Dragging a Rod

A uniform rod of unit length is on a flat, horizontal surface, supported at its two ends. The friction force is the same all over the surface. One end of the rod is pulled by a horizontal force, slowly and steadily, in a direction perpendicular to the original direction of the rod, along the black line in the illustration.

What will be the distance between the freely moving end of the rod and the black line when the pulled end of the rod is a unit distance from the starting point?

See also similar problems by Steven Chase and by Digvijay Singh .


The answer is 0.648.

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

Laszlo Mihaly
May 23, 2018

The pulled end of the rod will move with constant velocity, and its coordinates will be x 1 = v 0 t x_1=v_0t and y 1 = 0 y_1=0 , where \ell is the length of the rod. For the rest of the discussion we will assume the v 0 0 v_0 \rightarrow 0 limit i.e. the motion is "slow and steady".

The free end has coordinates of x 2 = cos α + x 1 x_2=\ell \cos \alpha+x_1 and y 2 = sin α y_2=\ell \sin \alpha , where α \alpha is the angle between the rod and the black line. Taking the time derivative and using ω = α ˙ \omega =\dot \alpha , v x = x ˙ 2 v_x=\dot x_2 , v y = y ˙ 2 v_y=\dot y_2 we get

v x = ω sin α + v 0 v_x=-\omega \ell \sin \alpha +v_0 , v y = ω cos α v_y=\omega \ell \cos \alpha

The velocity of the free end makes an angle β \beta to the reference line so that sin β = v y / v \sin \beta= v_y/v and cos β = v x / v \cos \beta= v_x/v , where v = v x 2 + v y 2 v=\sqrt{v_x^2+v_y^2} The magnitude of the friction force is constant, F F . The direction of the friction force is opposite to the velocity:

F x = F cos β F_x=-F\cos\beta , F y = F sin β F_y=-F\sin\beta

For the equation of motion we need to look at the torque acting on the rod and make it equal to zero. Using the pulled end as the axis of rotation, we get

F x sin α = F y cos α F_x\ell \sin \alpha= F_y \ell \cos \alpha

Inserting the friction force and the velocity expressed before, we get

( ω v 0 sin α + 1 ) sin α = ω v 0 cos 2 α \left(-\frac{\omega \ell}{v_0}\sin \alpha +1\right)\sin \alpha= \frac{\omega \ell}{v_0} \cos^2\alpha

or

sin α = ω v 0 = α ˙ v 0 \sin \alpha=\frac{\omega \ell}{v_0}=\frac{\dot \alpha \ell}{v_0}

This is a first order differential equation, that can be solved by separation of variables:

v 0 d t = d α sin α \frac{v_0}{\ell}dt=\frac{d\alpha}{\sin \alpha}

The integration yields

v 0 t = log ( tan α / 2 ) + c o n s t \frac{v_0}{\ell}t= \log (\tan \alpha/2) +const

The time is related to the position of the pulled end by t = x / v 0 t=x/v_0 , therefore

x = log ( tan α / 2 ) + c o n s t \frac{x}{\ell}= \log (\tan \alpha/2) +const

It turns out that if const =0 then at x = 0 x=0 we have α = 9 0 \alpha=90^{\circ} . When x = x=\ell , we get tan α / 2 = e \tan \alpha/2 = e , and we can solve that for α \alpha . Recalling that y 2 = sin α y_2=\ell \sin \alpha we get y 2 = 0.648 y_2=0.648

Ivo Zerkov (see below) pointed out that the response can be written in the form of 2 e 1 + e 2 \frac{2e}{1+e^2} .

The red curve below shows the position of the top of the rod as the bottom is pulled. A few instances of the positions of the full rod are also shown.

Notes:

  1. The motion is reversible. If we start with a rod that is nearly parallel to the reference line, and pull it backwards, the free end will follow the same curve.

  2. One of the consequences of the torque equation is that α = β \alpha=\beta . The motion of the end point of the rod is parallel to the rod, just as in this problem

  3. It is not necessary to require that the friction is the same everywhere. As long as the friction is isotropic (independent of the direction of motion) the result is the same.

Fun problem. Reminds me of one I posted a while back. But then again, I think that one may have been inspired by some previous ones of yours.

https://brilliant.org/problems/sliding-rod-with-friction-discontinuity/?ref_id=1476951

Steven Chase - 3 years ago

Log in to reply

I will insert a reference to it.

Laszlo Mihaly - 3 years ago

Correct answer according to your solution should be sin ( 2 tan 1 1 e ) = 2 e 1 + e 2 0.648 \sin{(2\tan^{-1}{\frac{1}{e}})}=\frac{2e}{1+e^2}\approx0.648 , not 0.645 0.645 .

Ivo Zerkov - 3 years ago

Log in to reply

You are right, I will correct it.

Laszlo Mihaly - 3 years ago
Steven Chase
May 23, 2018

Here's a numerical simulation. Interestingly, I consistently get about 0.648, instead of 0.645. I'm using very high resolution for the simulation. But close enough, I suppose.

import math
import random

#####################################

m = 1.0  # mass and length
L = 1.0

I = (1.0/3.0) * m * (L**2.0)  # moment of inertia

g = 10.0 # gravity

mu = 0.1 + 0.5 * random.random() # randomly generated friction coeff

W = m*g  # weight

t = 0.0  # time and time resolution
dt = 10.0**(-4.0)

v = 0.001  # very low velocity for dragging point
tf = 1.0 / v  # final simulation time

#####################################

theta = math.pi / 2.0   # theta initially 90 deg
thetad = 0.0   # first and second derivatives
thetadd = 0.0

#####################################


while t <= tf:

    theta = theta + thetad * dt      # numerical integration
    thetad = thetad + thetadd * dt

    x = v*t + L * math.cos(theta)    # free end coordinates and velocity
    y = L * math.sin(theta)

    xd = v - L * math.sin(theta) * thetad
    yd = L * math.cos(theta) * thetad

    ########

    vend = math.hypot(xd,yd)  # free end speed

    ux = xd / vend   # unit vectors corresponding to free end velocity
    uy = yd / vend

    ########

    Fx = -mu * (W/2.0) * ux    # friction forces on end
    Fy = -mu * (W/2.0) * uy

    ########

    rx = x - v*t   # lever components for torque calc
    ry = y - 0.0

    ########

    T = rx*Fy - ry*Fx   # torque

    thetadd = T / I   # second deriv of theta

    ########

    t = t + dt  # increment time

#####################################

# print results

print t
print theta
print x
print y

Indeed, it should be 0.648. I will correct it.

Laszlo Mihaly - 3 years ago

I think for x2 should be x1-lcosalfa, and that will change almost everything in final solution

Andreea Goia - 2 years, 2 months ago

Log in to reply

That depends on how is the angle alpha defined.

Laszlo Mihaly - 2 years, 2 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...