Monday, 27 July 2015

Complex functions and flow around a cylinder

Last semester I attended a course in complex analysis and I was introduced to the basics tool used in this field. I was fascinated by almost anything that I was taught and although some topics were a little bit difficult to grasp, I am sure it was worth it all the way. Complex Analysis is great for many applications, it is a very powerful tool that enables you to solve problems in the real domain that you would have not otherwise been able to solve.

Among the many applications of complex functions, potential flow is one that took my attention, mainly because it is easygoing for a novice in this field (as I am) but also because of the nice graphs one can plot using it and their physical interpretation.

While looking up for some materials I stumbled upon this paper.

It is quite a nice piece of paper, since it gives a little bit of introduction to complex functions and how they can be used to model flow.

An ideal flow is a flow that is both incompressible and irrotational, that’s to say, if F represents the fluid velocity at each point (x,y), then the following conditions must hold in order for the flow to be ideal:


The conditions above require that the flow has no vorticity (curl) and that the volume of the fluid does not change as it flows (divergence equal to zero).

Theorem 4.5 in the paper connects the velocity vector field and complex (analytic) functions. By then integrating the velocity one can find out the streamlines of the fluid flow.

If the velocity F(z) admits a complex anti-derivative,

clip_image002[11] and


Therefore the gradient of phi is equal to F and the real part phi(x,y) defines a velocity potential for the fluid flow. The harmonic conjugate psi(x,y) to the velocity potential, in fluid mechanics, is known as the stream function.

Example (4.8). Consider the complex potential function


By making some calculations we can find out


with streamlines


asymptotically horizontal at large distances.

By using Matlab, I implemented this example and plotted some plots about our Gamma function:


Now we plot the velocity (gradient of the real part of Gamma) and the the streamlines (psi).


Last but not least, the contour plot of the absolute value of Gama and of its real and imaginary part.


As you can see, the stream function shows that this model can be a first raw step in the way of modelling flow around a steady cylinder. This model however is far from being perfect: among other assumptions it assumes no friction and laminar flow. Laminar flow  can be assumed when the velocity of the fluid is very low and in many practical application it is not often the case. Anyway, this simple model surely can draw some attention to complex analysis and its tools.

No comments:

Post a Comment