University of Florida/Eml4507/Fead-s13-team4-R2

Problem R2.1 Model 2-D Truss With 2 Inclined Elements and Verify Assembly

edit

Given 2-D Truss Model and Constants

edit

Consider the 2D truss system

 
Figure 1: 2D Truss system analyzed in this problem. ([1])


Use the following constants to very calculations

 

 

 

 

 

 

 

 

 



Find:

edit

1) Explain the Assembly Process of the Above 2D Truss System

edit

2) Verify With CALFEM

edit

Solution: Assembly Process for 2d Truss System

edit

Construct local stiffness matrices

edit

The stiffness matrices will be constructed for individual elements 1 and 2. The rows and columns will be numbered to represent the force and degree of freedom. For element 1

 

For element 2

 

The variable K can be determined by

 

If the variables are plugged in, we get the following matrices:

For element 1

 

For element 2

 


It is clear that if we wish to combine the matrices, it will place coordinate 3,3 of element 2 on the location of 3,3 of element 1 due to them referencing the same point in the same direction. When combining the matrices, addition will be used since the matrices represent directional forces and forces always combine in the same direction by addition.

Combining local stiffness matrices into a global matrix

edit

The local matrices from element 1 and 2 will be combined as described above to give

 


Solution: Verify With CALFEM

edit

MatLab Code

edit

EDU>> Edof = [1 1 2 3 4

2 3 4 5 6];

EDU>> K = zeros(6);

EDU>> f = zeros(6,1);

EDU>> f(4) = -7;

EDU>> E1 = 3; E2 = 5;

EDU>> L1 = 4;L2 = 2;

EDU>> A1 = 1;A2 = 2;

EDU>> ep1 = [E1 A1]

ep1 =

    3     1


EDU>> ep2 = [E2 A2]

ep2 =

    5     2


EDU>> ex1 = [-3.464 0]; ex2 = [0 1.414];

EDU>> ey1 = [-2 0]; ey2 = [0 -1.414];

EDU>> Ke1 = bar2e(ex1,ey1,ep1)

Ke1 =

   0.5625    0.3248   -0.5625   -0.3248
   0.3248    0.1875   -0.3248   -0.1875
  -0.5625   -0.3248    0.5625    0.3248
  -0.3248   -0.1875    0.3248    0.1875


EDU>> Ke2 = bar2e(ex2,ey2,ep2)

Ke2 =

   2.5004   -2.5004   -2.5004    2.5004
  -2.5004    2.5004    2.5004   -2.5004
  -2.5004    2.5004    2.5004   -2.5004
   2.5004   -2.5004   -2.5004    2.5004


EDU>> K = assem(Edof(1,:),K,Ke1);

EDU>> K = assem(Edof(2,:),K,Ke2)

K =

   0.5625    0.3248   -0.5625   -0.3248         0         0
   0.3248    0.1875   -0.3248   -0.1875         0         0
  -0.5625   -0.3248    3.0629   -2.1756   -2.5004    2.5004
  -0.3248   -0.1875   -2.1756    2.6879    2.5004   -2.5004
        0         0   -2.5004    2.5004    2.5004   -2.5004
        0         0    2.5004    -2.5004  -2.5004    2.5004


EDU>> bc = [1 0;2 0;5 0;6 0];

EDU>> [a,r]=solveq(K,f,bc)

a =

        0
        0
  -4.3519
  -6.1268
        0
        0


r =

   4.4378
   2.5622
        0
  -0.0000
  -4.4378
   4.4378

Conclusion

edit

By comparing the value from the hand calculation and the value from the CALFEM program, we can see that the two matrices are identical. This verifies the assembly process of using addition when combining local matrices by addition.


Problem R2.2

edit

Disccription:

edit

We are asked to consider the L2-ODE-CC system referenced in (1) p.53-2 to have the following complex conjugate roots:

 

We are asked to find the homogeneous solution in standard form and to find the damping ratio,  

We are then to consider the initial conditions

 

We are to plot this solution and then use that plot to find the log decrement and compare it to the log decrement that would be obtained using (6) p.53-5b.

Finally, we're to find the quality factor, Q, and the loss factor,  



Solution:

edit

From R1.5, we have already obtained that for the above conditions,

 

with coefficients to the characteristic equation of

  and  

From (1) and (2) p.53-5, we have that

 

 

Knowing both a and b we get

 

 

We then have from equation (1) p.53-5d

 

 

Using the above graph of the homogeneous solution, accurate estimates were able to be made as to the various values of y at the peaks

Using equation (3) p.53-5c we are able to find

 
 
 

Conclusion

edit

We then arrive at

 

For comparison, the equation (6) p.53-5b gives

 

We now find the quality factor, Q. From (2) p.53-5d, we have

 

From (5) p.53-5d, we know that the loss factor is just the inverse of the quality factor so

 

Problem R2.3

edit

Problem Statement

edit

KS 2008 p.103 sec.2.7 pb.21 write a matlab program to assemble the global stiffness matrix, compute the unknown disp dofs, the reaction forces, the member forces; compare the results with exact hand calculation Statics, Mechanics of Materials).

run CALFEM to verify the results, but now do this problem in a different way (as in commercial FE codes).

first, establish 2 arrays: (1) the global node coordinate array “coord”, and (2) the elem connectivity array “conn”. next, write a matlab code to loop over the number of element to call the function “bar2e” of CALFEM to compute the element stiffness matrices. with this method, you let your matlab code construct the arrays ex1, ey1, ex2, ey2, etc. for you, using the info from the arrays “coord” and “conn”.


Solution

edit
MATLAB Code
edit

The following is a MATLAB code that assembles the global stiffness matrix, computes the unknown displacement dofs, reaction forces, and member forces

