AMATH 353
Lecture 2
Weston Barger
University of Washington
[email protected]
June 16, 2016
Partial Differential Equation
Definition
Partial Differential Equation (PDE) A partial differential
equation for a function u(x, t) is a differential equation that
relates derivatives of u(x, t).
Partial Differential Equation
Definition
Partial Differential Equation (PDE) A partial differential
equation for a function u(x, t) is a differential equation that
relates derivatives of u(x, t).
Question
How does the definition of partial differential equation differ from
ordinary differential equation?
Partial Differential Equation
Definition
Partial Differential Equation (PDE) A partial differential
equation for a function u(x, t) is a differential equation that
relates derivatives of u(x, t).
Question
How does the definition of partial differential equation differ from
ordinary differential equation?
Answer
It does not.
I Because u(x, t) is a function of two variables, partial
derivatives are involved.
Shorthand
We often use the shorthand
∂u ∂u
ut = , ux = ,
∂t ∂x
∂2u ∂2u
uxx = , utx =
∂x 2 ∂x∂t
etc...
Shorthand
We often use the shorthand
∂u ∂u
ut = , ux = ,
∂t ∂x
∂2u ∂2u
uxx = , utx =
∂x 2 ∂x∂t
etc... I sometimes write
∂u ∂u
∂t u = , ∂x u = ,
∂t ∂x
∂2u ∂2u
∂x2 u = , 2
∂xt u=
∂x 2 ∂x∂t
Advection Equation
Example
The PDE
ut + cux = 0 (1)
where c is a parameter is called the advection equation or the
transport equation.
Advection Equation
Example
The PDE
ut + cux = 0 (1)
where c is a parameter is called the advection equation or the
transport equation.
I Linear: (1) has no products of u with itself or derivatives of u
Advection Equation
Example
The PDE
ut + cux = 0 (1)
where c is a parameter is called the advection equation or the
transport equation.
I Linear: (1) has no products of u with itself or derivatives of u
I Homogeneous: u(x, t) = 0 is a solution
Advection Equation
Example
The PDE
ut + cux = 0 (1)
where c is a parameter is called the advection equation or the
transport equation.
I Linear: (1) has no products of u with itself or derivatives of u
I Homogeneous: u(x, t) = 0 is a solution
I First order: the highest derivatives are first derivatives
Diffusion Equation
Example
The diffusion equation a.k.a the heat equation is given by
ut = σuxx , (2)
where σ > 0 is a parameter.
Diffusion Equation
Example
The diffusion equation a.k.a the heat equation is given by
ut = σuxx , (2)
where σ > 0 is a parameter.
I Linear: (2) has no products of u with itself or derivatives of u
Diffusion Equation
Example
The diffusion equation a.k.a the heat equation is given by
ut = σuxx , (2)
where σ > 0 is a parameter.
I Linear: (2) has no products of u with itself or derivatives of u
I Homogeneous: u(x, t) = 0 is a solution
Diffusion Equation
Example
The diffusion equation a.k.a the heat equation is given by
ut = σuxx , (2)
where σ > 0 is a parameter.
I Linear: (2) has no products of u with itself or derivatives of u
I Homogeneous: u(x, t) = 0 is a solution
I Second order: second order in x, first order in t
Diffusion Equation
Example
I The two-dimensional diffusion equation is given by
ut = σ(uxx + uyy ),
where σ > 0 is a parameter. Could model heat spreading on a
metal sheet.
Diffusion Equation
Example
I The two-dimensional diffusion equation is given by
ut = σ(uxx + uyy ),
where σ > 0 is a parameter. Could model heat spreading on a
metal sheet.
I In three dimensions,
ut = σ(uxx + uyy + uzz ).
Could model heat spreading in a room.
Diffusion Equation
Example
I The two-dimensional diffusion equation is given by
ut = σ(uxx + uyy ),
where σ > 0 is a parameter. Could model heat spreading on a
metal sheet.
I In three dimensions,
ut = σ(uxx + uyy + uzz ).
Could model heat spreading in a room.
Question
Are these equations linear? Homogeneous? What order are they in
x and t?
Burger’s Equation
Example
The equation
ut + uux = 0 (3)
is known as the (inviscid) Burger’s equation. This equation can
be used to model fluid transport.
Burger’s Equation
Example
The equation
ut + uux = 0 (3)
is known as the (inviscid) Burger’s equation. This equation can
be used to model fluid transport.
I Nonlinear
Burger’s Equation
Example
The equation
ut + uux = 0 (3)
is known as the (inviscid) Burger’s equation. This equation can
be used to model fluid transport.
