Initial time step size in poromechanics runs with initialization

[ryar9534]

I believe that currently, a specified initial time step size is overwritten/not respected when poromechanics initialization is performed. For example, if the poromechanics initialization is done with a time step of size 1e11, the first step of the actual simulation will be whatever the max is, bypassing any specified initial time step from the input file.

I will take a look at fixing this.

[paveltomin]

#3396 addresses the issue