that thing can be written in this way, f sin omega t minus
omega 0 square x minus alpha x cube. So, taking these two equations, so one can, so these
are the first order equations. So, this x dot, and this is equal to z 2 d z 1, that
is equal to derivative of x, that is x dot. So, that thing equal to z 2 and d z 2, that
is x double dot, can be written in this form. So, this is f sin omega t minus omega 0 square
x minus alpha x cube. So, after getting these two first order equations, one can use this
method to find the solution. Last class, we have seen the solutions for different cases
and we have plotted the time response and phase portrait for these cases.
So, in case of the quantitative analysis, already we have seen the straight forward
expansion method.
So, in case of this straight forward expansion method, so we have seen, so given a equation,
let this is the original equation motion, x double dot plus alpha 1 x plus alpha 2 x
square plus alpha 3 x cube equal to 0, in which we have taken this alpha 1 equal to
omega square, omega 0 square. So, in case of the straight forward expansion, we have
expanded this response, that is x which is a function of t and a book keeping parameter
epsilon equal to epsilon x 1 t plus epsilon square x 2 t plus epsilon cube x 3 t in the
original equation. After substituting this equation, let this is equation 1, equation
2, and substituting equation 2 in equation 1 and ordering them of the order of epsilon,
so one can obtain this equation, equation 3 is for order of epsilon and equation 4 is
for order of epsilon square and equation 5 is for the order of epsilon cube.
So, after getting these equations, so one knows the solution of this. So, x double dot
plus omega 0 square x 1 equal to 0, so the solution of this equation. So here, as the
acceleration is proportional to displacement x 1, then the motion is simple harmonic.
In case of the simple harmonic motion, so the solution can be written as a sin omega
t. So, one can put the solution equal to a sin omega t. One can take this initial displacement
and velocity in polar form and write. So, one can write s 0 equal to a 0 cos beta 0
and b 0 equal to minus a 0 omega 0 sin beta 0 and substitute it in this equation and one
can find the solution.
So, after getting the solution of x 1, one can substitute that thing in equation of x
2 and one can get the solution in this form.
After getting the solution of x 2, one can substitute it in this equation of x 3.
One can obtain this equation, so in which one can get this x 3 double dot plus omega
0 square x 3 equal to alpha 2 square a cube by 3 omega 0 square into 3 cos omega 0 t plus
beta minus cos omega 0 t plus beta into cos 2 omega 0 t plus 2 beta minus alpha 3 a cube
into cos cube 3 0 cos cube omega 0 t plus beta. So, which can be also written as, 5
alpha 2 square by 6 omega 0 square minus 3 alpha 3 by 4 into a cube into cos omega 0
t plus beta minus alpha 3 by 4 minus alpha 2 square by 6 omega square a cube cos 3 omega
0 t plus 3 beta.
So, one can observe that, due to the presence of this term cos omega 0 t plus beta, so the
solution of this term, this part will contain a secular term. So, as the cos has a, so this
term has a frequency omega 0 and here you have this omega 0 square x 3. So, this will
lead to a secular term, which will have a solution in this form. The solution will be
10 alpha 2 square minus 9 alpha 3 omega 0 square by 24 omega 0 cube into a cube t, so
a cube into t into sin omega 0 t plus beta. So, as this time term is multiplied with this
term, so with increase in time or as t tends to infinity, so one can get the response to
be infinity.
But, in actual case, we have observed that for minimum potential, so the response of
this type of system will be periodic solution or the response or the solution is that of
a center. So, in this case, the solution should be periodic. But by doing this straight forward
expansion method, we are getting a solution, which shows that it will increase with time
or the system becomes unstable as the time progresses. So, the solution what we obtained
is not correct. So, one has to modify this procedure to correct the solution or to get
a correct solution. So, for this purpose, Lindstedt and Poincare, so they have proposed
a method.
