Suppose we have a functional integral between some end points and in order to find the function which optimizes it, I apply the Euler-Lagrange to the integrand. What I receive is a differential equation which I can put in the initial or final condition into. The doubt I have is, what conditions are required on the functional equation such that the curve I get in the end is consistent with both boundary condition?

  • 1
    $\begingroup$ The derivation of the Euler-Lagrange equation assumes Dirichlet boundary conditions. $\endgroup$
    – Ian
    Commented May 29, 2022 at 22:31
  • $\begingroup$ what does that mean in English o_o $\endgroup$ Commented May 29, 2022 at 22:33
  • $\begingroup$ Ok I see, that regularity thing I said is there is assumed $\endgroup$ Commented May 29, 2022 at 23:26
  • $\begingroup$ It means boundary conditions of the form $y(a)=y_a,y(b)=y_b$. (I'm a little surprised you know anything about calculus of variations but have never heard of Dirichlet boundary conditions.) $\endgroup$
    – Ian
    Commented May 29, 2022 at 23:53
  • $\begingroup$ My knowledge is basically from a physics context. Maybe I did hear of it before, but it didn't really click stick to me. @Ian $\endgroup$ Commented May 29, 2022 at 23:54

1 Answer 1


It seems OP is putting the cart before the horse. One cannot derive$^1$ the Euler-Lagrange (EL) equations without assuming appropriate boundary conditions (BCs) in the first place.

For a first-order Lagrangian, there are the following possible BCs:

  1. Essential/Dirichlet BCs,

  2. Natural BCs,

  3. Combinations thereof,

cf. e.g. my Math.SE answer here.


$^1$ Of course one can always write down the EL equations, but without appropriate BCs, there is no guarantee that the EL equations are relevant for the variational problem at hand.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .