Example 2 – free response and logarithmic decrement x1 x ( t1 )
"2 = =
x 2 x (t 2 )
We have seen that the free response of an under-damped linear single-DOF system is
given by Xe#$% n t1 sin(% d t1 + & )
=
Xe#$% n ( t1 +Td ) sin(% d ( t1 + Td ) + & )
x ( t ) = c cos " d t + s sin " d t = Xe#$" n t sin(" d t + % )
sin(% d t1 + & ) 1 1#$ 2
= = = e$% nTd = e 2 '$ /
which represents a harmonic oscillation with a time-decaying amplitude. e#$% nTd sin(% d t1 + 2' + & ) e#$% nTd
! What is the relationship between the damping ratio ! and the rate at which the Alternately, between the first and nth maximums we have:
oscillations die away?
! x1 x ( t1 )
"n = =
x n x (t n )
Xe#$% n t1 sin(% d t1 + & )
=
Xe#$% n ( t1 +nTd ) sin(% d ( t1 + nTd ) + & )
sin(% d t1 + & ) 1 1#$ 2
= = = e n$% nTd = e 2n'$ /
e#$% n nTd sin(% d t1 + 2n' + & ) e#$% n nTd
t3
t1 t2 or
! 1 $
ln(#n ) = (1)
2n" 1% $ 2
Solution This equation shows that the damping ratio ! is directly related to the ratio of successive
The local maximum values of response for the system occurs when: peaks in the decaying response. For a given “logarithmic decrement” ln("n ) , we can use
!
the above equation to solve for the damping ratio !. Generally we are interested in
dx ( t )
= 0 = "#$ n Xe"#$ n t sin($ d t max + % ) + $ d Xe"#$ n t cos($ d t max + % ) & problems with light damping, " << 1 . For this, 1" # 2 $ 1 and we see that:
dt
sin($ d t max + % ) $ !
= tan($ d t max + % ) = d & 1
cos($ d t max + % ) #$ n "= ln($n ) (2)
2n# ! !
% 1 ' $ * 2n-
t max =" + atan) d , + Discussion
$d $d ( #$ n + $ d a) Note that equation (1) (or equation(2)) does NOT depend on the natural frequency
! of the system or on the initial conditions.
From this, we see that these times are equally spaced, with the time spacing between the b) Equation (1) is often used in experimental vibration work to obtain an estimate of
peaks given by the damped natural period Td = 2" /# d . the damping ratio from a free response test. It is simple to use. One only needs to
!
measure to take the ratio of response peaks after n cycles. Taking the natural log
Let’s define the ratio of successive peaks x1 and x 2 in the above figure as "2 : with a factor of 1/2n" gives an estimate of the damping ratio.
! c) These results are also important in giving us a feel for what reasonable values of
damping ratio to expect in response. A plot of equation (1) is shown below. For
example, from this plot we see that successive peaks are reduced to 80% on each
! ! !
cycle for a damping ratio as small as 3% of critical ( " = 0.03); after 5 cycles, this
is a reduction to about 1/3 of its original value ( 0.8 5 = 0.33). For a 10% damping
!
!
ratio, " = 0.1, the peak response is reduced to 52% after each cycle; after 5 cycles,
the peak is reduced to about 4% of its original value ( 0.8 5 = 0.038 ). The point? It
does not take much damping (in terms of damping ratio) to quickly damp out the
free response. It is easy to remember that " = 1 is a critical value defining the
! damping threshold above which no oscillations occur. However, we should
!
remember that for damping ratios much smaller than unity, the response dies
away very quickly. There is an animation on the course website that shows this
effect. Check it out! !