I am currently looking at the theory of find the viscosity of and object through damped harmonic motion, and tho it can be done there is obviously a limitation with regrades to the medium. If the medium is too viscous then critical dampening will occur and we lose the ability to find the viscosity.

I was looking at the equation for motion of a driven system and was wondering if you could use a driven oscillator to work out the viscosity of a more viscous medium.

  • $\begingroup$ No I do mean critically damped. So with a under damped system you get an exponential decay, and I was just wondering if you could take a critically damped system and turn into a under damped system $\endgroup$ – james2018 Jan 23 at 23:11
  • $\begingroup$ Yes sorry, I have now rephrased the question more fully. $\endgroup$ – james2018 Jan 23 at 23:25

The solution to $m \ddot{x} + k x + c \dot{x} = 0$ can be characterized with the following parameter substitutions

  • $ k = m \omega_n^2 $
  • $ c =2 \zeta m \omega_n $

Where $\omega_n$ is the natural frequency of the system, and $\zeta$ the damping ratio. When $\zeta=1$ the system is critically damped when over 1 it is overdamped and under 1 it is underdamped.

The equation of motion is now

$$ \ddot{x} + \omega_n^2 x + 2 \zeta \omega_n \dot{x} = 0 $$

The solution is

$$ x(t) = \begin{cases} \exp(-\omega_n \zeta t) \left( C_1 \sin (\omega t) + C_2 \cos (\omega t) \right) & \zeta \leq 1, \;\omega = \omega_n \sqrt{1+\zeta^2} \\ \exp(-\omega_n \zeta t) \left( C_1 \sinh (\omega t) + C_2 \cosh (\omega t) \right) & \zeta > 1, \;\omega = \omega_n \sqrt{\zeta^2-1} \end{cases} $$

So if you find the exponential decay rate $\beta = \omega_n \zeta$ experimentally, you can then solve for $\zeta$ and then for $c$ if you know the mass $m$, and the stiffness $k$.


I assume it is a mechanical system that you are using and possibly a spring-mass system?

Having made the assumption that the resistive force is proportional to the speed of the mass, the equation of motion for a system undergoing free oscillations is $m\ddot x + b\dot x + kx =0$.

If $b^2 < 4km$ the system is under damped and so will perform oscillatory motion.
If at present your system is over damped you might consider increasing the mass $m$ and $k$ which often is the spring constant to move the system into the under damped regime.

The exponential envelope will be of the form $\exp\left( -\frac {b}{2m}t\right)$

If the system is to be driven then the equation of motion changes to something like $m\ddot x + c \dot x + kx =F_0\cos (\omega t)$.
Having a lot of damping is good news in that the driven system will reach a steady state (constant amplitude) in a relatively short period of time.

The equation for the amplitude of the driven system $A$ as a function of driver frequency $\omega$ is quite complicated $A(\omega) = \dfrac{F_0}{\sqrt{m^2(\omega^2-\omega^2_0)+b^2\omega^2_0}}$ where $\omega_0$ is the natural frequency of the system when it is undamped.

At resonance (maximum amplitude of the driven system) $A = \dfrac {F_0}{b\omega_0}$ however it does need you to know the amplitude of the force $F_0$ produced by the driver if you want to find $b$.

Now you can get away with not knowing what $F_0$ is by taking a number of readings of amplitude of the driven system as a function of driver frequency but to eliminate $F_0$ you will have to make sure that $F_0$ does not depend on the frequency of the driver.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.