ME 6303 – MECHANICAL VIBRATIONS

Lab 2: Numerical Integration and Video Analysis - Simulation of Coulomb

Friction

Prepared by:

Abraham Tobin- 201866878

Mariam Bassam-201559176

Report date: 29th September 2023

TABLE OF CONTENTS

INTRODUCTION 1

PROCEDURE 1

CALCULATIONS 4

RESULTS & DISCUSSION 5

CONCLUSION 8

REFERENCES 9

APPENDIX 10
INTRODUCTION
The objective of this experiment is to study the effect of coulomb friction on a mass spring
system. A video of a mass moving down an inclined plane was analyzed using vernier video
analysis software. The displacements and velocities obtained from the video were used to
calculate the friction coefficients. These friction coefficients are used in MATLAB to plot a
simulated graph of displacement vs time. The simulation results were compared to the
experimental results from lab 1 to observe any similarities.

PROCEDURE
The tilt-table test is performed to obtain the angle between the inclined wood surface and the
horizontal axis.
The mass is then allowed to move down the incline and the values of position and time are
obtained. A graph of displacement versus time is plotted using the data obtained from analyzing
the video in Vernier Software. The equation of the curve is obtained by plotting a quadratic best
fit curve (Figure 1). The acceleration is obtained by multiplying the coefficient of x2 by 2, which
gives an acceleration of -0.378 m/s2.

Figure 1: Mass Displacement versus Time

1
The geometry of the inclined surface was measured in order to calculate the kinetic and static
friction factor by using Vernier Analysis right before mass started sliding as shown in Figure 2.

Figure 2: Geometry of Inclined Surface

The resulting data are presented in Table 1, in which the angle of the tilted table is at 0.28 rad.
Table 1: Venier Analysis of Tilted Table

The free body diagram of the inclined mass is drawn and the forces are summed up to obtain a
force equation and calculate static and kinetic friction coefficients which is demonstrated under
calculation.

Figure 3: Inclined Mass FBD

2
The continuous function (Equation 1) is used to simulate Coulomb friction to handle transition
from static to dynamic friction. The range of velocity used in calculation is from -5m/s to 5m/s.

(1)

The friction was plotted against velocity range and the parameters were modified by trial and
error when adjusting a, b, g, n and μo to obtain peak and steady state friction factor values similar
to values in the calculation. Equation 1 was incorporated in MATLAB to simulate the transition
from static to kinetic friction.

The acceleration equation in MATLAB was modified to include non-linear friction as shown in
Equation 2; however, sgn(v) was not implemented in the equation as its effect has been
accounted for in tanh(v) from Equation 1.
𝑎 = − (μ 𝑁 𝑠𝑔𝑛(𝑣) − 𝑘 𝑥)/𝑚) (2)

The normal force N acting on hoisted mass was calculated using the following equation:
𝑁 = 𝑚𝑔𝑐𝑜𝑠θ (3)

The mass is hoisted by cable and placed on an inclined surface with angle of 75o, which is used
to obtain the normal force, N.

3
CALCULATIONS
The force equation is used to find the values of the static friction coefficient (μs) and dynamic
friction coefficient (μk). The value of acceleration is set to zero when finding μs. The calculations
are demonstrated below.

4
RESULTS & DISCUSSION

The friction coefficient curve in Figure 4 was obtained after modifying parameters in equation
(2) to obtain the static friction coefficient (μs) at the peak and kinetic friction coefficient (μk) for
velocities greater than zero. The parameters were modified so the coefficients obtained from the
curve are close to the coefficient values obtained from the calculations. The modified parameters
are: a= 7, b=100, g=2, n=2, and μo=0.247.

Figure 4: Friction Coefficient versus Velocity

The code used in Lab 1 was modified to include position and velocity of the mass versus time as
well the friction force versus velocity. Figure 5 represents the velocity and displacement of mass
as time varies. The red line represents the data obtained from running Vernier Video Analysis as
the mass sliding on an inclined surface while the blue line represents the MATLAB. The data
obtained from Vernier Video Analysis were uploaded to an excel sheet. It is observed that the

5
amplitudes of plots are similar. Both of the curves dampens after 2 seconds.However, the graph
from excel dampens slightly faster than the one obtained from MATLAB which can be
distinguished by the height of the third peak.

Figure 5: Mass Displacement and Velocity versus Time

The mcksim.m code was modified to account for Coulomb Friction as the viscous friction
equation under the ‘for-loop’ was replaced by the equation for friction force. The graph of
friction force versus velocity was obtained after implementing these changes as shown in Figure
6. The static friction force at the peak is 0.369N and kinetic friction force is 0.247N.

6
 Figure 6: Friction Force vs. Velocity

Comparing the Matlab and experimental results obtained from Lab 1 when accounting for
Viscous Friction with the results obtained from Lab 2 when accounting for Coulomb Friction, it
is clear that Coulomb Friction is best in describing the friction forces on sliding mass. The curves
generated in Lab 2 proves that Coulomb is the better method as both experimental and MATLAB
data closely align. On the other hand, the MATLAB and experimental data do not match when
Viscous friction method was used in Lab 1 as the experimental data show faster damping
compared to the MATLAB simulation.

7
CONCLUSION
Lab 1 and Lab2 were performed using two methods to account for friction force which are
Viscous and Coulomb friction methods. In both labs, Vernier Video Analysis software was used
to obtain experimental data, variables were calculated, and MATLAB was used for simulation.
The results show that Coulomb friction represents the friction forces experienced by the system.

8
REFERENCES

Rideout, G. (2023). ME6303 Mechanical Vibrations Lab 2 - Numerical Integration and Video
Analysis -Simulation of Coulomb Friction

SlidingMassCoulombFriction Video. (n.d.).

Vernier Video Analysis. (n.d.). Retrieved from https://videoanalysis.app/auth/.

Vibration, M.-6. M. (n.d.). Damped Free Vibration.

9
APPENDIX
lab2deriv.m

function derivs = lab2deriv(t,intvars)

%
% The function takes the current value of time t, and the current
% value of a vector called intvars, which has the same dimension
% as the initial conditions vector. In this example, intvars = [x,v].
% The function computes a column vector of derivatives.
%
% Give the elements of intvars physically meaningful names
%
x = intvars(1);
v = intvars(2);
%
global m k m_u m_o N n a b g
m = 0.5;
k = 44.2;
m_o=0.247;
n=2;
g=2;
a=7;
b=100;
thet1=75;
m_u = m_o*(tanh(v)+(a*v/(1+b*v^(2*g)+n*v)));
N=9.81*m*cos(deg2rad(thet1));
%
derivs(1) = v; %derivative of x
derivs(2) = (-m_u*N-(k*x))/m; %derivative of v
%
derivs = [derivs(1);derivs(2)]; % result must be a column vector,
% not a row vector

10
lab2Sim.m

clear all % clears variables from memory

% Define initial conditions and time window
ti = 3.6778; % start time (adjust this to 3.5 seconds)
tf = 5.756286; % end time
xo = -0.104; % adjust initial position accordingly
vo = 0; % adjust initial velocity accordingly
[T,OUT] = ode45(@lab2deriv,[ti tf],[xo vo]);
% [ti tf] is the time range. [xo vo] are initial conditions.
% mkderivs.m is a program that calculates the slopes of x and v
% ode45 takes the current values of x and v, along with the slopes,
% and finds the next value of x and v. It does this until t = tf.
global m k a b g n m_u m_o F_f N % this global declaration lets us define the
parameters
% in the derivs.m file and use them now for post-processing
% Plot outputs
%
%num1=xlsread('viblab.xlsx')
t_xl=xlsread('viblab.xlsx', 'vib','A2:A59');
figure % opens a new figure
subplot(2,1,1)
plot(T,OUT(:,1)); % plots “everything comma 1” of the X vector
ylabel('x [m]')
axis([3.6778 5.76 -0.11 0.11])
x_xl=xlsread('viblab.xlsx','vib', 'B2:B59');
hold on
plot(t_xl,x_xl)
hold off
legend('Simulation', 'Excel Data'); % Add legend
%
subplot(2,1,2)
plot(T,OUT(:,2));
y_xl=xlsread('viblab.xlsx','vib', 'G2:G59');
xlim([3.6778 5.76]);
ylim([-1.5 1.5])
hold on
plot(t_xl,y_xl)
ylabel('v [m/s]')
xlabel('Time [s]')
hold off
legend('Simulation', 'Excel Data'); % Add legend
%
% Calculating and plotting spring force
%
L = length(T); % finds the number of elements in T
v = OUT(:,2); % give the position column a physically meaningful
% name
for i = 1:L

11
for m_u=m_o*(tanh(v(i))+(a*v(i)/(1+b*v(i)^(2*g)+n*v(i))))
F_f(i) = N*m_u; %(i);
end
end;
figure
plot(v,F_f) % note that weVre plotting Fs vs. x, not vs. time
title('Friction Force')
xlabel('v [m/s]')
ylabel('F_f [N]')

12