Documente Academic
Documente Profesional
Documente Cultură
5 (54) - 2012
Series 1: Special Issue No. 1
(1)
where:
- [M ] inertial matrix of dynamic system;
- [C ] damping matrix of dynamic system;
- [K ] stiffness matrix of dynamic system;
46
x k (t ) x k ,i (t )
(2)
i 1
i 1
i 1
x k (t ) x k ,i (t ) k ,i i (t )
(3)
1, 2
x1 1,1
x
2, 2
2 2,1
.
.
x
n 1, 2
n 1
n 1,1
n,2
xn n ,1
1, n 1
2, n 1
.
.
. n 1, n 1
n , n 1
1, n 1
2, n 2
. .
n 1, n n 1
n , n n
{x(t )} [ ]{ (t )};
.
..
..
{x(t )} [ ]{ (t )};
(4)
{x(t )} [ ]{ (t )}.
where:
- [ ] modal matrix
- { (t )} displacements vector in natural
coordinates
.
- { (t )} velocity
Fig. 2. Vibrations mode and eigenvalues
for a system with n degrees of freedom
dynamic
coordinates
vector
in
natural
M. BOTI: Matlab program for the numerical solution of Duhamel convolution integral
..
- Fi* i {F (t )} corresponding
T
{x(t )} [ ]{ (t )};
-i
..
..
{x(t )} [ ]{ (t )};
(7)
{x(t )} [ ]{ (t )}.
The analysis of the dynamics of a system
with n degrees of freedom dynamics can
be seen as solving the system of equations
is reduced to calculation convolution
integral Duhamel.
For this reason this paper presents a
Matlab program for numerical integration
of convolution integral Duhamel.
If the t apply an H impulse of the
system, this is removed from equilibrium
.
47
x(t )
H i (t )
e
sin * (t )
m *
(8)
H * (t )
e
sin i* (t ) for t
*
m
i2
corresponding damping
2 i
mode i.
- [C ] [ M ] [ K ] Rayleigh damping
matrix
To solve decoupled equations of motion,
Duhamel convolution integral can be used.
Displacement response obtained using
The convolution integral Duhamel integral
on each dynamic degrees of freedom is:
t
1
ii (t ) * * F * ( )e ii (t ) sin i* (t )d
M i i 0
i 1...n
(6)
where,
- i* i 1 2 damped pulsation
48
0 at t .
After applying elementary impulse
displacement response of
the dynamic system dynamic on degree of
freedom is:
dx(t )
dH (t )
sin * (t )d
e
m *
(9)
F ( ) * (t )
e
sin i* (t )d
m *
Dynamic displacement response of the
system is achieved by superposition of
effects as a integral amount of motion
generated
by
elementary
pulse
succession
on the length
of time that force acts [1]:
F ( ) * ( t )
x(t )
e
sin i* (t )d
0 m *
f ( ) * ( t )
F0
e
sin i* (t )d
0 m *
f ( ) ( t )
F0
e
sin i* (t )d
0 m *
where,
- x 0 initial displacement for t=0
t )
x1
x1
h
h
[ f (a) f ( x1 )] [ f (b) f ( x1 )]
2
2
h
[ f (a) 2 f ( x1 ) f (b)]
2
ba
[ f (a) 2 f ( x1 ) f (b)]
4
To reduce the error of integration, the
interval [a,b] may be divided in n
subintervals with division h (b a) / n ,
M. BOTI: Matlab program for the numerical solution of Duhamel convolution integral
x1
x n 11
can
be
the
f ( x)dx
h
h
[ f (a) f ( x1 )] ... [ f (b) f ( x n 1 )]
2
2
h
[ f (a) 2 f ( x1 ) ..... 2 f ( x n 1 ) f (b)]
2
ba
e cos d e t sin t
0 m
F ( )
e sin d e t cot
0 m
t
t
A(t )e sin t B(t )e cos t
where,
F ( )
e cos d
A(t )
0 m
F ( )
B(t )
e sin d
0 m
To solve the Duhamel integral trapezoids
method will be used to calculate integrals
A(t) and B(t) with Matlab program [2].
49
50
f(i)=f1;
timp(i)=t;
% Variation representation force
%function in time
p=plot(t,f1,'--rs','LineWidth',2,...
'MarkerEdgeColor','b',...
'MarkerFaceColor','b',...
'MarkerSize',2);
hold on;
grid on;
% Calculation of values ya and yb
%from Duhamel convolution
integral
ya(i)=f1*1000*exp(niu*omega*t)*...
cos(omegad*t)*(1000*m*omegad)^-1;
yb(i)=f1*1000*exp(niu*omega*t)*...
sin(omegad*t)*(1000*m*omegad)^-1;
end
% Final values
yaa=[ya0
ya(1,:)' ];
ybb=[yb0
yb(1,:)'];
% Method trapezoids
t=0
subplot(2,1,2)
for i=2:npas+1
t=t+h;
A(1)=0;
B(1)=0;
A(i)=A(i-1)+h*(yaa(i)+yaa(i-1))/2;
B(i)=B(i-1)+h*(ybb(i)+ybb(i-1))/2;
% Response in motion for a ramp signal in
%case 1GLD dynamic system
d(i)=(A(i)*exp(-niu*omega*t)*...
sin(omegad*t)-...
B(i)*exp(-niu*omega*t)*...
cos(omegad*t))*(k/50);
% Graphical representation of the ratio of
%dynamic
displacement
and
static
%displacement
plot(t,d(i),'--rs','LineWidth',2,...
'MarkerEdgeColor','r',...
'MarkerFaceColor','r',...
'MarkerSize',2);
hold on;
grid on;
end