   function FEA_R2p3
   E = 10000; % Young's modulus
   A = pi*2^2/4; % Area
   % Lengths of each element
   L1 = 12;
   L2 = 12;
   L3 = 12;
   L4 = sqrt((12^2)+(9^2));
   L5 = 9;
   L6 = sqrt((12^2)+(9^2));
   L7 = 12;
   L8 = 9;
   L9 = sqrt((12^2)+(9^2));
   % topology matrix
   % [element# node1x node1y node2x node2y]
   Edof = [1 1  2  3  4;2  3  4 5 6;3  5  6 7 8;
           4 1  2  9 10;5  9 10 3 4;6  9 10 5 6;
           7 9 10 11 12;8 11 12 5 6;9 11 12 7 8];
   % element stiffness matrix in local coordinates for elements 1 through 9
   esmloc1 = (E*A/L1)*[1 0 -1 0;0 0 0 0;-1 0 1 0;0 0 0 0];
   esmloc2 = (E*A/L2)*[1 0 -1 0;0 0 0 0;-1 0 1 0;0 0 0 0];
   esmloc3 = (E*A/L3)*[1 0 -1 0;0 0 0 0;-1 0 1 0;0 0 0 0];
   esmloc4 = (E*A/L4)*[(L1/L4)^2 (L1/L4)*(L5/L4) -(L1/L4)^2 -(L1/L4)*(L5/L4);
                       (L1/L4)*(L5/L4) (L5/L4)^2 -(L1/L4)*(L5/L4) -(L5/L4)^2;
                       -(L1/L4)^2 -(L1/L4)*(L5/L4) (L1/L4)^2 (L1/L4)*(L5/L4);
                       -(L1/L4)*(L5/L4) -(L5/L4)^2 (L1/L4)*(L5/L4) (L5/L4)^2];
                       % [cos^2 cossin -cos^2 -cossin;cossin sin^2 -cossin -sin^2;
                       %  -cos^2 -cossin cos^2 cossin;-cossin -sin^2 cossin sin^2]
   esmloc5 = (E*A/L5)*[0 0 0 0;0 1 0 -1;0 0 0 0;0 -1 0 1];
   esmloc6 = (E*A/L6)*[(L1/L4)^2 -(L1/L4)*(L5/L4) -(L1/L4)^2 (L1/L4)*(L5/L4);
                       -(L1/L4)*(L5/L4) (L5/L4)^2 (L1/L4)*(L5/L4) -(L5/L4)^2;
                       -(L1/L4)^2 (L1/L4)*(L5/L4) (L1/L4)^2 -(L1/L4)*(L5/L4);
                        (L1/L4)*(L5/L4) -(L5/L4)^2 -(L1/L4)*(L5/L4) (L5/L4)^2];
                       % [cos^2 -cossin -cos^2 cossin;-cossin sin^2 cossin -sin^2;
                       %  -cos^2 cossin cos^2 -cossin;cossin -sin^2 -cossin sin^2]
   esmloc7 = (E*A/L7)*[1 0 -1 0;0 0 0 0;-1 0 1 0;0 0 0 0];
   esmloc8 = (E*A/L8)*[0 0 0 0;0 1 0 -1;0 0 0 0;0 -1 0 1];
   esmloc9 = (E*A/L9)*[(L1/L4)^2 -(L1/L4)*(L5/L4) -(L1/L4)^2 (L1/L4)*(L5/L4);
                       -(L1/L4)*(L5/L4) (L5/L4)^2 (L1/L4)*(L5/L4) -(L5/L4)^2;
                       -(L1/L4)^2 (L1/L4)*(L5/L4) (L1/L4)^2 -(L1/L4)*(L5/L4);
                        (L1/L4)*(L5/L4) -(L5/L4)^2 -(L1/L4)*(L5/L4) (L5/L4)^2];
                       % [cos^2 -cossin -cos^2 cossin;-cossin sin^2 cossin -sin^2;
                       %  -cos^2 cossin cos^2 -cossin;cossin -sin^2 -cossin sin^2]
   % element nodal coordinates
   % ex1 and ey1 are the x and y coordinates to the nodes of element 1
   ex1 = [0 12];  ey1 = [0 0];
   ex2 = [12 24]; ey2 = [0 0];
   ex3 = [24 36]; ey3 = [0 0];
   ex4 = [0 12];  ey4 = [0 9];
   ex5 = [12 12]; ey5 = [9 0];
   ex6 = [12 24]; ey6 = [9 0];
   ex7 = [12 24]; ey7 = [9 9];
   ex8 = [24 24]; ey8 = [9 0];
   ex9 = [24 36]; ey9 = [9 0];
   K = zeros(12);
   F = zeros(12,1); 
   F(6)=-1200; F(11)=400;
   ep = [E A];
   % assembly of the global stiffness matrix
   K =  assem(Edof(1,:),K,esmloc1);
   K =  assem(Edof(2,:),K,esmloc2);
   K =  assem(Edof(3,:),K,esmloc3);
   K =  assem(Edof(4,:),K,esmloc4);
   K =  assem(Edof(5,:),K,esmloc5);
   K =  assem(Edof(6,:),K,esmloc6);
   K =  assem(Edof(7,:),K,esmloc7);
   K =  assem(Edof(8,:),K,esmloc8);
   K =  assem(Edof(9,:),K,esmloc9);
   % boundary conditions
   bc = [1 0;2 0;8 0];
   % reaction force vector
   Q = solveq(K,F,bc);
   % finding element displacements
   ed1 = extract(Edof(1,:),Q); ed2 = extract(Edof(2,:),Q);
   ed3 = extract(Edof(3,:),Q); ed4 = extract(Edof(4,:),Q);
   ed5 = extract(Edof(5,:),Q); ed6 = extract(Edof(6,:),Q);
   ed7 = extract(Edof(7,:),Q); ed8 = extract(Edof(8,:),Q);
   ed9 = extract(Edof(9,:),Q);
   disp = [ed1;ed2;ed3;ed4;ed5;ed6;ed7;ed8;ed9];
   % reaction forces at each element; member forces (stress)
   N1 = bar2s(ex1,ey1,ep,ed1); MF1 = N1/A;
   N2 = bar2s(ex2,ey2,ep,ed2); MF2 = N2/A;
   N3 = bar2s(ex3,ey3,ep,ed3); MF3 = N3/A;
   N4 = bar2s(ex4,ey4,ep,ed4); MF4 = N4/A;
   N5 = bar2s(ex5,ey5,ep,ed5); MF5 = N5/A;
   N6 = bar2s(ex6,ey6,ep,ed6); MF6 = N6/A;
   N7 = bar2s(ex7,ey7,ep,ed7); MF7 = N7/A;
   N8 = bar2s(ex8,ey8,ep,ed8); MF8 = N8/A;
   N9 = bar2s(ex9,ey9,ep,ed9); MF9 = N9/A;
   RF = [N1;N2;N3;N4;N5;N6;N7;N8;N9];
   MF = [MF1;MF2;MF3;MF4;MF5;MF6;MF7;MF8;MF9];

