0% found this document useful (0 votes)
252 views

Lecture 16 - Matrix Method of Analysis of Trusses

The document discusses applying the matrix method of structural analysis to truss problems. It begins by reviewing the overall philosophy of breaking structures into small elements and writing the force-displacement relationship for each element. It then focuses on deriving the element stiffness matrix for individual truss members. The process involves numbering the joints and members, identifying the degrees of freedom at each joint, and determining the coordinates of joint locations. An example truss problem is shown and the document demonstrates writing the force-displacement relationship for an arbitrary member to obtain its stiffness matrix. Subsequent lectures will cover applying loads and boundary conditions to solve example truss problems.

Uploaded by

Akshay
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
252 views

Lecture 16 - Matrix Method of Analysis of Trusses

The document discusses applying the matrix method of structural analysis to truss problems. It begins by reviewing the overall philosophy of breaking structures into small elements and writing the force-displacement relationship for each element. It then focuses on deriving the element stiffness matrix for individual truss members. The process involves numbering the joints and members, identifying the degrees of freedom at each joint, and determining the coordinates of joint locations. An example truss problem is shown and the document demonstrates writing the force-displacement relationship for an arbitrary member to obtain its stiffness matrix. Subsequent lectures will cover applying loads and boundary conditions to solve example truss problems.

Uploaded by

Akshay
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 16

Matrix Method of Structural Analysis

Prof. Amit Shaw


Department of Civil Engineering
Indian Institute of Technology, Kharagpur

Lecture – 16
Matrix Method of Analysis Trusses

Hello everyone, welcome to the second phase of the course and the first phase which was
panning over first 3 weeks, we discussed some of the basic concepts which constitute the
premise of this method, method of matrix method of structure analysis.

There, we discussed the underlying philosophy of the method, and the underlying
philosophy is if we have a structure then we need to break this structure into small small
pieces, and for each piece for instance if it is truss then divide the entire truss into small
small each member. Then for each member similarly for beam and frames for each
component we need to write the force displacement relation using the stiffness definition.

And then once we have the forces first plane relation for all the members then we need to
assemble them to get the global system of equations and then we apply the boundary
conditions and solve them. What we will be doing in the second phase which is starting
today we will see how this is to be done for different structural components.

Today this week we will start with truss and then subsequent weeks we also see the
similar exercise for beams and frames. So, the what essentially we will be discussing is
the of course, how that method is to be a is to be translated for truss problem, and also
some of the implementation, some of the difficulties or the some of the implementation
issues that we may come across while applying this method, ok.
(Refer Slide Time: 02:03)

You see, so today is this is the 16th lecture we will start with application of matrix
method analysis for truss. And today’s topic is element stiffness matrix we will see how
the equation once you have discretized this, once you have divided the entire structure
into small small elements small small segments for each segment how to write the force
displacement relation and that is we are going to do it for truss today,

See here it is latent element stiffness matrix you can write it is member stiffness matrix
as well, ok.

(Refer Slide Time: 02:37)


So, let us take a truss problem. Now, if we the first step is we remove the take only this
truss, remove the loads, remove the boundary conditions, how to use the load and the
boundary conditions we will see in the subsequent lectures. Now, this is the structure we
have

Now, the first is the structure we need to break the structure into several elements in
these cases several members, ok.

(Refer Slide Time: 03:04)

Now, then for each member we for each member, for each member we need to write the
element for different members we need to write the force displacement relation means
the force is equal to is equal to stiffness into displacement right. This relation we do it for
every member and then assemble it and today what we see is we how to write this
expression for a given member, ok.
(Refer Slide Time: 03:39)

Now, let us take one example, this is an example this is a again a statically determinate
structure and we can solve it and this is the solution of this problem. But let us not right
now bother about this solution because we have not come to that stage where we get the
solution, let us first write the element stiffness member stiffness matrix.

(Refer Slide Time: 04:02)

Now, see the first thing is remove the loads and the and the and the boundary conditions.
Remove the load means it is just do not do not bother about loads and boundary
condition right now. We will bring we will see how the external load and the boundary
conditions to be taken into account that will seems a subsequent, subsequently, ok.

So, now, you see actually the if we look at the numbering of the member this is note this
is point number joint number A, joint number B, joint number C. But this is not the way
we identified joints in this method instead of writing ABC or XYZ what we do is we
write say this is this is a first joint number 1, joint number 2, and joint number joint
number 3 ok.

