The solution of a partial differential equation can only be found if the boundary conditions and the initial conditions are given. Therefore, for every problem statement the boundary conditions are very important. However, in many textbooks the boundary condition of the wave equation (or the Helmholtz equation) is not considered when free space propagation is discussed.
Question 1: What is the correct boundary condition (of the wave or helmholtz equation) for a propagation of a wave in free space? I have read that the eigenmodes of the wave equation in free space are plane waves. However, for finding the eigenmodes, there need to be known boundary conditions as far as I know.
Question 2: In many textbook examples the solution of the Helmholtz equation is given by a superposition integral (Example 1). However, I always wonder if this solution set is really complete. In most cases, an ansatz respectively an assumption is used to find this solution of the PDE. This is a contradiction to my undergraduate math lectures. In these lectures, the solution of a PDE was always found by a) finding the eigenfunctions of the PDE from the boundary conditions and b) calculating the eigenfunction coefficients from the initial condition. There was no 'guess' involved in the process of solving the given PDEs.