So, in that case, one can take, so this one, in the first step, one can take this time,
non dimensional time tau equal to omega t and where omega is an unspecified function
of epsilon and that omega can be written in this form. So, omega epsilon equal to omega
0 plus epsilon omega 1 plus epsilon square omega 2. Like previous case, here also one
can take x t epsilon equal to epsilon x 1 tau plus epsilon square x 2 tau plus epsilon
cube x 3 tau plus. So, one can take as many term as possible. But it is difficult to take
higher order terms. So, for that purpose, one can write a symbolic code to find the
solution. So, in this Lindstedt Poincare method, the first stage step is to convert this equation
by using this non dimensional time tau equal to omega t. Then substitute; write this omega
in terms of epsilon. So, that is where epsilon is a book keeping parameter. So, using this
book keeping parameter, so omega epsilon is written as omega 0 plus epsilon omega 1 plus
epsilon square omega 2 and this. So, this is the basics of this Lindstedt Poincare method.
By substituting this equation and proceeding in the previous way, one can find the solution
of the system.
So, for example, in this case, so if we have the equation d square x by d t square plus
alpha 1, so plus alpha 1 x plus alpha 2 x square plus alpha 3 x cube, let us take up
to 3 terms equal to 0. So, where one can write this alpha 1 equal to omega 0 square and then
substituting this equation, so substituting the expression for omega epsilon and x epsilon
in this equation, so one can write this equation. So, if you differentiate this, so d square
x by d t square, so one can write this as d by d tau of d x by d t. So, into d tau by
d t, so this becomes, so we have to substitute this equal to omega. So, d tau by d t equal
to omega. So, one can put this equal to omega and then it can be written as d by d tau.
So, inside also one can write this d x by d tau into d tau by d t into d tau by d t.
So, this is omega and this is omega. So, this will become omega square into d square x by
d tau square. So, the first term can be, so with respect to this, when one differentiate,
so this becomes omega square into d square x by d tau square. So, by substituting that
thing, so one can have omega 0 plus epsilon omega 1 plus epsilon square omega 2 whole
square into d square by d tau square. So, this is x x equal to epsilon x 1 plus epsilon
square x 2 plus epsilon x cube plus higher order terms plus m equal to 1 to n alpha n.
So, one can substitute these things, epsilon x 1 plus epsilon square x 2 plus epsilon cube
x 3 plus higher order terms to the power n equal to 0. So, this term one can expand for
n equal to 1. So for example, one can write this part for n equal to 1. So, for n equal
to 1, this becomes omega 0 square into epsilon x 1 plus epsilon square x 2 plus epsilon cube
x 3 and for the second term, so that is alpha 2. So, alpha 2 into, so it will be epsilon
x 1 plus epsilon square x 2 plus epsilon cube x 3 plus higher order term whole square plus;
let me take up to 3 terms, so then it will be epsilon x 1 plus epsilon square x 2 plus
epsilon cube x 3, so to the power cube.
Now, expanding this one, the second one, so this is alpha 2 into, so if I will make it
square, then this will be a square plus b square. So, this is a plus b plus c whole
square. So, it will be a square plus b square plus c square plus 2 a b plus 2 b c plus 2
c a. So, I will have this epsilon square x 1 square plus epsilon 4 th x 2 square plus
epsilon 6th x 3 square plus 2 into epsilon cube into x 1 x 2 plus 1 into epsilon 4 th
into x 1 into x 3 plus 2 into epsilon to the power 5 in x 2 into x 3. Similarly, one can
expand this x 1 plus. So, epsilon x 1 plus epsilon square x 2 plus epsilon cube x 3.
But in this case, if we are limiting our analysis to up to this 3rd order, then we can take
only the first term, that is your epsilon x 1 cube. Higher order terms one can neglect
because the higher order terms, this term will come, epsilon to the power 6 x to the
power 6 x to the power cube and this term will come, x to the power 9 into x to the
power 9 x to the x to the power x 3 to the power 3 and similarly, higher order terms
will come.
