Formulation of governing differential equation of tapered beam

The fundamental beam vibrating equation:

Shape factors : n m

And as mentioned on the problem the depth assumed to be constant

Here we considered only the free vibration, so considering the motion to be of form y(x, t) =z(x) sin ( t),
so applying the following relation to the fundamental beam equation we get:

letting u = x/l (where u varies from 0 to 1) the following equation is obtained

By proper approximation i.e. α=h0/h1 and β=b0/b1 and k Egh 12 / 2 2 1 above

is the final equation of motion for a double-tapered beam with rectangular cross-section. It was
solved by numerical integration to give values of (lk) for various taper ratios for clamped-free
beam with boundary conditions i.e. at x= 0 or u = 0, d2z/du2= 0 and z= 0, at x= l or u= 1, dz/du=
0 and z= 0. This after solving leads

For the Second Part Of the Problem When a system is subjected to free vibration and the system is
considered as a discrete system in which the beam is considered as mass-less and the whole mass is
concentrated at the free end of the beam. The governing equation of motion for such system will be,

Where m is a concentrated mass at the free end of the beam and k is the stiffness of the system. The
transverse stiffness of a cantilever beam is given as
for a rectangular cross section

Consider a small element of length dx at a distance x from the free end .The beam displacement at
this point is given by (Timoshenko and Young, 1961),

Here fl3/3EI is the deflection at free end of the cantilever beam. Now the velocity of the small
element at distance x is given by,

Hence, the kinetic energy of the element is given by

and the total kinetic energy of the beam is

where m1l = mb. If we place a mass of 33/140mb at the free end of the beam and the beam is
assumed to be of negligible mass, then

Total kinetic energy possessed by the beam =

Hence two systems are dynamically same. Therefore, the continuous system of cantilever beam
can be changed to single degree freedom system as shown in Fig (b) by adding the 33 /140mb of
mass at its free end.

Values of the mass density for various beam materials are given in Table 2.1. If any contacting
type of transducer is used for the vibration measurement, it should be placed at end of the beam
and then the mass of transducer has to be added into the equivalent mass of the beam at the free
end of the beam during the natural frequency calculation. If mt is the mass of transducer, then the
total mass at the free end of the cantilever beam is given as,

The Matlab Code :

function [phi,wn] = eigenModes(geometry,BC,Nmodes)

%
%% check input
if ~isnumeric(BC), error('BC must be an integer between
1 and 4');end
if max(ismember([1,2,3,4],BC))==0, error('BC is unknown.
Please, choose BC = 1,2,3 or 4');end
if ~isnumeric(Nmodes), error('Nmodes must be an
integer');end

name = [{'L'},{'E'},{'nu'},{'rho'},{'I'},{'y'},{'m'}];
for ii=1:numel(name),
if ~isfield(geometry,name{ii}),
error([' The field ',name{ii},' is missing in the
structure "geometry".'])
end
end

%%
y= geometry.y;
Nz = numel(y); % number of discrete elements

% volume V and mass m of the beam

m = geometry.m;
L = geometry.L;

% get the non trivial solution of f

if Nmodes<10,
Ndummy=1:Nmodes^2; % the number is arbitrary fixed as
the square of the number of Nmodes.
else
Ndummy=1:100; % the number is arbitrary fixed as the
square of the number of Nmodes.
end
tolX = 1e-8;
tolFun = 1e-8;
options=optimset('TolX',tolX,'TolFun',tolFun,'Display','off
');
h =fsolve(@modeShape,Ndummy,options);
% small values of h come from numerical errors
h(h<0.4)=[];
% Analytically, many solutions are identical to each other,
but numerically, it is not the
% case. Therefore I need to limit the prceision of the
solutions to 1e-4.
h = round(h*1e4).*1e-4;
% the uniques solution are called beta:
beta = unique(h);
beta = beta(1:min(numel(beta),Nmodes));

%modes shapes calculations

if Nmodes>numel(beta),
warning([' Mode shape required is too high. The number
is set to',num2str(numel(beta)),' \n\n'])
Nmodes=numel(beta)
end
wn =
beta.^2.*sqrt(geometry.E*geometry.I./(m.*geometry.L^4)); %
in rad
phi = zeros(Nmodes,Nz);

for ii=1:Nmodes,
switch BC,

case 1,% pinned-pinned

phi(ii,:)=sin(beta(ii)*y./L);

phi(ii,:)=phi(ii,:)./max(abs(phi(ii,:)));

case 2 % clamped-free

phi(ii,:)=(cosh(beta(ii)*y./L)-
cos(beta(ii)*y./L))+...

((cosh(beta(ii))+cos(beta(ii))).*(sin(beta(ii)*y./L)-
sinh(beta(ii)*y./L)))./...
(sin(beta(ii))+sinh(beta(ii)));

phi(ii,:)=phi(ii,:)./max(abs(phi(ii,:)));

case 3 % clamped-clamped

phi(ii,:)=sin(beta(ii)*y./L)-
sinh(beta(ii)*y./L)-...
 ((sin(beta(ii))-
sinh(beta(ii))).*(cos(beta(ii)*y./L)-
cosh(beta(ii)*y./L)))./...
(cos(beta(ii))-cosh(beta(ii)));

phi(ii,:)=phi(ii,:)./max(abs(phi(ii,:)));

case 4 % clamped-pinned

phi(ii,:)=sin(beta(ii)*y./L)-
sinh(beta(ii)*y./L)-...
((sin(beta(ii))-
sinh(beta(ii))).*(cos(beta(ii)*y./L)-
cosh(beta(ii)*y./L)))./...
(cos(beta(ii))-cosh(beta(ii)));

phi(ii,:)=phi(ii,:)./max(abs(phi(ii,:)));
end

end

% Function to solve is different for each BC:

function [f] = modeShape(y)

switch BC,
case 1,% pinned-pinned
f=sin(y).*sinh(y);
case 2 % clamped-free
f=cos(y).*cosh(y)+1;
case 3 % clamped-clamped
f=cos(y).*cosh(y)-1;
case 4 % clamped-pinned
f=cos(y).*sinh(y)-sin(y).*cosh(y);
end
end

end

%%======
%% Definition of the geometry
clearvars;close all;clc;
geometry.L = 60; % beam length (m)
geometry.E = 200*10^9; % Young Modulus (Pa)
geometry.nu = 0.3; % Poisson ratio
geometry.rho = 7800; % density (kg/m^3)

% Rectangular beam
%B = geometry.L/100;
B = 1;
H = 2;
%H = geometry.L/100;
Iy = B*H.^3/12; % quadratic moment
Ix = H*B.^3/12; % quadratic moment
geometry.I = Iy; % affectation of quadratic moment
% number of discretisation points for the beam
geometry.y = linspace(0,geometry.L ,100);
V = B*H*geometry.L;
geometry.m = geometry.rho.*V./geometry.L; % in kg/m

% Number of modes
Nmodes =4; % number of mode wanted

%% modal analysis: Case 2

BC = 2;% clamped-free

[phi,wn] = eigenModes(geometry,BC,Nmodes);

figure
for ii=1:Nmodes,
subplot(Nmodes,1,ii)
box on;grid on
plot(geometry.y,phi(ii,:));
ylabel(['\phi_',num2str(ii)])
title(['w_',num2str(ii),' = ',num2str(wn(ii),3),'
rad/s']);
end
set(gcf,'color','w')
xlabel('y (m)');