Transcript Sect. 2.7

Astronomy
before computers!
Sect. 2.7: Energy Function & Energy
Conservation
• One more conservation theorem which we would
expect to get from the Lagrange formalism is:
CONSERVATION OF ENERGY.
• Consider a general Lagrangian L, a function of the
coords qj, velocities qj, & time t:
L = L(qj,qj,t)
(j = 1,…n)
• The total time derivative of L (chain rule):
(dL/dt) = ∑j(∂L/∂qj)(dqj/dt) + ∑j(∂L/∂qj)(dqj/dt) + (∂L/∂t)
Or:
(dL/dt) = ∑j(∂L/∂qj)qj + ∑j(∂L/∂qj)qj + (∂L/∂t)
• Total time derivative of L:
(dL/dt) = ∑j(∂L/∂qj)qj + ∑j(∂L/∂qj)qj + (∂L/∂t) (1)
• Lagrange’s Eqtns: (d/dt)[(∂L/∂qj)] - (∂L/∂qj) = 0
Put into (1)
(dL/dt) = ∑j(d/dt)[(∂L/∂qj)]qj + ∑j(∂L/∂qj)qj + (∂L/∂t)
Identity: 1st 2 terms combine
(dL/dt) = ∑j(d/dt)[qj(∂L/∂qj)] + (∂L/∂t)
Or: (d/dt)[∑jqj(∂L/∂qj) - L] + (∂L/∂t) = 0
(2)
(d/dt)[∑jqj(∂L/∂qj) - L] + (∂L/∂t) = 0 (2)
• Define the Energy Function h:
h  ∑jqj(∂L/∂qj) - L = h(q1,..qn;q1,..qn,t)
• (2)

(dh/dt) = - (∂L/∂t)
 For a Lagrangian L which is not an explicit function of time
(so that (∂L/∂t) = 0)
(dh/dt) = 0 & h = constant (conserved)
• Energy Function h = h(q1,..qn;q1,..qn,t)
– Identical Physically to what we later will call the
Hamiltonian H. However, here, h is a function of n indep
coords qj & velocities qj. The Hamiltonian H is ALWAYS
considered a function of 2n indep coords qj & momenta pj
• Energy Function h  ∑jqj(∂L/∂qj) - L
• We had (dh/dt) = - (∂L/∂t)
 For a Lagrangian for which (∂L/∂t) = 0
(dh/dt) = 0 & h = constant (conserved)
• For this to be useful, we need a
Physical Interpretation of h.
– Will now show that, under certain circumstances,
h = total mechanical energy of the system.
Physical Interpretation of h
• Energy Function h  ∑jqj(∂L/∂qj) - L
• Recall (Sect. 1.6) that we can always write KE as:
T = M0 + ∑jMjqj + ∑jMjkqjqk
M0  (½)∑imi(∂ri/∂t)2 , Mj ∑imi(∂ri/∂t)(∂ri/∂qj)
Mjk  ∑i mi(∂ri/∂qj)(∂ri/∂qk)
Or (schematically) T = T0(q) + T1(q,q) + T2(q,q)
– T0  M0 independent of generalized velocities
– T1  ∑jMjqj linear in generalized velocities
– T2  ∑jMjkqjqk quadratic in generalized velocities
• With almost complete generality, we can
write (schematically) the Lagrangian for most
problems of interest in mechanics as:
L = L0(q,t) + L1(q,q,t) + L2(q,q,t)
L0  independent of the generalized velocities
L1  linear in generalized the velocities
L2  quadratic in generalized the velocities
– For conservative forces, L has this form. Also
does for some velocity dependent potentials,
such as for EM fields.
L = L0(q,t) + L1(q,q,t) + L2(q,q,t)
(1)
• Euler’s Theorem from mathematics:
If f = f(x1,x2,.. xN) = a homogeneous function
of degree n of the variables xi, then
∑ixi(∂f/∂xi) = n f
(2)
• Energy Function h  ∑jqj(∂L/∂qj) - L
(3)
• For L of form (1):
(2)  h = 0L0 + 1L1+2L2 - [L0 + L1 + L2]
or
h = L2 - L0
L = L0(q,t) + L1(q,q,t) + L2(q,q,t)
 Energy function h  ∑jqj(∂L/∂qj) - L = L2 - L0
• Special case (both conditions!):
a.) The transformation eqtns from Cartesian to
Generalized Coords are time indep.
 In the KE, T0 = T1 = 0

T = T2
b.) V is velocity indep.  L2 = T = T2 & L0 = -V
 h = T + V = E  Total Mechanical Energy
• Under these conditions, if V does not depend on t,
neither does L & thus (∂L/∂t) = 0 = (dh/dt)
so h = E = constant (conserved)
Energy Conservation
• Summary: Different Conditions:
Energy Function h  ∑jqj(∂L/∂qj) - L
ALWAYS: (dh/dt) = - (∂L/∂t)
SOMETIMES: L does not depend on t

(∂L/∂t) = 0, (dh/dt) = 0 & h = const. (conserved)
USUALLY: L = L0(q,t) + L1(q,q,t) + L2(q,q,t)

h  ∑jqj(∂L/∂qj) - L = L2 - L0
SOMETIMES: T = T2 = L2 AND L0 = -V

h = T + V = E  Total Mechanical Energy
 Conservation Theorem for Mechanical Energy:
If h = E AND L does not depend on t, E is conserved!
• Clearly, the conditions for conservation of
energy function h are DISTINCT from those
which make it the total mechanical energy E.
 Can have conditions in which:
1. h is conserved & = E
2. h is not conserved & = E
3. h is conserved &  E
4. h is not conserved &  E
Most common case in classical (& quantum) mech. is case 1.
• Stated another way: Two questions:
1. Does the energy function h = E for the system?
2. Is the mechanical energy E conserved for the system?
• Two aspects of the problem! DIFFERENT questions!
– May have cases where h  E, but E is conserved.
– For example: A conservative system, using generalized
coords in motion with respect to fixed rectangular axes:

Transformation eqtns will contain the time

T will NOT be a homogeneous, quadratic function
of the generalized velocities!
 h  E, However, because the system is conservative, E is
conserved! (This is a physical fact about the system,
independent of coordinate choices!).
• It is also worth noting:
The Lagrangian L = T - U is independent
of the choice of generalized coordinates.
The Energy function h  ∑jqj(∂L/∂qj) - L
depends on the choice of generalized
coordinates.
• The most common case in classical (&
quantum) mechanics is h = E and E is
conserved.
Non-Conservative Forces
• Consider a non-conservative system: Frictional forces
obtained from the dissipation function ₣ . Derivation
with the energy function h becomes:
(dh/dt) + (∂L/∂t) = - ∑jqj(∂₣ /∂qj)
• Ch. 1: The formulation of ₣ shows it is a
homogeneous, quadratic function of the q’s.
 Use Euler’s theorem again: ∑jqj(∂₣ /∂qj) = 2₣

(dh/dt) = - (∂L/∂t) - 2₣
• If L is not an explicit function of time (∂L/∂t =0 )
AND h = E: (dE/dt) = - 2₣
– That is, under these conditions, 2₣ = Energy dissipation rate.
Symmetry Properties & Conservation
Laws (From Marion’s Book!)
• In general, in physical systems:
A Symmetry Property of the System
 Conservation of Some Physical Quantity
Also: Conservation of Some Physical Quantity
 A Symmetry Property of the System
• Not just valid in classical mechanics! Valid in
quantum mechanics also! Forms the foundation of
modern field theories (Quantum Field Theory,
Elementary Particles,…)
We’ve seen in general that:
Conservation Theorem: If the Generalized
Coord qj is cyclic or ignorable, the
corresponding Generalized (or Conjugate)
Momentum, pj  (∂L/∂qj) is conserved.
• An underlying symmetry property of the system:
If qj is cyclic, the system is unchanged (invariant)
under a translation (or rotation) in the “qj direction”.

pj is conserved
Linear Momentum Conservation
• Conservation of linear momentum:
If a component of the total force vanishes,
nF = 0, the corresponding component of total
linear momentum np = const (is conserved)
• Underlying symmetry property of the system:
The system is unchanged (invariant) under a
translation in the “n direction”.

np is conserved
Angular Momentum Conservation
• Conservation of angular momentum:
If a component of total the torque vanishes,
nN = 0, the corresponding component of total
angular momentum nL = const (conserved)
• Underlying symmetry property of the system:
The system is unchanged (invariant) under a
rotation about the “n direction”.

nL is conserved
Energy Conservation
• Conservation of mechanical energy:
If all forces in the system are conservative, the
total mechanical energy E = const (conserved)
• Underlying symmetry property of the system:
(More subtle than the others!) The
system is
unchanged (invariant) under a time reversal.
(Changing t to -t in all eqtns of motion)

E is conserved
Summary: Conservation Laws
• Under the proper conditions, there can be up to
7 “Constants of the Motion”  “1st Integrals
of the Motion”  Quantities which are
Conserved (const in time):
Total Mechanical Energy (E)
3 vector components of Linear Momentum (p)
3 vector components of Angular Momentum (L)