So now, arranging of the order of epsilon, so one can write this. So, this is order of
epsilon. For example, for this order of epsilon one can have, so omega 0 square into, so this
is d square, this is d square epsilon x 1 by d t square. So, order of epsilon, this
first term will come. So, this will become, so taking that omega 0 square out, so here
also we have a omega 0 square epsilon x 1. This is the term, first term and this is the
first term and this corresponds to the first term. So, one can write this equation in this
form, d square x 1 by d tau square plus x 1 equal to 0. Similarly, for the second term
order of, so this is order of epsilon square. So, in case of the order of epsilon square,
one can write omega 0 square d square x 2 by d tau square plus x 2 will be equal to
minus 2 omega 0 omega 1 into d square x 1 by d tau square minus alpha 2 x 1 square.
Similarly, for order of epsilon cube, one can write this omega 0 square into d square
x 3 by d tau square. So, this is d tau square plus x 3 equal to minus 2 omega 0 omega 1
d square x 1 by d tau square minus 2 alpha 2 x 1 x 2 minus omega 1 square plus 2 omega
0 omega 3 d square x 1 by d tau square. So, the solution of this equation is well known
and one can write the solution of this equation. So, x 1 will be equal to a cos tau plus beta.
So, by substituting this, where a and beta are constant, so by substituting this x 1
in this equation, so one can write, so omega 0 square d square x 2 by d tau square plus
x 2 will be equal to minus 2 omega 0 x 1. So, differentiating this thing twice, so it
will become minus a cos tau plus beta minus alpha 2 and x 1 square. So, x 1 square will
give a square cos square tau plus beta.
One can expand this cos square term, cos square tau plus beta equal to, so 1 plus cos 2 beta
1 plus cos 2 into tau plus beta by 2. While substituting that thing, one can get this
equation. So, omega 0 square d square x 2 by d tau square plus x 2 will be equal to
2 into omega 0 omega 1 into a cos tau plus beta minus half alpha 2 a square into 1 plus
cos 2 tau plus beta. Now, one can see that these terms or the coefficient of this, so
this frequency equal to 1 and here also frequency equal to 1. So, this will lead to a secular
term. Already we know that the secular term should not be present in the response. So,
the response is bounded. So, there should not be any secular term. So, this term should
be eliminated from this equation. So, to eliminate this term from this equation, which is 2 omega
0 omega 1 a cos tau plus beta. So, as this cos will have a value from minus 1 to plus
1 and this omega 0 is not equal to 0. So, the only remaining term which should be 0
is equal to omega 1. So, from this we got one condition, that this omega 1 equal to
0.
So, while initially we are assuming the solution in Lindstedt Poincare method or we have assumed
this omega equal to omega 0 plus epsilon omega 1 plus epsilon square omega 2 plus epsilon
cube omega 3, in this case only omega 0 is known to us, so which is equal to or omega
0 square equal to alpha 1. But the other terms we have to find by eliminating the secular
terms. So, in this case, we found that this omega 1 equal to 0. So, by eliminating this
secular term, so we have found this omega 1 equal to 0. So, the solution, remaining
solution of this, that is omega 0 square d square x 2 by d tau square plus x 2 equal
to minus half alpha 2 a square into 1 plus cos 2 tau plus beta, so it has 2 components.
So, the first component, so you can write or you can have the particular integral. So,
for these two components, one component equal to minus half alpha 2 a square and the other
component equal to minus half alpha 2 a square into cos 2 tau plus beta.
The solution of the second component can be written or the solution of this first component
will be alpha 2 a square by 2 omega 0 square. So, this thing can be obtained in this way.
So, the particular integral when I am writing, so this equation can be written d square x
2 by d tau square plus x 2. So, this is equal to half minus half alpha 2 a square by omega
0 square. So, omega 0 square is divided here. So, this is the equation. So, for this, when
one has to find the particular solution, so this one can write this thing as; so the auxiliary
equation, so one can write this d square plus 1 x 2, so this is the, this is equal to alpha
2 a square by 2 omega 0 square.
Then, this particular solution x 2 can be written as 1 by 1 plus d square into alpha
2 a square by 2 omega. So, this negat
…

Need subtitles from another video?