Topic 3 & 4 - Second Order PDEs Flashcards
What makes a PDE linear?
A PDE is linear when it contains no product of dependent-variable terms.
e.g. y(dy/dx) is a non-linear term, as the derivative tells us that ‘y’ is a dependent variable.
How can non-linear PDEs be solved?
Non-linear PDEs cannot be solved analytically, however, they can be solved with numerical methods.
What does a PDE need to be homogeneous?
If every term in the equation for a PDE contains the dependent variable in some form, it is homogenous.
Homogeneous or inhomogeneous?
Inhomogeneous as not all terms contain the dependent variable.
The constant 1 is what makes the equation inhomogeneous.
Why is a curly d used for PDEs over a regular dx?
A curly d is used to communicate that a multivariable function is involved.
What is the significance of this function?
Instead of having one variable changing the other, and this being graphed, for this equation every point on the plane is an input.
Does separation of variables work on non-linear PDEs?
No, it doesn’t.
What would this non-linear PDE look like if you plugged in u(x,y) = f(x)*g(y)
The middle term is not the same as d2u/dx2.
This middle term has two lots of u.
Which is significant as you have to insert u twice. So you end up with one more f(x) and g(y) compared to d2u/dx2.
If the discriminant is positive, what sort of PDE is it?
Hyperbolic
If the discriminant is 0, what sort of PDE is it?
parabolic
If the discriminant is less than 0, what sort of PDE is it?
Elliptic
What sort of PDE is laplace’s equation?
Elliptic
What sort of PDE is the heat equation?
Parabolic
What sort of PDE is the wave equation?
Hyperbolic
What is the general form of a 2nd order Linear PDE?
What is the discriminant function of this equation?