Structural Dynamics

Prof. P. Banerji
Department of Civil Engineering
Indian Institute of Technology, Bombay

Lecture - 10
Pulse Loading

Hello again, so we have finished looking at in the last lecture on how to deal with
periodic loading, so till now we have been looking at loads, which are essentially
therefore very long duration. In other words we know we only consider a situation,
where the loading is there. For example, harmonic loading only think that we says that
harmonic load is that time period, but then the number of cycles we know goes on
forever. Similar, if a periodic load also we say that we do not know when it started, we
do not know when it ends, we just taking particular period during the time taking a
window.

(Refer Slide Time: 01:17)

Today, we are going start looking at something that is known as pulse loading, pulse
loading is essentially it is a short duration load. In other words the loading starts at a
specific time and ends at the specific time, and what are pulse loads, these are typically
what do we mean by short duration loads.
(Refer Slide Time: 01:50)

Short duration essentially implies that if I call t d, as the duration of the pulse, this is of
the same order of magnitude as the time period of the structure, when I say same order of
magnitude I essentially mean that they are similar to each other. It is not as if t d is equal
to it just means that t d could be t by 2, could be t by 4, could be 2 t, but not more than 2 t
for sure.

So, if you say that then what happens let us look at, so let me take and here what I am
going to look at is we are going to look at mathematically defined loads. So, that we can
get close form solutions and let us start off with the simplest kind of load, and that is a
rectangular pulse. So, this is starts at 0, this end the t d the loading is p 0, so essentially it
is starts at 0, and it is p naught till t d and then t d it goes to 0. Obviously, the rise and the
fall or instantaneous, this is the mathematical abstraction reality all we say is that the
risen fall are very short compared to the time duration.

So, in other words p of t is equal to p 0 between 0 and t d, and 0 for t less than 0 and t
greater than t d, so now, the question becomes well what how do we solve this problem.
Well, immediately there is no difference in the I mean you know we know, how to solve
any you know partially differential equation. We know all of that, but there is there are 2
interesting aspects to this problem, one that I have to solve this is the 2 phases problems,
phase 1 and phase 2.
Phase 1 is known as during that period, when the duration when the loading is non 0 and
phase 2 is after the duration of the load, we have what is known as a free vibration phase,
so this is forced vibration, and this is free vibration. So, now, let us look at this particular
problem here what we are going do is we are going to solve for this rectangular pulse, we
are going solve both the phase 1 and phase 2 problems. And note that again our sole
interest here is in getting the peak value of the response, that is all we are interested in
because we are not interested in the time duration.

Now, interestingly enough one particular point that requires and this we say when we are
look at damping, remember when I was looking at free vibration I say that damping takes
some time to start off. And the viscous damping after 1 or 2 cycles of vibration, that is
only the time when viscous damping takes of now since, t d is at the same order as t we
are not going to get two cycles of vibration at all. So, for all practical purposes we can
solve this problem as an un-damp problem both for the force vibration and free vibration,
because again understand we are not interesting in the time history at all, we interested
only in the peak value why because the peak value is what we need to design a structure
for it.

(Refer Slide Time: 06:41)

So, let us now look at it, so the problem is no longer the solution of m u double dot plus c
u, but we can drop the c u term, why not because c u is 0, but because the viscous
damping takes time to happen and t d is not is not sufficiently long enough for the
loading to happened. So, this is the problem that we solve and the problem here is one of
now, note that since it is a short duration load u 0, that is and u dot 0 which are what,
initial displacement, initial velocity.

These are important and we will solve the problem as A, so I am looking at phase 1
solution right now, so how do I solve it particular solution, particular solution is p naught
upon k and the homogeneous solution is well u is equal to a sin omega t plus b cosine
omega t plus p naught upon k and A and B I get from here. So, if you look at it if I
differentiate it u dot is equal to A omega cosine omega t sorry, minus B omega sin
omega t in the plus note that these is a constant, so differentiation is 0. So, if we put u 0,
since sin omega t, t equal to 0.

What we get is B cosine omega t plus p naught upon k is equal to 0? And this cosine
omega t is 1, so you may get that b is equal to minus p naught upon k and again if we
took u dot 0 is equal to A omega which equal to 0 this implies that a equal to 0. These
implies that u of t is equal to p naught upon k into 1 minus cosine omega t, so this is the
solution of the problem and again this is the reason I am writing now u dot of t and these
is equal to p naught omega upon k into sin omega t, so this is the solution.