I Nonlinear
I Homogeneous
Burger’s Equation
Example
The equation
ut + uux = 0 (3)
is known as the (inviscid) Burger’s equation. This equation can
be used to model fluid transport.
I Nonlinear
I Homogeneous
I First order
This equation provides a classic example of shock waves. (Queue
Mathematica)
Wave Equation
Example
The equation
utt = c 2 uxx (4)
is known as the wave equation. This equation is used to model a
plucked guitar string.
Wave Equation
Example
The equation
utt = c 2 uxx (4)
is known as the wave equation. This equation is used to model a
plucked guitar string.
I Linear
Wave Equation
Example
The equation
utt = c 2 uxx (4)
is known as the wave equation. This equation is used to model a
plucked guitar string.
I Linear
I Homogeneous
Wave Equation
Example
The equation
utt = c 2 uxx (4)
is known as the wave equation. This equation is used to model a
plucked guitar string.
I Linear
I Homogeneous
I Second order
Korteweg-deVries
Example
The equation
ut + uux + uxxx = 0
is known as the Korteweg-deVries (KDV) equation. This
equation is used for modeling solitary waves or solitons.
Korteweg-deVries
Example
The equation
ut + uux + uxxx = 0
is known as the Korteweg-deVries (KDV) equation. This
equation is used for modeling solitary waves or solitons.
I Nonlinear
Korteweg-deVries
Example
The equation
ut + uux + uxxx = 0
is known as the Korteweg-deVries (KDV) equation. This
equation is used for modeling solitary waves or solitons.
I Nonlinear
I Homogeneous
Korteweg-deVries
Example
The equation
ut + uux + uxxx = 0
is known as the Korteweg-deVries (KDV) equation. This
equation is used for modeling solitary waves or solitons.
I Nonlinear
I Homogeneous
I Third order in x, first order in t
Korteweg-deVries
Example
The equation
ut + uux + uxxx = 0
is known as the Korteweg-deVries (KDV) equation. This
equation is used for modeling solitary waves or solitons.
I Nonlinear
I Homogeneous
I Third order in x, first order in t
YouTube
PDE intuition
Question
Why are the previous examples special? I can write down PDEs all
day.
PDE intuition
Question
Why are the previous examples special? I can write down PDEs all
day.
Answer
Many of the PDEs shown can be derived directly using
physics/physical intuition.
PDE intuition
Question
Why are the previous examples special? I can write down PDEs all
day.
Answer
Many of the PDEs shown can be derived directly using
physics/physical intuition.
I Fix x0 . Then ut (x0 , t) denotes the rate of change of u as t
changes.
PDE intuition
Question
Why are the previous examples special? I can write down PDEs all
day.
Answer
Many of the PDEs shown can be derived directly using
physics/physical intuition.
I Fix x0 . Then ut (x0 , t) denotes the rate of change of u as t
changes.
I We put a probe at spot x0 and measure the signal over time.
PDE intuition
Question
Why are the previous examples special? I can write down PDEs all
day.
Answer
Many of the PDEs shown can be derived directly using
physics/physical intuition.
I Fix x0 . Then ut (x0 , t) denotes the rate of change of u as t
changes.
I We put a probe at spot x0 and measure the signal over time.
I Now freeze time t0 . Then ux (x, t0 ) gives slope of u at t0 .
PDE intuition
Question
Why are the previous examples special? I can write down PDEs all
day.
Answer
Many of the PDEs shown can be derived directly using
physics/physical intuition.
I Fix x0 . Then ut (x0 , t) denotes the rate of change of u as t
changes.
I We put a probe at spot x0 and measure the signal over time.
I Now freeze time t0 . Then ux (x, t0 ) gives slope of u at t0 .
I We take a snapshot of u at time t0 .
PDE intuition
Question
Why are the previous examples special? I can write down PDEs all
day.
Answer
Many of the PDEs shown can be derived directly using
physics/physical intuition.
I Fix x0 . Then ut (x0 , t) denotes the rate of change of u as t
changes.
I We put a probe at spot x0 and measure the signal over time.
I Now freeze time t0 . Then ux (x, t0 ) gives slope of u at t0 .
I We take a snapshot of u at time t0 .
I ux and ut are thought of as velocities. uxx and utt are
thought of as accelerations.
Example
Consider
ut + cux = 0. (5)
(queue Mathematica)
Example
Consider the heat equation
ut = σuxx , σ > 0. (6)
Figure : Blue line is a solution to (6). Arrows represent ut (x, t0 ).
Example
What do solutions to
ut + cux = 0. (7)
do?
Example
What do solutions to
ut + cux = 0. (7)
do?
Figure : Blue line is a solution to (7). Arrows represent ut (x, t0 ).
Question
Which direction is this wave “advecting?”
