Vortex shedding is often investigated in the flow past a cylinder. In this case visualisation methods ((e.g. PIV) are often applied. I think you can get a number of papers on this on Scopus/WoS. 

There is an excellent high-level article that describes this in Physics Today from February 2013, p.41 by Thomas Peacock and George Haller. (http://georgehaller.com/reprints/PhysToday.pdf)

Dear chirag, if you are looking at saddle points in flow past square cylinder, just go through yoon 2010 publication "Flow past a square cylinder with an angle of incidence".

http://dx.doi.org/10.1063/1.3388857

Dear Chirag,

Saddle points can generally be regards as separatrix of different regions with distinct behaviors. It can be visualized in both 2D and 3D flow regions. Bear in mind that the argument that "distinct trajectories either diverge to infinity, or converge to a limit point or a limit cycle" is only valid in 2D. When the phase space dimension increases up to 3D and higher, the trajectories can enjoy the freedom of staying confined in the phase space (strange attractor). That is, the visualization of saddle points is only considered meaningful in 2D flows.

Thank you.

Very kind wishes,

Wang Zhe