Yes, any sub _set_ of a relational structure is (the underlying set of) a sub _structure_. Re: your second question, the answer is also yes if I'm interpreting it correctly: if $\mathcal{A}$ is a structure with underlying set $A$, and $B\subseteq A$ is a subset which is closed under the function symbols of $\mathcal{A}$, then $B$ is (the underlying set of) a substructure of $\mathcal{A}$.
Put another way: **the only way "substructure" differs from "subset" is by requiring closure under the function symbols.** (Note that I'm thinking of constants as functions, here - a constant symbol is a $0$-ary function symbol.)