Can Partial Differential Equations Have Non-Separable Solutions?

Click For Summary

Discussion Overview

The discussion centers around the nature of solutions to partial differential equations (PDEs), specifically whether non-separable solutions exist and how they might be identified. The conversation touches on theoretical aspects of PDEs, methods of solution, and the limitations of common approaches like separation of variables.

Discussion Character

  • Debate/contested
  • Technical explanation

Main Points Raised

  • One participant suggests that PDEs are typically solved by separation of variables, which assumes solutions can be expressed as products of functions dependent on individual variables.
  • Another participant counters that separation of variables is not the predominant method for solving PDEs, arguing that most are solved numerically and that classroom examples do not reflect the broader reality.
  • A third participant notes that many PDEs are solved using Fourier series methods, implying a connection to separable solutions.
  • A later reply emphasizes the interest in solutions that do not conform to separable forms, providing an example of a potential non-separable solution, e^(xy).

Areas of Agreement / Disagreement

Participants express differing views on the commonality of separation of variables as a solution method for PDEs. There is no consensus on the existence or identification of non-separable solutions, and the discussion remains unresolved.

Contextual Notes

Participants highlight the limitations of common solution methods and the potential for alternative approaches, but do not resolve the conditions under which non-separable solutions might be found.

pivoxa15
Messages
2,250
Reaction score
1
The title should have been partial differential equations.

PDEs are solved usually by separation of variables but that assumes each solution is a product of two functions which are only dependent on one variable only.

But could there exist solutions which are not in the this form? If so how would you find them?
 
Physics news on Phys.org
I take issue with that. PDEs are not 'usually solved' by separation of variables since only a tiny fragment of PDEs are solvable in this manner. The majority of PDEs that are solved are surely done numerically.

What happens in a classroom example rarely approximates the normal state of affairs.

Try googling for existence and uniqueness of solutions to PDEs. I know there are results in the one variable case (the Lipschitz condition, for example), but I don't know about the multivariable one.
 
And many partial differential equations are solve by Fourier series methods.
 
HallsofIvy said:
And many partial differential equations are solve by Fourier series methods.

That is after you assume the separation of variable solutions though? I was asking for solutions with variables that are unseparable (i.e. e^(xy) as a solution)
 

Similar threads

  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 18 ·
Replies
18
Views
5K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 5 ·
Replies
5
Views
3K
  • · Replies 5 ·
Replies
5
Views
4K
  • · Replies 36 ·
2
Replies
36
Views
6K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 9 ·
Replies
9
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 5 ·
Replies
5
Views
2K