(Refer Slide Time: 10:13)

Now, that this solution means what, let us look at it and that is that we have a situation,
where phase 1 is u of t is equal to p naught upon k into 1 minus cosine, let us look at
phase 2. Phase 2 is what well little put down u dot t only here, is equal to p naught
omega k sin omega t, so now, if we if we you put together what do we get we get it the
following phase 2 basically becomes a problem on m u dot plus k u is equal to 0. Why?

Because, essentially you have situation where you know, again we looking at only peak
value, so the peak value will happen in first one or 2 cycles, so damping does not come
into picture, so this we know the solution to this solution is and where you have u of t d
and u dot of t d as initial conditions. So, then I can write I know the solution to this, these
if I just put it shift from t d and this solution basically becomes, u dot t d upon omega
into sin omega t plus u t d cosine omega t, again the values itself is irrelevant, so what we
want is we only want the peak value.

So, the u peak is equal to u t d squared plus omega the whole square, so that is the peak
value, in phase to this the peak value, so I can plug in these are my response, I can plug
those in and if take p naught upon k outside I get u dot t d is this plus whole squared. So,
these basically becomes 1 minus 2 cosine omega t plus cosine squared omega t and u
upon omega, so this is omega again p naught goes, so this becomes plus sin squared
omega t d that is the peak value. So, if I look at this becomes 1, so it becomes 2 minus
square root, so if I look at that particular solution this is 1, so it becomes 2 minus 2
cosine squared.

(Refer Slide Time: 13:49)

So, this becomes peak value becomes u p is equal to p naught upon k 2 minus 2 cosine
omega d t squared, so this is equal to from trigonometric equality, this becomes 2 p
naught upon k into sin. Now, it becomes omega upon 2 sin omega upon 2, so omega is 2
pi upon t, so this becomes sin pi t d upon t so; that means, if this is the peak value, let me
just write this down. So, the peak value in phase 2 is equal to 2 p naught upon k into sin
pi t d by t, and if I rewrite this as dynamic factor, then this is 2 sin pi t d upon t.

These basically comes from the fact that this into this essentially becomes 2 times 1
minus cosine is equal to sin squared pi you know omega t d upon 2, so that is way this
becomes square root of sin squared becomes a you know the 2 two here. And so it
becomes 2, so it becomes four square root of 4 is 2 and so this is what at comes out, so
these is the dynamic amplification factor if peak is in phase 2.

Now, let us look at phase 1, suppose when will now let us look at the phase 1
displacement, understand that for the maximum to occur in phase 1 this has to equal to 0,
do you understand that. So, I am going to write the following and that is that we have a
situation where the peak value for the peak to line phase 1 p naught omega upon k into
sin omega t is equal to 0, and omega is what omega is equal to, so now note for this to be
equal to 0, since these cannot be 0 this has to be equal to 0.

(Refer Slide Time: 16:39)

So, in other words let me rewrite this sin 2 pi t upon t is equal to 0, so what does that
mean this implies that your 2 pi t upon t is equal to let us look at it what does 0 imply, 0
can be well it cannot be 0 because t equal to 0 is 0 is a trivial solution. So, it has to be
what is the next one, it is n pi now; obviously, now the duration is small, so which is the
first one, so in other words if t d by t is greater than look n pi the first one is pi upon 2 is
half.

So, if t d by t is greater than half then peak can occur in phase 1, and what is that peak
then well 2 pi upon t, so omega t is equal to pi. So, if I take omega t is equal to pi look at
this, this is cosine omega t, so cosine of pi what is cosine of pi minus 1, so this is
becomes 2. So, you are u peak is equal to 2 p naught upon k, so this implies that if t is
greater than half, then peak can occur and this is the peak; however, if t d by t is less than
half peak in phase 2 and what is that peak value the peak value is equal to.

So, in other words D is equal to 2 and peak is in phase 2 D is equal to 2 sin pi t d upon t,
note very interestingly that if t d by t equal to half. Then where does the peak occur if t d
by t equal to half then the peak occurs at t equal to t d and what is the value put t d equal
to t here what you get sin pi, sin pi is what sorry, made a mistake here. Here, is if t d by t
is equal to half, this is sin pi by 2 sin pi by 2 is 1, so d is equal to 2, so it exactly is D
equal to 2. So; that means, what we have is if t d by t is greater than or equal to half D is
equal 2 and if t d is less than half then D is equal to, so we have solved the problem of
the peak.