Now, now, once we have identified the joints and if you have a if you have say 4 n
number of joints in a in a truss then it has to be number like this 1 to n. Now, here one
point I just I am mentioning now, we will come to this point once again when we do the
numbering, numbering is also not very it is not very arbitrary way we should number
different joints. We will see that we will see that you have to number in a specific way.
So, that the matrix is not the end at the end of the data matrix that you that you that we
arrive that matrix becomes banded matrix. And what is banded matrix and all we will see
that.

So, at this point keep that in mind the numbering of different joints is not arbitrary. Yes,
you can number the way you want, but make sure there are certain requirements comp
computational requirement that needs to be satisfied while you do this numbering, ok.

Now, once we have this. So, let us number the members also. So, this member number 1,
member number 2 and member number 3. Now, once we have the member then let us
suppose at this joint at this joint let us, at this joint to every joints we have 2 degrees of
freedom, ok. So, suppose these 2 degrees of freedom is, so at this point we have one this
one horizontal displacement say u here and one vertical displacement say it is ux and uy,
but again we do not write the we do not identify the degrees of freedom by ux, uy, vx,
and vy and so on, instead of that let us they degrees of freedom it is if you are using uth
then it is u 1 and it is u 2. So, whenever we say that the first degrees of freedom in this
case it is the displaced horizontal displacement of this joint. Second degrees of freedom
of this structure is the horizontal vertical displacement at this joint.

So, similarly so we have we have say it is 3 4 and it is 5 6, the total 6 degrees of freedom
we have each degree, each joints 2 degrees of freedom, ok. Now, each degrees of
freedom similar to nodes or points and the member each degrees of freedom we need to
give an id and this is 1 2 3 4 at the different ids for the different degrees of freedom.

Now, once we have that the next is we need to take 3 members we have here 3 members
separately. So, this is member number 1, member number 2 and member number 3. Now,
see member number 2 we have 2 at this joint it is degrees of freedom 1 and 2 at this joint
degrees of freedom 5 and 6. Again for 2 point common point 2 is the common node
between member number 1 and main member number 2. So, naturally degrees of
freedom at node at point at this point would be 1 and this is will be 2. So, this is how the
degrees of freedom for different member, ok.

Now, we have to write the stiffness matrix or the force displacement relation for each
separate member, ok. Now, so if we take this length is the length of each is L, this is L,
this is L, this is L so we can find out what is the coordinates of each joints as well, ok.
These points are important we will be, using this point we can in this case it is a 60
degree, but it could be any arbitrary angle say theta. So, these, these, these coordinates of
these 2 points are equal to find out this angle theta and the length of this member ok.

(Refer Slide Time: 08:51)

Now, next is suppose take and any arbitrary as I said our objective is now, is to find out
the stiff find out the force displacement relation for each member. So, let us, let us find
out a force displacement relation for any member which is arbitrary oriented and once
we have the force displacement relation for that or the stiffness matrix for that member,
then depending on the degrees of freedom and depending on the coordinates of the
coordinates of the coordinates of the points we can have the similar equations for
different members.

So, let us take any member arbitrary member say member m and which the this is node i
and this is node j and the degrees of freedom at node i is i 1, i 2 and this is j 1 and j 2, ok.
It is you know depending on i 1 and i 2, j 1 and j 2 and depending on the value of m and
depending on the coordinate of this ith member and jth member we can have different
member in different orientation.

Now, take us for writing, but the ease in right for writing let us say this member is 1 and
2 and the degrees of freedom is 1 2 here and 3 4 here, ok. So, we will find out what is the
force displacement relation for this member. So, for so in this case we take i is equal to 1,
j is equal to 2 and i 1, i 2 is equal to 1 2, j 1 j 2 is equal to 3 4, ok. So, that we do not have
to write i 1, i 2, j 1, j 2 again and again, writing will be easier in terms of 2 3, 1 2 3 4, ok

So, now, suppose as I said these are the degrees of freedom u 1, u 2, u 3, u 4 are the
corresponding degrees of freedom. And then the forces at members force in 2 different
direction say it is P 1, P 1, P 2, P 3 and P 4; P 1, P 2, P 3 and P 4, ok. So, when we when
whenever we say P 4 it automatically means P 4 is the force in vertical direction at joint
2, ok.

