Thermodynamics without entropy or analogies among dynamical systems • please note the distinction between Static and Dynamic energy storages • small w denotes total energy, capital W denotes energy density • systems can be “lumped parameter” or distributed • Four variable models -- note the “oddball” down in the lower left corner • yes, Virginia, dynamic is really different from static energy storage (you always knew there was something strange about kinetic energy and momentum) • Primary network variables are the ones whose product is power (or power flux) and whose ratio is impedance -- the use of any other variable pair as the primary variables impedes building complicated systems from elementary building blocks; although it can, at times, be thermodynamically advantageous
Copyright 2001 Kevin G. Rhoads
v
q
voltage
charge
Push
stuff
λ
i
flux-turns linkage current
Thud
Flow
Impedance = v / i
C
v d⁄dt
wdynamic = ∫i dλ
wstatic = ∫v dq
q d⁄dt
λ
i L
Power = v i
Uncertainty Pair
Electrical
Copyright 2001 Kevin G. Rhoads
Lumped Element
F
x
Force
position
Push
stuff
p
v
momentum
velocity
Thud
Flow
Mechanical Impedance = F / v
1/k
F d⁄dt
wdynamic = ∫v • dp Uncertainty Pair
wstatic = ∫F • dx
x d⁄dt
p
v m
Power = F v
Linear Mechanical
Variables (i.e., F, x, v, p) may be scalars or vectors.
Copyright 2001 Kevin G. Rhoads
Lumped Element
T L
θ
Torque
position
Push
stuff
velocity
Thud
Flow
(angular)
ω
momentum
Mechanical Impedance = T / ω
1/k
θ
T d⁄dt
wdynamic = ∫ω dL
wstatic = ∫Tdθ d⁄dt
ω
L I0
Power = T ω
Uncertainty Pair
Rotational Mechanical Copyright 2001 Kevin G. Rhoads
Lumped Element
T
Q
?
I[Q´]
Temperature
heat, energy
Push
stuff
heat flow
Thud
Flow
Nameless CThermal
Thermal Impedance = T / I
T
Q
d⁄dt
d⁄dt
?
I ??
Thermal An example of an incomplete analogy. There is nothing to put into the “thud” slot in
Lumped Element
general; although for studies of TTT processing, “soakage” might be a reasonable candidate.
Copyright 2001 Kevin G. Rhoads
E
D
Electric field Intensity
B
H
Magnetic Flux Density
Wdynamic = ∫H • dB Uncertainty Pair, in some extended sense
*
Magnetic Field intensity
Push
stuff
Thud
Flow
ε0
Wave Impedance = |E| / |H| ∂⁄∂t, ∇×
Displacement Flux Density
E
D
B*
H µ0
∂⁄∂t, ∇×
Wstatic = ∫E • dD
Power Flux Density = E × H
Electrical Fields
instead of B one could use A, this does break the
symmetry of the analogy and is not recommended
Copyright 2001 Kevin G. Rhoads
Distributed System
P
Q [V]
Pressure Differential
Volume
p
Q´
Momentum Density
Volume rate of flow
stuff
Thud
Flow
Gas Law
Flow Impedance = P / Q´
P
Wstatic = ∫P dQ
Q
∂⁄∂t
Wdynamic = ∫Q´ dp
Push
∂⁄∂t
p
Q´ ρ
Power Flux Density = P Q´
Gas Flow Fluid Flow Copyright 2001 Kevin G. Rhoads
Distributed System
Pop Quiz (answers ahead) • What is the pneumatic equivalent of “ground bounce” • What is the electrical equivalent of “stiction” • Comparing the Mississippi River and a fire hose • which is higher impedance? • why? • Is a direct analogy between the old style automotive ignition and hammering in a nail reasonable? • If yes, why? • If no, why not?
Copyright 2001 Kevin G. Rhoads
P
ξ
Pressure variations displacement (Intensity)
Push
stuff
p
v
momentum density
Thud
Flow
velocity
Acoustic Impedance = P / v
1/M*
ξ
P ∂⁄∂t
Wdynamic = ∫v dp ? Uncertainty Pair, in some extended sense
*
Wstatic = ∫P dξ ∂⁄∂t
p
v ρ
Power Flux = P v
Acoustic
M is bulk modulus; for isothermal it is (for small strains) equal to the unperturbed pressure
Distributed System
for sound, however, adiabatic is more accurate than isothermal, so multiply by 1.4 for diatomic gases
Copyright 2001 Kevin G. Rhoads
• What is the pneumatic equivalent of “ground bounce” • Pressure rise in the pneumatic return line upon operation • What is the electrical equivalent of “stiction” • One example is crossover distortion in Class B push-pull stages • Comparing the Mississippi River and a fire hose • which is higher impedance? • Fire hose: impedance = Push/Flow • why? • Higher pressure (more push), lower flow • Is a direct analogy between the old style automotive ignition and hammering in a nail reasonable? • If yes, why? • Yes - both involve storing energy for rapid release as dynamic storage is stopped -- it is hard to get similar results with static storage • If no, why not?
Copyright 2001 Kevin G. Rhoads