Vous êtes sur la page 1sur 9

ModelingofaSimplePendulum

SimpleLinearModel:

Thesimplestanalyticalmodelforastandardpendulumcanbederivedbyassumingno
drag force and small variations in the angle of the pendulum. Assume that the entire
mass of the pendulum is concentrated in the bob and that the length from the pivot
pointtothecenterofgravityofthebobisgivenby .
FindthesumofthemomentsaroundthepointOtakingthecounterclockwisedirection
aspositive.Notethatthetension, ,inthethinroddoesnotcontributetothemoment
aboutthepointO.

0
0

Themassmomentofinertiaofthependulum, ,isgivenby:

Theangularaccelerationofthependulum, ,isthesecondderivativeoftheangle .

Thedifferentialequationgoverningthedynamicsofthependulumbecomes:

Dividing through by
differentialequation:

and rearranging terms yields the following second order

Although this equation appears simple enough, it is nonlinear in nature due to the
presence of a transcendental function. This nonlinear equation can, however, be
linearizedbyassumingsmallvariationsintheangleofthependulum, .
Forsmallangles(expressedinradians),
isapproximatelyequalto .Forexample:

22.5

0.3927

0.3827

Therefore for small variations in angle, the linear differential equation describing the
dynamicsofasimplependulumcanbeapproximatedby:

ThisequationcanbesolvedanalyticallybyconsideringthebasicpropertiesofLaplace
transforms. Recall that the second derivative of a function can be expressed in the
Laplacedomainasfollows:

Thedifferentialequationdescribingthedynamicsofthependulumbecomes:

Assume that the pendulum is released from some initial angle,


0).
(

, no initial velocity

Solvingfor yields:

ThiscaneasilybeputintotheformofthefollowingbasicLaplacetransform:

The angle of the pendulum, , as a function of time based on a simple, linear model
becomes:

Although it is not necessary to use MatlabSimulink to visualize this function, for


comparisonpurposestheSImulinkmodelbecomes:

SimpleNonLinearModel:

Althoughfindingananalyticalsolutiontothesimplenonlinearmodelisnotevident,it
canbesolvednumericallywiththeaidofMatlabSimulink.Startbyisolatingthehighest
orderderivativefromthelinearmodel:

TheMatlabSimulinkmodelthatsolvesthisequationisgivenby:

NonLinearModelIncludingaConstantAerodynamicDragForce:

Assumethattheaerodynamicdragactsonlyonthebob;i.e.,thedragonthethinrodis
negligible.Thedragforce, ,imposesanadditionalmomentaboutthepointO.
Thedifferentialequationgoverningthedynamicsofthependulumbecomes:

Isolatingthehighestorderderivativeof yields:

TheMatlabSimulinkmodelthatsolvesthisequationisgivenby:

Note that the drag force must change sign based on the angular velocity of the
pendulum.Inotherwords,althoughthedragforceistakenasconstant,itmustalways
actoppositetothedirectionofmotion.

NonLinearModelIncludingAerodynamicDragForcethatisafunctionofVelocity:

Theaerodynamicdragforincompressibleflowoveranybodyisgivenby:

where isthevelocityoftheincompressiblefluid, isthedensityofthefluid, isthe


frontalarea(note:notthesurfacearea),and isthedragcoefficient.
ThedragcoefficientforacylindricalbodyisafunctionoftheReynoldsnumberasshown
inthefigureonthefollowingpage.

RecallthattheReynoldsnumberisgivenby:

where isthediameterofthecylinderand istheviscosityoftheincompressiblefluid


(inthiscaseair).

The MatlabSSimulink mo
odel for thependulumw
with variablee aerodynam
mic drag isggiven
by:

NonLinearModelIncludingVariableAerodynamicDragandFrictionatthePivot:

Therecouldpotentiallybeasignificantamountoffrictionatthepivotpoint.
The MatlabSimulink model including aerodynamic drag and Coulomb friction at the
pivotpointisgivenby:

Theresponseofthesystembasedonaninitialpendulumangleof22.5degreesis
illustratedbelow.

Thefollowingtablesummarizestheparametersusedinthesimulation:
T0 = 22.5*pi/180;
L = 0.6;
g = 9.81;

%% Initial pendulum angle (radians)


%% Length of rod (distance to CG) (m)
%% Acceleration due to gravity (m/s^2)

M = 0.06;
FD = 0.0007;

%% Mass of bob (kg)


%% Drag Force (N)

DB = 0.0333;
HB = 0.0508;
A = DB*HB;

%% Diameter of bob (m)


%% Height of bob (m)
%% Frontal area of bob (m^3)

RHO = 1.23;
MEW = 1.79E-5;
CD = 1;

%% Density of air (kg/m^2)


%% Viscosity of air (kg/(m s))
%% Drag Coefficient

KD = 0.5*CD*A*RHO;
TFF = 0.0005;

%% Drag constant (kg/m)


%% Friction torque at pivot (N-m)