Now, what we need to find out? So, this is your displacement u 1, u 2, u 3, u 4 and the
associated forces are P 1, P 2, P 3, P 4. So, what we are interested now is to find out a
relation between the force and the displacement. So, here we have 4 degrees of freedom
and the 4 corresponding 4 forces. So, these are the forces and corresponding associated
displacements are this, and how these forces and the displacement are related to each
other. And so, they are related to each other to a stiffness matrix k. You remember k is a
supers superscript 1, this k 1 means this is the node number this is suppose this is
element number 1 or member number 1, ok. Any arbitrary any arbitrary element any
arbitrary member we can write the stiffness matrix is say k e, where is the id of the
element in this case this is k m ok.

Whenever we say k m it means the stiffness matrix the member stiffness matrix for the
mth member. So, k 1 is the stiffness matrix of the first member number 1, ok. And
another thing is usually as you can use your notation, but usually we use a small k for
element stiffness matrix, element stiffness matrix, ok.

Now, today what we do is we need to we need to see how to determine this k 1 or k for
any arbitrary member.

(Refer Slide Time: 12:36)

Now, if you recall in the second class, second week we discussed this if we have a truss,
truss is a two force member. So, essentially you if you if you take any member we have a
force along the longitudinal axis of the member. Now, you see there are, now, let me
introduce two coordinate system, one is the local coordinate system and one is the global
coordinate system.

You see local coordinate system is the coordinate system along for instance if we take F
1 and F 2. If we assume a coordinate system like this if we assume a coordinate system
which is x is in this direction this is your x direction and this is your y direction, say it is
y dash and x dash because x and y you have already used for something else. So, the
local coordinate system local coordinate system ok. So, F 1 and F 2 these 2 forces are
with respect to local coordinate system, ok.

Now, if we have F 1 and F 2 two forces defining with respect to local coordinate system
and the associated displacements are u 1 and u 2 again in the local coordinate system,
then if you will you recall we discussed in the second way that F 1 and F 2 and u 1 and u
2 are related with this, and this is the stiffness matrix this is the stiffness matrix k. And
this stiffness matrix is the local stiffness matrix, this is local stiffness matrix, ok.

If we have a different member for instance, if we have a member like, if you have a
member like this, if you have a member like this, ok. And then in this member if we have
a if we have the coordinate system like this for instance this is this is this is x direction,
this is x dash direction and then this is y dash direction. And then corresponding forces
are say this is the corresponding forces this is F, F 2 and this is F 1, F 1 and similarly we
have u 2 and u 1, u 1 here, u 1 here. So, u 1, u 2 and F 1 if F 2 again will be related by
the same stiffness matrix, there will be no change in the stiffness matrix and this is
stiffness matrix is with respect to local coordinate, locals coordinate system.

But you know what at the, when we when we at the end of the day we will not be dealing
with local stiffness matrices local, force displacement relation we have to we have to add
assemble all the force displacement relation to get the global system of equation. And
therefore, we need to write the form of the element stiffness matrix the k, we instead of
this is with respect to local coordinate system, but instead of local coordinate system we
need to transfer them we from local coordinate system to global coordinate system, ok.

So, once we have the global coordinate system in this element stiffness matrix for global
coordinate system then irrespective of the orientation of the orientation of the truss
member always, ok. If we when we when we say that that force displacement relation
with respect to local coordinate system and global coordinate system it essentially means
when we say the local stiffness matrix it relates the forces and the displacement defined
in local coordinate system. When we say global stiffness matrix it means or the stiffness
matrix with respect the global coordinate it means that it relates the force and
displacement defining global coordinate system.

Now, at the end of the day we need to solve the entire structure right. So, entire structure
all the forces and the displacement is are which are defined with respect to a cord fixed
coordinate system and therefore, it is very important that we transfer this element
stiffness matrix which is written in local coordinate system to global coordinate system.
(Refer Slide Time: 16:46)

So, when you do that means now, we have to write the we have to write, you remember
last in the in the previous slide just now, we saw your equation was defined with respect
to this and this coordinate system and where we have the force one this. But now, this is
local and then this is global with respect to global coordinate, where the displacements
the degrees of freedom in the forces are defined with respect to global coordinate system
which is x and y here.

And that is the reason why at every joints you have two, the two displacement, one is
global x direction, global y direction, global x direction, global y direction and similarly
force is also a global x, global y and global x, global y direction, ok.
(Refer Slide Time: 17:41)

Now, let us do this. Let us um. So, this is the thing that we need to find out and that we
discuss this we have to do now, ok. So, let us do let us do that.

(Refer Slide Time: 17:59)

You see suppose take any. So, what we know is if we have say take first take a member
and in this member we have this is the coordinate system where we have the x dash and
this is y dash, ok. And then we have defined here is u this is say u 1, u 1 dash and this is
u 2 dash. This is your node number say this is node number 1 and this is point numb joint
number 2, ok.
Now, now, we have the same thing here we have this is u 1 and then this is u 2 then u 3
and u 4, this is node number 1, this is number 2, ok. And suppose this angle is theta
suppose this angle is theta, ok, and what we know is what we know is that F 1, F 1 and F
2 the this is equal to ae by L 1 minus 1 minus 1 1 and this is u 1 dash u 1 dash and u 2
dash, right.

Now, you see you can we write now, u 1 this though they are written in terms of different
coordinate system u 1, u 2 and u 1 dash they are they are defined here in terms of
different coordinate system, but they have a relation, right. Because essentially they both
tell us what is the movement or what is the displacement of joint 1. So, we can find out a
relation between u 1, u 2 and u 1 dash and that relation is very straightforward we can
just apply the we can just take the components of u 1 and u 2 along this direction. And if
we do that we can write u 1 dash is equal to u 1 cos theta cos theta plus, plus u 2 sin
theta. And similarly u 2 dash is equal to u 3 cos theta plus u 4 sin theta.

And now, this we can write as u 1 dash u 2 dash which is this is equal to cos theta, sin
theta, 0, 0, 0, 0, cos theta, sin theta and then this is u 1, u 2, u 3 and u 4 ok. Now, suppose
cos theta is equal to let us take lambda x is equal to cos theta and lambda y is equal to sin
theta and then we can write these expression these expression as this is called, this is say
this is called this is called T. T is the, T for transformation what it transform it is a
transformation it gives it relates the displacement defined in one coordinate system to
displacement define it another coordinate system. So, this is transform that you how the
transformation of displacement can take place ok, that is why it is called transformation
matrix.

Now, so T can be written as once we have this then. So, T we can write. So, T is equal to,
so T is equal to lambda x, lambda y, 0, and 0, 0, 0, lambda x, and lambda y ok. This is
very important, ok, we will come to this expression once again or rather several times ok.

So, what we have done so far is we have displacement in one local coordinate system
and the displacement defining degrees of freedom rather defining global coordinate
system, then we have seen how these 2 displacement can be transformed, ok.

Now, let us do the same exercise for forces as well and if we do that then forces also can
be written. So, let us take, let us take let us let us take we can go we can we can we can
suppose now, the forces will be here it is say F 1, it is F 1, F 1, and this is F 2 and
similarly here it is F 3 and F 4 right. And here it is say sorry instead of F 1 you write P 1
because that is the that is the that is how we define forces and in local coordinate system
it is F 1 that is how we defined and it is F 2, ok.

Now, this gives you a relation between u 1 dash u 2 dash and u 1, u 2, u 3, u 4 what we
do the similar exercise which relate F 1 and F 2 with P 1, P 2, P 3, P 4. Now, if we do that
if we do that then you check that you can write like this P 1 is equal to P 1 is equal to F 1
cos theta and P 2 P 2 is equal to F 1 sin theta. Similarly P 3 is equal to it is just the taking
the component of forces F 2 cos theta and P 3, P 4 is equal to F 2 sin theta, F 2 sin theta.

(Refer Slide Time: 24:35)

Now, you recall if I write P 1, P 2, P 3, P 4; P 1, P 2, P 3 and P 4 and that become if we


substitute cos theta as lambda x and sin theta as lambda y. Then this become lambda x 0,
then lambda y, 0 and then 0, lambda x, 0, lambda y and then F 1; F 1 F 2, ok. So, so this
directly comes from these two equation, ok.

So, now, again this is if you recall T was T T was like this T was lambda x, lambda y, 0,
0, 0, 0, lambda x and lambda y this was transformation T. So, naturally this becomes T
dash, ok. Now, P is related to F 1, F 2, F 1, F 1, P 1, P 2, P 3, P 4 related to F 1 F 2
through T dash, ok.

Now, next if recall that F 1 F 2 is related to u 1 dash u 2 dash with this. So, if I substitute
in the in the in this expression if I substitute F 1, if you substitute F 1 F 1 F 2 from this
what we have is what we have is, so this can be written as now, so, this can be written as
say this is P is a vector which consists of P 1, P 2, P 3, P 4 is equal to this is say T dash
which is transformation of T. And then F 1 F 2 F 1 F 2 was AE by L, AE by L into u 1
dash and u 2 dash ok, which is directly from this expression, a 1, a 1 not only that we
have another term here which is which I forgot to write. Let us erase it. It should be, yes
it should be 1, minus 1, minus 1, 1 and then u 1 dash and u 2 dash, ok.

Now, once again you remember last u 1 and u 2 dash u 1 and u 2 dash is related to u 1, u
2, u 3, u 4 by this T. So, what we can write is we can substitute u 1, u 2, u 3; u 1 dash u 2
dash as so this becomes P which is this and then we can take AE by L out and then
transformation T dash and then 1, minus 1, minus 1, 1 and then this become T into u 1, u
2, u 3, u 4, right.

So, essentially this becomes P is equal to your P 1, P 2, P 3, P 4. So, let us write P 1, P 2,


P 3; P 1, P 2, P 3 and P 4 this is something k into u 1, u 2, u 3, u 4 and that k this k is this
the entire thing is this k, entire thing is this k ok. So, this is this k.

Now, if I substitute T dash T dash and T, T from this expression and T dash from these,
these T and T dash are transpose and do this operation matrix operation and then we get
an expression like this. We get an expression like this. We get an expression like this, ok.

(Refer Slide Time: 29:54)


So, for any arbitrary for any arbitrary member any arbitrary member if the it was 1 and 2
if you recall in our in our the expression that we derived it was 1, it was 1, it was 2 and it
was 3 and it was 4, and it is it was 1 and it was 2 something like this. So, it is 1, it is 2, it
is 3, it is 4, 1 2 3 4 and this is, this is 2, and this is 1 this is 2, ok. So, if we do this we
have the expression for T dash and the and if we substitute that then we get k m is equal
to or element or member stiffness matrix for any arbitrary member m member stiffness
matrix is equal to this, ok.

Now, this i and j depending on the depending on the, depending on the, depending on the
your member i j will be and the i j and these degrees of freedom will be different. For
instance, if we have if we have for instance for if we go back to; for instance for this case
for if we consider this member your i is to, your i is 1 and j is 2 and degrees of freedom
are this is 5 6 1 2 and for this i is 2, j is 3 and degrees of freedom is 1 2 3 4.

So, similarly, so the stiffness matrix what we get k 1 this relates degrees of freedom 5 6 1
2 with the forces 5 6 1 2. Similarly we get k 2 here which is degrees of freedom for
member 2 which relate displacement u 1, u 2, u 3, u 4 with the forces P 1, P 2, P 3, P 4.
Similarly stiffness matrix k 3 that we get here that that relates forces u 3 P 3, P 4, P 5, P 6
with the displacement u 3, u 4, u 5 and u 6. And if we just write a general form of the
stiffness matrix then that general form will be that general form is this, that general form
is this, where lambda x is the lambda x and lambda y are this is the angle if we take this
angle theta, ok.

Now, this is the stiffness matrix for a given member this is also called member stiffness
matrix, ok. And now, you see there are stiffness matrix in this case has to be 4 by 4
matrix because we have 4 degrees of freedom and this was 4 forces. So, this gives you
the stiffness matrix 4 by 4

Now, let us just before we before we stop let us try to understand what the interpretation
of each element in the stiffness matrix for instance you see lambda this is this means
these relates i 1 to i 1, ok. Means it is the relation between force in this direction and the
displacement in this direction. And suppose these lambda x this relates i 1 and j 1 this
relates force in i 1 direction and displacement in displacement for j 1. Similarly, similarly
if it is similarly if we take this they need relates also corresponding force and
corresponding displacement.
Another important observation that you can see from the stiffness matrix the diagonal
terms all the diagonal terms are positive and the stiffness matrix is symmetries. What is
the interpretation physical interpretation of stiffness matrix symmetric you can recall
structural analysis one you study reciprocal theorem it is very similar to that, ok.

And what is the physical interpretation of all these diagonal term being positive, and why
whether the diagonal terms can be negative or not, what happens, what is the properties
of the of the stiffness matrix, whether the stiffness matter is singular or not or is there any
other properties the stiffness matrix stiffness matrix has any element stiffness matrix has.
Those things we will be discussing towards the end of this course when we have when
we discuss various implementation issues, ok.

So, I will stop here today next class what we do is next class once we have the element
stiffness matrix or the member stiffness matrix. Next class we will see with one example.
We compute the member stiffness matrix for different members and then how to
assemble all these stiffness matrices to get the global stiffness matrix, ok. Stop here
today. See you in the next class.

Thank you.

You might also like