1
$\begingroup$

Recently, I took a course on numerical methods where we learned about Galerkin's Method of Weighted Residuals. We were told that it forms the basis of the Finite Element Method .

However, in most introductory texts I've seen on FEM, they start with a simple example where they manually write out the equations to minimize the potential energy at each node. For instance, see this example of a system of springs. There are no test functions, so it doesn't seem to be the MWR.

So, my question is:

  1. Is there a name for this other method?
  2. What are the practical and numerical differences between it and the Galerkin Method of Weighted Residuals? (e.g., is it easier to set up for certain problems? Does it have better numerical precision or stability?)
$\endgroup$

closed as unclear what you're asking by Paul Mar 22 '16 at 20:02

Please clarify your specific problem or add additional details to highlight exactly what you need. As it's currently written, it’s hard to tell exactly what you're asking. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

  • $\begingroup$ The spring (or in higher dimensions, truss) examples are presented first since they are easy to think about, especially for the undergraduate mechanical engineers for whom many of these introductory texts are written. They serve to show the basic workflow of a finite element solution (element tangent --> assembly --> solution) without muddying the water with test functions and other formalities, which can be addressed in detail later. Once you're comfortable with the actual mathematics involved, you can probably safely ignore the physical example system. $\endgroup$ – Tyler Olsen Mar 20 '16 at 17:43
  • 1
    $\begingroup$ The finite element method is a specific case of Galerkin's method, defined by a special choice of piecewise polynomial functions as test and trial functions. I already gave my opinion on introducing finite element methods via systems of springs or similar in this answer (contrary to @TylerOlsen, I think this muddies the water rather than the other way around) -- the test functions are there as well, but hidden in the way the systems are assembled in a purely local fashion. $\endgroup$ – Christian Clason Mar 20 '16 at 19:19
  • $\begingroup$ @ChristianClason Just read the other answer, and that's fair enough, but you'd have to be in a math class for that approach. I had it introduced to me in a structural mechanics class, so the audience was right for the mechanical interpretation, I suppose. $\endgroup$ – Tyler Olsen Mar 20 '16 at 19:34
  • $\begingroup$ 1. I often hear "this other method" referred to as the direct stiffness method. 2. Weighted residuals exploits the strong form of the PDE, while FEM exploits the weak form of the PDE. See this answer for more details. $\endgroup$ – Paul Mar 20 '16 at 20:32
  • 1
    $\begingroup$ I think your question already has an answer here: scicomp.stackexchange.com/questions/7845/… $\endgroup$ – Paul Mar 22 '16 at 1:30

Browse other questions tagged or ask your own question.