Truss analysis (method of joints/sections) only balances forces, never moments, at each joint. That is only valid if the joint truly cannot carry a moment.
So truss members are assumed to be joined by smooth pins.
A third way to approach this is to think about why the idealization is made at all, by comparing it against how truss joints are actually built in practice (welded, riveted or bolted gusset plates) and asking which idealization best represents pure axial-load transfer.
Comparing the idealization against real construction confirms that the simplifying assumption used is frictionless, moment-free pins.
Therefore, the correct answer is smooth pins.
A steel wire of $20$ mm diameter is bent into a circular shape of $10$ m radius. If modulus of elasticity of wire is $2\times10^{5}\ \text{N/mm}^2$, then the maximum bending stress induced in wire is: