We can derive easily the expression for tension and velocity for a particle in vertical circular motion. We know that in a string tension occurs when the attached object pulls the rope away from the point where the rope is rigidly attached.
We see that when $\theta$ is acute then there is a radial component of $mg$ which pulls the object, which in turn the rope away from the center of the circle. Thus it produces tension in the rope. But when the angle is obtuse (like $90^o+\theta$), then we see that the component of $mg$ also act radially inwards, so there is nothing which pulls the rope away from the center and also the velocity is tangential.
$T=\frac{mu^2}{r}-2mg+3mgcos\theta$
So, $T=0$, when $cos\phi=\frac{2gl-u^2}{3gl}$
So, when $u=\sqrt{2gl}$, then $T=0$ at $\theta=90^o$
But when $\sqrt{2gl}<u<\sqrt{5gl}$ at $90^o<\theta<180^o$
If we think intuitively, $T$ should be $0$ at all $\theta$ such that $90^o<\theta<180^o$, irrespective of any initial speed greater than $\sqrt{2gl}$. Because at $\theta>90$, the radial component of $mg$,i.e., $mgsin\theta$ also act inwards, so there is nothing which pulls the rope outwards and so alone this $mgsin\theta$ should provide the centripetal acceleration.
We can think $T$ to exist when $\theta>90^o$, when there is some external agency which holds the rope and gives the object a vertical circular motion by pulling the rope to provide centripetal acceleration, but if we think the rope is attached to the nail, there should not be any tension at $\theta>90^o$.
Please help me in explaining intuitively why tension exists in rope at $90^0<\theta<180^o$ when there is nothing which pulls the rope.