Beruflich Dokumente
Kultur Dokumente
LAGRANGES EQUATIONS
(Analytical Mechanics)
1
Ex. 1: Consider a particle moving on a
fixed horizontal surface.
z
Let, r P be the
F1
position and F be O
P y
the total force on x m (x,y,0)
the particle.
The FBD is: -mgk F1
f N
p
The equation of motion is mr F(r, r, t )
2
In component form, the equation of motion is
mxP Fx ( x, y, z, x , y , z, t )
myP Fy ( x, y, z, x , y , z, t )
mzP Fz ( x, y, z, x , y , z, t )
Also, motion is restricted to xy plane
z = 0 - equation of constraint
It is a geometric restriction on where the
particle can go in the 3-D space.
Clearly, there is a constraint reaction
(force) that needs to be included in the
total force F. 3
Ex. 2: Consider a particle moving on a surface.
z
m
O Surface
f(x,y,z)=c
x y
Z m1
C
O m2
X Y
Z m1
C
O
m2
X Y
13
Fundamental difference:
A geometric constraint restricts the
configurations that can be achieved
during motion. Certain regions
(positions) are inaccessible
A kinematic constraint only restricts the
velocities that can be acquired at a given
position. The system can, however,
occupy any position desired (e.g.: one can
reach any point in the skating rink - it is
just that one cannot move in arbitrary
direction). 14
We can also write the constraints in the form:
(in differential form)
n
a ji ( q1 ,...., qn , t )dqi a jt ( q1 ,...., qn , t )dt 0,
i 1
j = 1, 2, ., d
In differential form
( x2 x1 )(dx2 dx1 ) ( y2 y1 )(dy2 dy1 )
( z2 z1 )(dz2 dz1 ) 0
or ( x1 x2 )dx1 ( x2 x1 )dx2 ( y1 y2 )dy1 ( y2 y1 )dy2
( z1 z2 )dz1 ( z2 z1 )dz2 0
or
(q1 q4 )dq1 (q4 q1 )dq4 (q2 q5 )dq2
(q5 q2 )dq5 (q3 q6 )dq3 (q6 q3 )dq6 0
21
constraint #4:
( x1 x2 )( y2 y1 ) ( y1 y 2 )( x2 x1 ) 0
In Differential form:
(q5 q2 )(dq1 dq4 ) (q4 q1 )(dq2 dq5 ) 0
or
(q5 q2 )dq1 (q4 q1 )dq2 (q5 q2 )dq4
(q4 q1 )dq5 (0)dq5 (0)dq6 0
22
differential form of constraints (in
general): n a dq a dt 0, j 1,2,3,.....m
ji i jt
i 1
25
e.g.: consider the constraint #3 (j = 3)
( q1 q4 )dq1 ( q4 q1 )dq4 ( q2 q5 )dq2
( q5 q2 )dq5 0
Here a j1 q1 q4 , a j 2 q2 q5 , a j 3 0
Thus a j1 / q2 0 a j 2 / q1
Similarly a j 4 q4 q1 , a j 5 q5 q2 , a j 6 0
and a j1 / q4 1 a j 4 / q1 , etc.
This constraint is integrable.
26
Also, a j1 / q3 0 a j 3 / q1
a j5
a j1 / q5 0 ; a j5 q5 q2
q1
a j1 / q6 0 a j 6 / q1 ; a jt / q1 a j1 / t 0
a j 2 / q3 0 a j 3 / q2
.
.
27
Now, consider constraint #4:
( q5 q2 )dq1 ( q4 q1 )dq2 ( q5 q2 )dq4
( q4 q1 )dq5 0
Then a41 q5 q2 , a42 ( q4 q1 ), a43 0,
28
Classification: an N particle system is said to be:
Holonomic - if all constraints are geometric,
or if kinematic - are integrable (reducible to
geometric).
Nonholonomic - if there is a constraint which
is kinematic and not integrable.
Schleronomic - all the constraints, geometric
as well as kinetic, are independent of time t
explicitly.
Rheonomic - if at least one constraint
depends explicitly on time t.
29
Possible and Virtual Displacements
Suppose that a system of N particles, with
position vectors r1 , r 2 ,, r N has d
geometric constraints
i ( r1 , r 2 , r 3 ,...., r N , t ) 0, i 1,2,3,...., d ,
and g kinematic constraints
N
l ji r j Dj 0 , i 1,2,..., g
i 1
and
N
l ji r j Dj 0 , i 1,2,..., g (2)
i 1
31
Possible velocities: the set of all velocities
which satisfy the (d + g) linear equations of
constraints.
3N > (d + g) infinity of possible velocities.
One of these is realized in an actual motion
of the system. Let
d ri ri dt , i 1,2,...., N
These are the possible (infinitesimal)
displacements. They satisfy
N
ij d r j
i
dt 0 , i 1,2,...., d (3)
j 1 t 32
N
and l ji d r i D j dt 0 , j 1,2,...., g (4)
i 1
ij (d r j d r j ) 0 , i 1,2,..., d
i 1
and N
l ji ( d r j d r j ) 0 , i 1,2,..., g
i 1
Possible velocities v u
v vR u vr
(v R relative velocity) P Surface
Possible displacements f(x,y,z)=c(t)
z
dr v dt (v R u )dt
Two possible displacements:
x y
dr (v R u )dt , d r (v R u )dt
36
Thus, a virtual displacement is
r dr dr (v R v R )dt rR rR
n
dr
u
u r
Note that et
d r is along absolute velocity direction,
whereas r is along relative velocity or
tangent to the surface (frozen constraint)
(set dt = 0). 37
Degrees-of-freedom:
N - number of particles
(d + g) - geometric + kinematic constraints
there are n = 3N - (d + g) independent
virtual displacements
Problem of Dynamics:
Given a system with - external forces
Fi F i (r, r, t ), i 1, 2,...., N ;
Initial positions r io , and initial velocities v io
compatible with constraints; we need to
38
determine the motion of the system of particles,
i.e., the positions ( r i (t )), the velocities ri , and the
constraint or reaction forces R i , i 1,2,, N .
mi
ri Fi R i , i 1,2,..., N
(3N equations)
N
ij r j i / t 0, i 1,2,..., d
j 1
(d equations)
N
l ji d r i D j dt 0, j 1,2,..., g
i 1
(g equations)
39
In these equations, the unknowns are: r i , R i -
6N unknowns
Thus, additional relations required:
6N - (3N + d + g) = 3N - (d + g) n
(equal to the number of degrees-of-freedom)
Need to define concept of workless constraints.
6.4 Virtual Work
Definition: A workless constraint is any
constraint such that the virtual work (work
done in a virtual displacement) of the
constraint forces acting on the system is zero
for any reversible virtual displacement. 40
Ex 10: Consider a double pendulum.
The positions are: X
r1 x1 i y1 j O
l
r 2 x2 i y 2 j A
Constraints are: 1 m
( x1 y1 ) 0
2 2 2
(1) 2
l
Y m
or x1 x1 y1 y1 0 B
(differential form)
2
( x2 x1 ) ( y2 y1 ) 2
2
0 (2)
41
or ( x2 x1 )( x2 x1 ) ( y2 y1 )( y 2 y1 ) 0
(differential form)
Consider FBDs
Then, T1 1
1
1
tan ( x1 / y1 ) A
1 ( x2 x) T2
2
tan
y2 y1 mg 2
F mr A
Y
FBD: Ry
The reaction A
force is: Rx
R=Rx i+ Ry j
mg
44
Newtons 2nd law gives
x : Rx mx (1)
y : Ry mg my (2)
Constraint on motion is:
2 2 2
x y l 0 (3)
or xx yy 0 ( r r 0)
(differentiated form)
or x dx + y dy = 0
(differential form)
45
Counting: 3 equations
4 variables x, y , Rx , Ry
external as well
as constraint forces
not accounted for in workless constraint forces.
52
Static equilibrium for the ith particle
F i R i 0, i 1,2,..., N
ij r j 0, i 1,2,..., d geometric
j 1
N
l ji r i 0, j 1,2,..., g kinematic
i 1
O
The constraints are: the ends must remain in
contact with respective surfaces. 56
To properly set up the problem, we need to
first define a coordinate system so that the
appropriate position vectors can be defined.
Then, we can define the constraints and the
virtual displacements.
Let z A and zB- positions of A and B along
the inclined surfaces. Also, - the angle of
inclination of the rod. The constraints are:
cos ( Z A Z B )cos
sin (Z B Z A )sin
57
The variables are described here on the
picture more clearly:
B
G
h A
zB
O
zA
58
A possible set of virtual displacements
consistent with constraints are shown here:
G B
A
h B
G zB
zA A zB
zA O
59
FBD:
B
G
A RB
W
RA
N
l ij d r j Di ( r, t )dt 0, i 1,2,...., g
j 1
64
Then, the relations satisfied by virtual
displacements are
N
( fi / r j ) r j 0, i 1,2,...., d
j 1
and
N
l ij r j 0, i 1,2,...., g
j 1
Workless constraints
N DAlemberts
(mi
ri F i ) ri 0
i 1 Principle
(a single scalar equation)
particle of mass m O
r (a b cos t ),
a b 0. r(t)
P e
X Y
e r , e in OPZ plane;
r Z e
e to OPZ plane er
mg
67
position of the ball: r P re r
velocity: r P re r re r
or r P re r r e r sin e
acceleration: r ( r r 2 r 2 sin 2 )e r
(r 2r r sin cos )e ( r sin
2
2r sin 2r cos )e
virtual displacement:
possible velocity r P re r r e r sin e
possible displ. d r dre r rd e rd sin e
constraint: r a b cos t 68
or dr (b sin t )dt
sinceconstraint frozen dt 0 dr 0
virtual displacement: r r e r sin e
External force acting:
F mg cos er mg sin e mgK
DAlemberts Principle
(mr F ) r 0
Note: on the FBD of the particle, tension
force also acts along the rod - a workless
constraint force
69
mr[ g sin (r
2r r sin cos )]
2
71
Thus, there are:
4 (or 6 counting zs) variables or generalized
coordinates
2 (or 4 if z included) constraints ( z1 0, z2 0)
n = degrees of freedom = 2.
Need only 2 independent variables (for
geometric constraints case) to specify the
configuration at any given time
e.g.: let y1 , y2 be the two chosen independent
coordinates. Then, we can write
x1 (l12 y12 ), x2 (l22 y22 ) 72
r1 l12 y12 ) i y1 j r1 ( y1 )
2 2 2 2
r2 { l 1
y 1
(l 2
y )}i ( y1
2
y2 ) j
r 2 ( y1 , y2 )
Note: geometric constraint now
automatically satisfied.
Another possible choice of generalized
coordinates are:
1, 2 angles with y axis
73
Then
x1 l1 sin 1 , y1 l1 cos 1
x2 l2 sin 2 , y2 l2 cos 2
r1 l1 (sin 1 i cos 1
j) r1 ( 1 , 2 )
r2 l1 (sin 1 l2 sin 2 )i
(l1 cos 1
l2 cos 2 ) j r2 ( 1, 2 )