With this, the output is as follows:

   GlobalStiffnessMatrix = K 
   dispDofs = disp
   reactionForces = RF
   memberForces = MF
GlobalStiffnessMatrix = 1.0e+003 *
   3.9584    1.0053   -2.6180         0         0         0         0         0   -1.3404   -1.0053         0         0
   1.0053    0.7540         0         0         0         0         0         0   -1.0053   -0.7540         0         0
  -2.6180         0    5.2360         0   -2.6180         0         0         0         0         0         0         0
        0         0         0    3.4907         0         0         0         0         0   -3.4907         0         0
        0         0   -2.6180         0    6.5764   -1.0053   -2.6180         0   -1.3404    1.0053         0         0
        0         0         0         0   -1.0053    4.2446         0         0    1.0053   -0.7540         0   -3.4907
        0         0         0         0   -2.6180         0    3.9584   -1.0053         0         0   -1.3404    1.0053
        0         0         0         0         0         0   -1.0053    0.7540         0         0    1.0053   -0.7540
  -1.3404   -1.0053         0         0   -1.3404    1.0053         0         0    5.2988         0   -2.6180         0
  -1.0053   -0.7540         0   -3.4907    1.0053   -0.7540         0         0         0    4.9986         0         0
        0         0         0         0         0         0   -1.3404    1.0053   -2.6180         0    3.9584   -1.0053
        0         0         0         0         0   -3.4907    1.0053   -0.7540         0         0   -1.0053    4.2446
dispDofs =
        0         0    0.3056   -1.4992
   0.3056   -1.4992    0.6112   -2.1836
   0.6112   -2.1836    1.0695         0
        0         0    0.8260   -1.4992
   0.8260   -1.4992    0.3056   -1.4992
   0.8260   -1.4992    0.6112   -2.1836
   0.8260   -1.4992    0.5204   -1.9258
   0.5204   -1.9258    0.6112   -2.1836
   0.5204   -1.9258    1.0695         0
reactionForces = 1.0e+003 *
   0.8000
   0.8000
   1.2000
  -0.5000
        0
   0.5000
  -0.8000
   0.9000
  -1.5000
memberForces =
 254.6479
 254.6479
 381.9719
-159.1549
        0
 159.1549
 254.6479
 286.4789
-477.4648

It can be seen here that the problem done by hand corresponds to the force output of the MATLAB program.


CALFEM MATLAB
edit
   % CALFEM
   % global coordinate matrix
   Coord = [0 0;12 0;24 0;36 0;12 9;24 9];
   % element connectivity array
   conn = Edof;
   % global nodal dof matrix
   Dof = [1 2;3 4;5 6;7 8;9 10;11 12];
   % contructing element arrays
   [ex,ey]=coordxtr(conn,Coord,Dof,2);
   K=zeros(12);
   EA=[E A]; % Young's modulus, Area
   % compute element stiffness matrix Ke and global stiffness matrix
   for i=1:9
      Ke = bar2e(ex(i,:),ey(i,:),EA); % element stiffness matrix
      K = assem(conn(i,:),K,Ke);      % global stiffness matrix
   end
   % element displacement vector
   ed=extract(conn,Q);
   % reaction forces
   for i=1:9
      N(i)=bar2s(ex(i,:),ey(i,:),ep,ed(i,:));
   end
   % member forces
   MF1C = N(1)/A;  % stress in AC
   MF2C = N(2)/A;  % stress in CE
   MF3C = N(3)/A;  % stress in EF
   MF4C = N(4)/A;  % stress in AB
   MF5C = 0;  % stress in BC
   MF6C = N(6)/A;  % stress in BE
   MF7C = N(7)/A;  % stress in BD
   MF8C = N(8)/A;  % stress in DE
   MF9C = N(9)/A;  % stress in DF
   MFC = [MF1C;MF2C;MF3C;MF4C;MF5C;MF6C;MF7C;MF8C;MF9C];

The output to the CALFEM code is as follows:

   CGlobalStiffnessMatrix = K 
   CdispDofs = ed
   CreactionForces = N'
   CmemberForces = MFC
CGlobalStiffnessMatrix = 1.0e+003 *
   3.9584    1.0053   -2.6180         0         0         0         0         0   -1.3404   -1.0053         0         0
   1.0053    0.7540         0         0         0         0         0         0   -1.0053   -0.7540         0         0
  -2.6180         0    5.2360         0   -2.6180         0         0         0         0         0         0         0
        0         0         0    3.4907         0         0         0         0         0   -3.4907         0         0
        0         0   -2.6180         0    6.5764   -1.0053   -2.6180         0   -1.3404    1.0053         0         0
        0         0         0         0   -1.0053    4.2446         0         0    1.0053   -0.7540         0   -3.4907
        0         0         0         0   -2.6180         0    3.9584   -1.0053         0         0   -1.3404    1.0053
        0         0         0         0         0         0   -1.0053    0.7540         0         0    1.0053   -0.7540
  -1.3404   -1.0053         0         0   -1.3404    1.0053         0         0    5.2988         0   -2.6180         0
  -1.0053   -0.7540         0   -3.4907    1.0053   -0.7540         0         0         0    4.9986         0         0
        0         0         0         0         0         0   -1.3404    1.0053   -2.6180         0    3.9584   -1.0053
        0         0         0         0         0   -3.4907    1.0053   -0.7540         0         0   -1.0053    4.2446
CdispDofs =
        0         0    0.3056   -1.4992
   0.3056   -1.4992    0.6112   -2.1836
   0.6112   -2.1836    1.0695         0
        0         0    0.8260   -1.4992
   0.8260   -1.4992    0.3056   -1.4992
   0.8260   -1.4992    0.6112   -2.1836
   0.8260   -1.4992    0.5204   -1.9258
   0.5204   -1.9258    0.6112   -2.1836
   0.5204   -1.9258    1.0695         0
CreactionForces = 1.0e+003 *
   0.8000
   0.8000
   1.2000
  -0.5000
        0
   0.5000
  -0.8000
   0.9000
  -1.5000
CmemberForces =
 254.6479
 254.6479
 381.9719
-159.1549
        0
 159.1549
 254.6479
 286.4789
-477.4648

The results of all methods verifies their accuracy.

Problem R2.4

edit

Description:

edit

We are to find an expression for the excitation referenced in (1) p.53-6 and then find the actual solution and the amplification factor, A for the following conditions:

 
 

We are then to find the complete solution.


Solution:

edit

 

Let us guess the solution is:

 

then

 
 
 

We then write

 

It follows that

  and that  

We solve for A and B to be

 
 

Once given the conditions

 
 

we can find

 

Plugging these values in yields

 
 

The particular solution is then

 

Equations (3) p. 53-8 and (2) p. 53-8 give us

 
 

The following is the graph of the homogeneous solution:

 

The following is the graph of the particular solution:

 

The following is the graph of the total solution,  :

 

A graph of all three plots is produced below:

 

Problem R2.5 CALFEM 3.4 Manual exs1 on a Spring System and Verification

edit

Given a Linearly Elastic Spring System

edit

Consider a system of 3 linearly elastic springs

 
Figure 1: Spring System from CALFEM Manual 3.4 p9.3-2 ([2])

Use the following diagram for the node locations

 
Figure 1: Spring System node numbering from CALFEM Manual 3.4 p9.3-2 ([3])

Find:

edit

1.Find the results when running exs1 from CALFEM

edit

2.Verify the results of CALFEM to manual calculations

edit

Solution: Results of CALFEM

edit

Construct the topology matrix

edit

This matrix contains the node numbers and degrees of freedom.