(Refer Slide Time: 21:07)

And if you look at this, here you look at this one, this one is a sinusoidal function which
is this sinusoidal function, goes here at t d upon t equal to half it is 2 which is what we
get here. And after that if it is longer duration, then t d is equal to 2 which is what we get
here, so this one you look at it, this is what is known as the shock spectrum for a
rectangular load.

So, shock spectrum essentially if I give you this, if you know t d the duration of the load
of a rectangular load and you know t you just go here read it of and that gives you the
dynamic amplification factor and multiply that by p naught upon k and you have your
response. So, in other words the shock spectrum you know replaces I have done the
mathematics here, over here I have done the mathematics, but if I had this I did not need
the mathematics the shocks spectrum if someone give me the shock spectrum I could not
then look at it and say this is my duration look at it read it that is the duration. For shock
spectrum actually gives you us full idea about how to go about it, so now, I will just do
another specific duration short duration load, before I move on to some other kind of
loading interesting loading that you can do and that is let us look at the shock duration.

(Refer Slide Time: 23:03)

Another kind of shock duration load is the one that I have draw here which known as the
half sin curve, so let us try to get these shock spectrum from all mathematics, what is the
half sin, half sin is essentially this. Where this is duration this and this is given as this is p
naught, so p t is equal to this is p of t p of t is equal to p naught sin pi t upon t d for 0 less
than and is equal to 0 for t less than 0 and t greater than t d, this is my half sin pulse.

So, now, how do I solve for this well again, I go through the same step solve the phase 1
solution and phase 1 solution for phase 1 is basically 0 t d this. Actually, this particular
case we can do this and the solution and of course, you know we have the fact that u 0 is
equal to u dot 0 is equal to 0 at rest initial conditions. And therefore, u of t is equal to and
again I am solving what the un-damp problem, so un-damp problem becomes A sin
omega t plus B cosine omega t plus the solution of this, the solution of this is nothing but
if I put this in and solve by this thing.

I already know what this is, this is equal to plus p naught upon k into 1 minus beta
squared, beta squared is omega. Now, the omega is what pi upon t d is omega upon
omega, which is 2 pi, so this becomes what it becomes t by t d squared, so I am going to
put this as 1 over 4, because I have remember this pi upon 2 and the other one is 2 pi.

So, omega bar is pi upon 2 t d and omega is equal to 2 pi sorry, pi upon t d omega bar is
pi upon t d and omega bar is 2 pi upon t, so beta is equal to pi upon t d upon omega bar
upon omega, so this becomes then t upon 2 pi. Pi, pi cancels, it becomes 2, so this is
what it have has and then into sin pi by t t d, this is my solution of the first equation and I
can solve for A 1 and A 2. If I solve for A 1 and A 2, I get that B sorry, B is equal to 0
and because note that if I put t equal to 0, this one becomes b this goes to 0 this goes to 0,
so therefore, B is equal to 0, when, u 0 equal to 0 implies that B is equal to 0 and from u
dot equal to 0, I get that A 1 is equal to minus T by 2 d into p naught by k upon 1 minus I
am going take this inside t upon 2 d the whole square, so this is A 1. And so therefore, if
I look at the total solution this becomes the following, I am just now going to substitute
A 1 here and take out the p naught upon k outside.

So, this is going to be equal to u of t is equal to p naught upon k and then note that I have
1 upon 1 minus t upon 2 d this is 1 minus beta square, you already solve this I am not
going to solve it again, so that one also comes out. So, inside I have what I have sin pi t
upon t d minus T upon 2 d into sin 2 pi by t, and again this 2 pi by t can be actually
written as T d upon t sin pi the radial matter, let not a confuse the issue here, so this is the
solution in phase 1, for a half sign pulse.
(Refer Slide Time: 28:12)

And once we have that I can actually get u dot also u dot turns out to be equal to it is
easy to do p naught upon k and I will get pi upon t d and so that is 1, so that is cosine pi t
upon t d minus cosine of 2 pi t upon T, so this is what I get as my solution for the
equation. So, now, let me first start of by, so we why do I do the u dot t in the second
phase, well that is because I mean sorry, in the first phase, because if the if maximum is
in phase 1. Then this has to be equal to 0 and this is equal to 0, I am sorry this is this 1
minus t upon 2 d, this is still there and so over here what you have is.

So, I get cosine pi t upon t d minus cosine 2 pi t upon t has to be equal to 0, that is the;
obviously, thing that you will land up getting and in other words, so note that for this to
happen, this has to be equal to this, because this minus this as to be equal to this. So, if
this has to be this all; that means, is that pi t upon t d is equal to 2 pi n plus or minus
sorry, this is the one, so just this wrong.

This is not correct what you get is this one comes on this side, so it is basically what your
getting is t 2 pi t is equal to pi t upon t d plus or minus plus 2 pi n, and note that this one
will this is valid for the first n, so it is 2 pi. So, if I do that I get 2 pi t upon t is equal to pi
t upon t d plus 2 pi, so then I get this side to be coming in and I can solve for t, over here.
So, any way I mean this is either here nor there this one, I need to solve this for time t
depending on my t d and t, so this is the time at which I have the peak and then that one I
plug back into here and I will be able to get peak value of the displacement.
So, I am not going to substitute all of those, but so this becomes little bit more complex
if you look at this, this one does not follow as clean a line as we got for the other one, it
is a much more complex a solution. Now, the question then becomes that can I substitute
in here, so I can just put over here and I get I can just take out the pi and I get t upon t is
equal to sorry, 2 t I will just make it t upon t is equal to t by 2 t d plus sorry, is 2 pi is
gone because I have, so plus 1 minus.

So, if I take this in I get t upon t plus t upon 2 d t is equal to 1 and therefore, I can just
say that t into 1 plus T plus 1 by 2 t is equal to 1, so that is I am sorry that is what we get
here. And I can solve for t by saying that look t is equal to 1 upon 1 plus t upon 1 plus 2
d, so now, depending on the omega and time duration and the time period we can
compute the time at which this happens and then you plug this in here and you will get it.
So, that is that is when it happens in phase 1 now the question comes that what happens
in phase 2, well let see what happens in phase 2.

(Refer Slide Time: 36:48)

In phase 2, what are we know that it is going to be u max is equal to u t d squared plus u
dot t d upon omega the whole square, so we have to do this. So, we have to find out what
u of t d is, so u of t d substituting into that becomes p naught upon k 1 minus t upon 2 t d
the whole squared into now sin of pi because t is equal to t d. So, t upon t d is 1, sin of pi
minus t upon 2 t d into sin, now you see this is equal to 2 pi t d upon t, so sin of pi is
what is sin of pi 0. So, what we get is u of t d is equal to minus p naught by k into t upon
2 t d all over 1 minus y of 2 t d squared into sin 2 pi t d by t this is u of t d and similarly
substituting in here.

What we get it as is if I put t d t is equal to t d because cos pi and this becomes cosine t d
upon t, so this essentially a substitute that u dot of t d is equal to p naught upon k pi by t
d upon 1 minus t upon 2 t d the whole squared into cosine of pi what is cosine of pi
minus 1. And so therefore, cosine of pi is minus 1 sorry, the cosine of 0 is 0 cosine of pi
is 1 sorry, it is 1, so therefore, what we get is cosine of pi is 1 minus cosine of 2 pi t d
upon t. So, that is what we get sorry making a mistake here, cosine of pi is minus 1, I am
very sorry extremely sorry cosine of pi is minus 1, so it is minus 1 here and minus here.

So, u d and u t d is these, so now, if I write this is getting fairly complex, but just stay
with me what you get is if I do u dot squared, and u dot is divided by omega. So, let me
see what I get as u t d by omega implies, actually u dot t d into t upon 2 pi, so essentially
if I put this in your upon omega this becomes essentially t upon 2 pi, so this becomes
again t upon 2 t d.

(Refer Slide Time: 41:12)

So, therefore, essentially what you get is u max in phase 2 is equal to you got I am going
to take the entire p naught upon k into 1 upon 1 minus t upon 2 d the whole squared, I am
going to keep this entirely outside and inside what I have is the following. I have sin
squared 2 pi t d upon t and over here multiplying this into t upon 2 d the whole and on
this side I have minus minus I get plus 1 plus cosine 2 pi t d upon t the whole squared.
And inside I have I have t upon 2 t d the whole square, again this is there in both of them
second take that out, so this becomes again p naught upon k I will put t upon 2 t d outside
inside 1 minus t minus 2 t d squared and inside I have 1 plus cosine sorry, 2 cosine 2 t d
by t. And then plus cosine squared 2 pi t d by t plus sin squared 2 pi t d by t these 2
become 1, so this becomes and then of course, square root I forget this square root here is
equal to p naught upon k into t upon 2 t 1 minus t upon 2 t d squared then into 2 plus 2
cosine 2 pi x t d square root.

This become if you look at this becomes 2 cosine of it becomes 2 into 2 cosine squared
pi t d upon t x and so if I do this square root of that this becomes equal to p naught upon
k into t upon 2 t 1 minus and the 2 comes out. So, I have essentially I will just re-write it
t upon 2 t d whole squared multiplied by 2 the 2 into 2 4 comes out square root is 2, and
the square root of cosine squared is cosine of pi d upon t, this is your u max, if max is in
phase 2.

Now, you know in the other case we got a beautiful situation, where we could find out a
value of t upon t d or t d upon t remember at in rectangular, if t d upon t is less than half.
It will in phase 2 and we know the value and if t d upon t is greater than half, then d is
equal to 2 here if you note that you cannot really do that purely because this process.
This is the t at which u t is maximum getting, this is the fairly complex a process, so
what we do is how do we solve this problem, how do how do I plot this red curve which
is for the half sin wave.

Well, the easiest thing is I know that at any time t d upon t I solve this, this I can solve
given t d upon t I can solve for this and get of course, u max upon p naught upon k that
gives me the dynamic amplification. This is completely dependent on t d by t for a
particular value I know what this value is, so this is the maximum if it occurs in phase 2
then for same t d by t I can solve for this one. So, if solve for this one this one
essentially, becomes something like I could rewrite this in this form that t is equal to 1
upon t into 1 plus.

So, note that this time duration does not only depend upon t d by t, it also depends on t
itself, so therefore, this will be this maximal occur at different times for different t d’s t d
by t. So, therefore, all you need to is find out this for given t d by t find out this for a
given t d by t and of course, time you know you know that you get this t and for that t
substitute into this and get the dynamic amplification which is u t upon p naught u max
by substituting that, t in here.

And get that value and compare this value with that value, whichever is the what will it
be whichever is the larger value is the one, that is operational. Typically, when you are
very, very low very, very small this operates because when t d by t is very much less
than 1, then you know it does not get a chance to do this. And therefore, since it does not
get a chance to do this one it will always be this, so if you look at this particular part for
a small part it is this, but once t d by t becomes slightly larger than what happens? When
you have to compare this and this because of the solution that this solution does not only
depends t d by t it also depends on t itself. So, that is the region why you find out both
whichever is maximum is the value that you get, so much for this and similarly what we
can do is if we come here.

(Refer Slide Time: 49:46)

We draw another curve for a symmetrical triangular is typically of this form, I will just
show it to you a symmetrical triangular is I am not solving this, I leave it up to you to
solve with if you wish. To, p of t, t symmetrical triangular is t d t d by 2, p naught why
do not you people you take this is an exercise, take this is an exercise and find out this
show that this is the shock spectrum for a symmetrical triangular. Of course, again for
this also you know the t d by t a value for which it is in phase 1 and phase 2 is not
obvious of course, in the smaller part it will be in phase 1 that know for sure, because
when t d by t is very small the duration of the load is not large, enough for it to be
occurring in phase 1, so this area is always in phase 2. However, just like in this case we
had half in this we did know when it was coming in this case also it is going to be
slightly problematic, but you can plot it you can put in values of t d by t and t and you
can get it.

So, that is what is known as pulse load, I have shown you that in pulse load short
duration loads you did not only have one solution like we use to have for harmonic and
periodic you have phase 1 solution phase 2 solution and how to compute the peak I also
showed you. For the rectangular we knew exactly when it will occur in phase 1 minus
occur phase 2 for the half sign I showed you well, for phase 2 you can get a close form
solution for phase you should not a close form solution for the peak value, and so you do
not know where it is.

However, the point that I am trying to make is that you do not have to solve anything,
any mathematical solution, if your given the shock spectrum and you have given the
particular kind of load. If you know what kind of load you have you can always read of
here p d upon of 1.75 read 1.5 it is symmetrical triangular. And this is the value it is
mean that and multiplied by p naught upon k and you have your may peak responses that
is all you need, so shock spectrum actually is very, very useful for pulse loads were you
do have solve any problem.

Thank you very much, see you next time, bye.