Indian Institute of Technology Kanpur

National Programme on Technology Enhanced Learning (NPTEL)

Course Title
Basics of Finite Element Analysis

Lecture – 11
Gradient and Divergence Theorems
Part-II

by
Prof. Nachiketa Tiwari
Dept. of Mechanical Engineering
IIT Kanpur

Hello, again welcome to basics of finite element analysis. In last class we had introduced
two different theorems, the gradient theorem and the divergence theorem. And today
what we will do is, we will explore these theorems further, and see how we can apply
them in context of weighted residuals. And how can -- we can use theorems to shift
differentiability operator from one function to the other function.

And this technique becomes very useful in context of problems when they are two-
dimensional or three-dimensional in nature.

(Refer Slide Time: 01:00)

So let us consider a function G times w, so G is a scalar and w could be a weighting
function okay. And then we take its gradient and we integrate it over the whole surface.
So I can express it as integral and I will first -- I apply operator on G and then the second
case I will operate it on w. So it is w times gradient, so this is a vector of g + g times
gradient of w, dx, dy.

So I can write this as w times gradient of dx, dy integral equals integral of gradient of G
times w minus integral of G gradient of w, dx, dy Ω. So now we continue this thing and
we know that the left side can be written as integral of Ω, so this is equation a1, so it is
actually not Ω it is w times gradient of G times dx, dy is equal to using the gradient
theorem.

So this we keep it as is, and here we apply the gradient theorem. So now we have
developed this equation.

(Refer Slide Time: 03:22)

And then the next thing we do is that we apply the gradient theorem on this particular
term. So what we get is left side does not change, w times gradient of G times dx, dy
equals, and here I am going to apply the gradient theorem, but before that I will write this
term here, so I am also reorganizing my terms g times grad of w dx, dy and for this term I
will use the gradient theorem.

So this is our domain integral and I can express it in terms of boundary integral using the
gradient theorem. So that term comes to plus boundary integral of G Ω not Ω, I am sorry
w. And there should be a unit normal here times ds okay. So this is one equation okay, so
what do you see in this equation that here the gradient was applied on the scalar G and
that has gotten shifted to w variable.

So if w is the weighting function then it gets shifted to w, and then of course we have a
boundary integral. So this is one important statement and we use this to transfer
differentiability. So in one dimension we use integration by parts so this is somewhat of a
similar process here also.

(Refer Slide Time: 05:31)

Because we start from here and then we shift differentiability operator are not necessarily
differentiate with differentiability, but the gradient operator from one function to the
other function.

(Refer Slide Time: 05:44)

So this was one equation and now we will develop another equation.
(Refer Slide Time: 05:51)

So consider this, so suppose there is some function w times gradient of G, and I take a
divergence of this entire function. So it is a dot product, so this is equal to -- first I
operate it on w, and then inner product with G. So this is 1, so gradient of w and gradient
of G, the inner product of these two plus w times grad square of G dx, dy. So I am going
to rearrange this w times grad square of g dx, dy Ω equals this is there, so the G is dy
minus inner product of gradients of w and g dx, dy.

Now what I will do is, I will apply the divergence theorem on this guy. So this is the
vector, because w times gradient of G is a vector. And when I operate this with the grad
vector and take its inner product this is basically the divergence of that vector right, and it
is the integral of the divergence of that integral of that vector. So that is nothing, but the
boundary integral of this thing ds minus -- and the right side left side does not change.

So this is the second equation okay, so here also this is the boundary term, so several
times you may have seen that there is a differential equation something like this grad
square FE equals some constant r0, some physical phenomena is defined like this in heat
transport equation, diffusion equation, several equations. And if you multiply that by the
residue w, and then you do all this process then this is the right side you have the gradient
operator being applied on w and on G.

And again you have a balanced system the differentiability requirement again become
balanced. So here its second order differentiability requirement, here its first order
differentiability requirement. And if its fourth order equation then you can get iterate it
and you can get similar equations. And then of course each time you do this you get a
boundary term.

So this is the analog of what we learnt two lectures earlier in context of 1d systems.
Whereby using the integration by parts we were able to shift differentiability operator
from one function to the other function.

(Refer Slide Time: 10:21)

Here we are doing somewhat same thing using divergence and gradient theorems. So this
gives you a mathematical foundation how you can shift the differentiability operator from
one function to the other both in one-dimensional domains, and also in for two
dimensional systems. And you can use similar mathematics for three dimensional
systems as well. So this concludes our lecture for today we will again meet tomorrow and
that will be the last lecture of this week, thank you bye.

Acknowledgement
Ministry of Human Resource & Development

Prof. Satyaki Roy

Co-ordinator, NPTEL IIT Kanpur

NPTEL Team
Sanjay Pal
Ashish Singh
Badal Pradhan
Tapobrata Das
Ram Chandra
Dilip Tripathi
Manoj Shrivastava
Padam Shukla
Sanjay Mishra
Shubham Rawat
Shikha Gupta
K. K. Mishra
Aradhana Singh
Sweta
Ashutosh Gairola
Dilip Katiyar
Sharwan
Hari Ram
Bhadra Rao
Puneet Kumar Bajpai
Lalty Dutta
Ajay Kanaujia
Shivendra Kumar Tiwari

an IIT Kanpur Production

©copyright reserved