 

Construct the global stiffness matrix and load vector f

edit

 

The load vector is in position 2 with F=100

 

Generating the element stiffness matrices

edit

The element stiffness matrices are generated using the CALFEM function spring1e. The matrices are generated using k and 2k where k=1500

 

 


Assembling the element stiffness matrices into the global stiffness matrix

edit

The assembly is done using the assem fucntion from CALFEM.

Assembling the second element at the first row:

 

Assembling the first element at the second row:

 

Assembling the second element at the third row:

 


Solve the system of equations using boundary conditions

edit

The CALFEM fucntion solveq is used to solve the system of equations subject to certain boundary conditions.

 

 

 


Evaluate the element forces from the element displacements

edit

 

 

 

Evaluate the spring forces from function spring1s

edit

 

 

 

Solution: Results of verification

edit

This section verifies the CALFEM code by manual calculations

Nodal equilibrium equations

edit

 

 

 

Setting up the equation to solve

edit

 


 

Implementing the boundary conditions

edit

 

Inverting the global stiffness matrix to find displacements

edit

 

Displacements 1 and 3 are zero due to boundary conditions.

Calculate forces in springs from equation

edit

The forces in the springs are found by the following equations. Negative forces display compressions and positive tensions.

 

 

 

 

Discuss if the solution is verified

edit

The solution by CALFEM is verified since both the displacements and the forces are the same by both methods.

Problem R2.6

edit

Problem Statement

edit

Find the displacements of the nodes as well as the forces in the springs(tension/compression). Also determine the reaction forces at the walls.  


Nodal Equilibrium Equations

edit
  
   
  
  
  
 

 
 


Using Matlab
Solutions
Displacements  

These are the displacements at  ,  , and   respectively. The negative sign shows that the nodes are moving in the direction of the applied force which is against the convention of positive displacement left to right.

Now it is possible to find the forces in the spring. Negative forces denote compression and positive forces denote tension.

Forces in Springs

edit
 
 
 
 
 
 
 

Now,

Reaction Forces

edit
 
 
 
 

Verification with Calfem

edit

a values are the nodal displacements 1 through 5.
r values are the reaction forces at the walls.
es1-es6 are the forces in the springs. Negative values denote compression and positive values denote tension. These answers match those previously calculated.


===Code===
Edof=[1 1 2;2 2 4;3 2 3;4 1 3;5 3 4;6 4 5];
ep1=500;
ep2=400;
ep3=600;
ep4=200;
ep5=400;
ep6=300;
Ke1=spring1e(ep1);
Ke2=spring1e(ep2);
Ke3=spring1e(ep3);
Ke4=spring1e(ep4);
Ke5=spring1e(ep5);
Ke6=spring1e(ep6);
K=zeros(5,5);
K=assem(Edof(1,:),K,Ke1);
K=assem(Edof(2,:),K,Ke2);
K=assem(Edof(3,:),K,Ke3);
K=assem(Edof(4,:),K,Ke4);
K=assem(Edof(5,:),K,Ke5);
K=assem(Edof(6,:),K,Ke6);
bc= [1 0; 5 0];
f=zeros(5,1);
f(3)=-1000;
[a,r]=solveq(K,f,bc)
ed1=extract(Edof(1,:),a);
ed2=extract(Edof(2,:),a);
ed3=extract(Edof(3,:),a);
ed4=extract(Edof(4,:),a);
ed5=extract(Edof(5,:),a);
ed6=extract(Edof(6,:),a);
es1=spring1s(ep1,ed1)
es2=spring1s(ep2,ed2)
es3=spring1s(ep3,ed3)
es4=spring1s(ep4,ed4)
es5=spring1s(ep5,ed5)
es6=spring1s(ep6,ed6)

'''Solution'''
a =

         0
   -0.8542
   -1.5521
   -0.8750
         0


r =

  737.5000
   -0.0000
    0.0000
    0.0000
  262.5000


es1 =

 -427.0833


es2 =

   -8.3333


es3 =

 -418.7500


es4 =

 -310.4167


es5 =

  270.8333


es6 =

  262.5000




--EML4507.s13.team4ever.Wulterkens (discusscontribs) 07:21, 6 February 2013 (UTC)