Looking For A Real Spring

Suppose we have a spring of length L L and spring constant k k . Unlike your typical massless spring, ours is real and has uniform mass density given by σ = M / L \sigma = M/L .

We fix one end of the spring to a table, compress it, and release it, so that it starts vibrating. Find its frequency of oscillation in rad / s \si{\radian/\second} .

Given your answer to 4 significant figures.

Details and Assumptions :

  • k = 1 N / m k = 1\ \si{\newton/\meter} .

  • M = 1 kg M = 1\ \si{\kilo\gram} .


The answer is 1.732050808.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

2 solutions

Josh Silverman Staff
Jul 7, 2016

This problem departs from the oscillating spring in that the material of the spring has mass. Instead of propelling a mass at the end of the spring through a large motion, the driving force from the compressed/extended spring is used to accelerate and decelerate mass that is—on average—closer to the fixed end, and thus, requires less energy. Therefore, we should expect the massive spring to vibrate with a frequency faster than the massless spring with a massive weight at the end.

It would be cumbersome to approach this problem using Newton's second law as we need to consider the distribution of mass along the length of the spring—not just the bob at the end like usual. Therefore, we take the Lagrangian approach which makes it easier to deal with distributed quantities like the many particles that make up the spring.

If the particles of the spring vibrated with a uniform velocity, then we could simply write the kinetic energy of the spring as 1 2 m v 2 \frac12 mv^2 . However, when the spring vibrates, the particles near the fixed end oscillate with a very small amplitude while the particles near the free end have a bigger amplitude. In particular, if the free end vibrates with amplitude A A , a particle halfway along the spring will have an amplitude, and therefore a speed half as large. Thus, if the end vibrates with speed s ˙ \dot{s} , a particle a distance l l along the spring will have the speed u ( l ) = l L s ˙ u(l)=\frac{l}{L}\dot{s} , and kinetic energy δ T = 1 2 u ( l ) 2 σ d l \delta T = \frac12 u(l)^2 \sigma dl .

As the mass of the spring is evenly distributed, we can write a mass density σ = M / L \sigma = M/L so that the mass of a small segment of the spring of length d l dl is given by σ d l \sigma dl .

To get the kinetic energy T T of the entire spring, we can simply add up the kinetic energy of the small mass elements, T = δ T = 1 2 u ( l ) 2 σ d l T = \int \delta T = \int \frac12 u(l)^2 \sigma dl . We find

T = σ 2 0 L u ( l ) 2 d l = σ 2 L 2 s ˙ 2 0 L l 2 d l = M s ˙ 2 6 \begin{aligned} T &= \frac{\sigma}{2} \int\limits_0^L u(l)^2 dl \\ &= \frac{\sigma}{2L^2}\dot{s}^2\int\limits_0^L l^2 dl \\ &= \frac{M\dot{s}^2}{6} \end{aligned}

The potential energy of the system is given by the Hooke potential U = 1 2 k s 2 U = \frac12 ks^2 so that the Lagrangian is

L = T U = M s ˙ 2 6 1 2 k s 2 \begin{aligned} \mathcal{L} &= T - U \\ &= \frac{M\dot{s}^2}{6} - \frac12 ks^2 \end{aligned}

Taking the derivatives—according to the Euler-Lagrange equations—we find

d d t d L d s ˙ = d L d s M 3 s ¨ = k s s ¨ = 3 k M ω 2 s \begin{aligned} \frac{d}{dt}\frac{d\mathcal{L}}{d\dot{s}} &= \frac{d \mathcal{L}}{ds} \\ -\frac{M}{3} \ddot{s} &= ks \\ \ddot{s} &= -\overbrace{\frac{3k}{M}}^{\omega^2} s \end{aligned}

Which is the expression for a harmonic oscillator with frequency ω = 3 k M \omega = \sqrt{\frac{3k}{M}} .

Got close with a numerical solution in Excel....slightly less than 2.

Matt Owings - 4 years, 10 months ago

Log in to reply

I would be very interested to see your approach. What number do you actually get?

Josh Silverman Staff - 4 years, 10 months ago

Log in to reply

As the spring has mass , shouldn't we consider its gravitational potential energy term also?

Ankit Kumar Jain - 3 years, 3 months ago

Log in to reply

I divided the spring into 10 spring-mass models in series. Each of the 10 subdivisions were assigned a massless spring with a k of 10 N/M and attached to a point mass of 0.1 kg. each point mass therefore had force exerted on it by 2 springs....except for the point mass at the free end of the spring, it received force from only the 10th spring. I set initial conditions X1-X9=0 and X10=0.05. V1-V10=0 and time step =0.005. A graph showing 12 seconds of displacement of the end of the spring vs time appeared to show 2 frequency components, a low freq component at around 2 rad/sec and a high freq component at around 8 rad sec. Since the values for k and m given were simple integer values of 1, I eyeballed the graph and concluded the answer is most likely 2 rad/sec. Since that was incorrect, I examined my work and concluded I had made a mistake assigning initial conditions. I gave X10 a value of 0.05, X9=0.045.......X1=0.005. This yielded a smooth cosine wave with a frequency of 1 rad/sec. That being incorrect, I reexamined my results from my first try, setting X1-X9=0 and carefully counting time between maximums in the spreadsheet corresponding to major crests in the graph.....The frequency was 1.743 rad/sec, very close to the frequency you arrived at analytically. I should have been more careful! But now I am confused....it seems to me that the second way I assigned initial conditions to my model should have been the correct way. What am I missing?

Matt Owings - 4 years, 10 months ago

After rethinking initial conditions again I set all Intermediate displacements X1-X10 at 0.005. I think this would be right because if we exert a tensile force on the end of the spring in a static state that force will be equally transmitted through all series segments and thus each will undergo equal stretch. This approach still yields a frequency of 1 rad/sec....I am still trying to figure out what is wrong with my model!

Matt Owings - 4 years, 10 months ago

Log in to reply

Sorry I haven't responded yet. Your approach is very cool/ingenious. Once I free up some time, I'm going to come back to it.

Josh Silverman Staff - 4 years, 10 months ago

First write the energy of the spring during motion i.e. after the spring is released.

The spring will possess elastic potential energy and kinetic energy.

We know that when one end of a spring with mass is fixed its kinetic energy= 1 6 \frac{1}{6} M v 2 v^{2}

Its elastic potential energy = 1 2 \frac{1}{2} k x 2 x^{2}

Total Energy ,

1 6 \frac{1}{6} M v 2 v^{2} + 1 2 \frac{1}{2} k x 2 x^{2} = E

On differentiating w.r.t time on both sides,

k x v kxv + M v a 3 \frac{Mva}{3} = 0 0

a a = - 3 k x M \frac{3kx}{M}

Comparing to the equation, a a = - ω 2 ω^{2} x x

ω ω = 3 k M \sqrt{\frac{3k}{M}}

Substituting the given values,

ω ω = 3 \sqrt{3}

= 1.732 1.732

Why didn't you consider the gravitational potential energy term? I am asking so because if we include the potential energy term , it is not an SHM in that case as the equation is of the form a = -kx +c. Where am I wrong?

Ankit Kumar Jain - 3 years, 3 months ago

Log in to reply

F=-kx+c represents an SHM with just a shifted mean position

Satwik Murarka - 3 years, 3 months ago

Log in to reply

Thanks for the explanation , I missed that point

Ankit Kumar Jain - 3 years, 3 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...