# Why do curtains in a closed room behave weirdly when ceiling fan is switched on?

I just noticed today that the curtains in my room behave weirdly when fan is switched on. Let me give a rough sketch of what happens before and after the fan is switched on. Below is the side-view.

The window wall has four curtains and this is the case for the middle two curtains.

Before switching on the fan in the room, the curtain is nearly parallel to the window wall behind it. But when the fan is switched on it starts behaving weirdly and three things happen mostly, some times in order then without order:

1. The curvature of the curtain becomes as in the fig. 2 that is a bulge in middle appears and the lower part kind of goes nearer.
2. Sometimes the bulge gets sucked inwards with the lower part still being near.
3. The entire curtain middle and lower both get swept away as shown in fig. 3

I thought that Bernoulli's equation would be able to explain this and below is the diagram for the application for the same.
$$\mathrm{P_0 + 0 = P + \frac{1}2\rho v^2}$$This comes with the assumption that the velocity of air behind the curtains is 0. Now the outward pressure difference is: $$\mathrm{\triangle P = \frac{1}2\rho v^2}$$ but this can only explain the outer bulge and nothing else. Thus I have got two questions here;

Why does this happen?Also is it possible mathematically to find out the curvature of the curtain in fig.2?