CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   Flow over a cylinder (http://www.cfd-online.com/Forums/main/10944-flow-over-cylinder.html)

 Anna March 7, 2006 23:36

Flow over a cylinder

Hello,

I'm doing a research about the flow over a circular cylinder. I have a few things that I don't understand. My questions are:

1) What is the difference between horizontal cylinder, vertical cylinder, circular cylinder and vertical/horizontal circular cylinder?

2) I found that in some paper, they used the cartesian coordinate in the equation for this problem. Some other paper used cylindrical polar coordinate in the equation. Is it we can use any coordinate for this problem or there is a certain condition in choosing the coordinate?

Can somebody help me? -Anna-

 Kasyap March 8, 2006 04:13

Re: Flow over a cylinder

Dear Anna...

well...i don't know the precise answer for ur first question...but as far as the 2nd quest. considered i think the selection of cord sys. depends on ur convenience...for a cylinder like problem it is obvious that analysis wud be simplified if u use a cylindrical sys

best regards Kasyap

 ramp March 8, 2006 04:42

Re: Flow over a cylinder

Hi !!

[1]. When the flow is approaching over a cylinder placed horizontal in the flow stream (side view will show a cross section of cylinder) is horizontal flow While in case of the vertical cylinder (side view will show a length of cylinder) is vertical cylinder flow

The cylinder may be of two type: one is the circular cylinder which has a circular cross section (say radius r) and other is square cylinder which has the square cross section (say side of square r).

The horizontal and vertical cylinder is nothing but cylinder (circular or square) place in flow streams.

[2]. One can solve the problem in any coordinate systems.

You can also refer the following two books to get the feel of the things about circular cylinder:

1. Zdravkovich, M.M. (1997). Flow around circular cylinder Volume 1, Fundamentals. Oxford University Press, New York, USA.

2. Zdravkovich, M.M. (2003). Flow around circular cylinder Volume 2, Applications. Oxford University Press, New York, USA.

Best regards, Ramp

 Anna March 8, 2006 05:39

Re: Flow over a cylinder

Dear Kasyap and Ramp,

Many thanks for the information given.

Best regards, Anna

 changkiang March 8, 2006 07:24

Re: Flow over a cylinder

well, for the 1st, maybe it depends on whether we take the gravity into account, or there is no difference.

 ramp March 8, 2006 08:49

Re: Flow over a cylinder

Hi !!

The placement of the cylinder and the direction of flow is two different things. The placement of cylinder (horizontal or vertical ) does not account for the gravity.

The flow directions can be horizontal, upward and downward flow. The gravity comes into picture when the flow is vertical (i.e., upward or downward).

Best regards.... Ram

 worasit March 10, 2006 01:06

Re: Flow over a cylinder

Hi! Anna

I think for the 1st question, the major difference to be considered between vertical and horizontal cylinder is 'end effect' not 'gravity'. For short vertical cylinder, end effect can reduces several percent of drag coefficient. Generally, when you consider horizontal cylinder, end effect will be neglected.

For 2nd question, you can use any coordinates system to solve your problem.

Best regards

 ramp March 10, 2006 04:09

Re: Flow over a cylinder

Hi !!

The horizontal and vertical cylinders are completely different situations..... The drag will also not be same for the both schematics.

I think, there will be significant end effects in case of the finite length cylinder (vertical or horizontal) only, not for the infinite length cylinder.

Best regards... ramp

 Anna March 24, 2006 04:18

Re: Flow over a cylinder

Hello again,

I've read some papers about unsteady boundary layer flow past an impulsively started circular cylinder. Based on my understanding,

1) the problem begin with no backflow, then exist a separation point at a finite time, followed by flow reversal

2)for a steady boundary layer flow, there is a singularity at the separation point

3) for unsteady problem, the question whether there exist a singularity is still couldn't be answered.

4) finite-time singularities do not develop in solutions to the Navier-Stokes equations.

5) the flow reversal first occur at the rear stagnation point.

How can we explain the application of this topic in a real situation especially for the small time solution (no flow reversal) for all points along the cylinder surface?

The solution at the stagnation point is important to be find (I think so), what about the solution for other points along the cylinder surface, is it as important as at the stagnation point?

Regards, Anna

 ramp March 24, 2006 15:32

Re: Flow over a cylinder

Hi !!

I think your understandings are correct. If you have not, then have a look on the following two papers...

[1] Williamson, C.H.K. (1996) Three-dimensional wake transition, Journal of Fluid Mechanics 328 pp. 345-407

[2]Williamson, C.H.K. (1996) Vortex dynamics in the cylinder wake, Annual Review of Fluid Mechanics 28 pp. 477-539

These will give you detailed info about stationary circular cylinder.

>How can we explain the application of this topic in a real situation especially for the small time solution (no flow reversal) for all points along the cylinder surface?

Do you mean the applications of these studies (flow without separation behind the cylinder?). They can be applicable in high viscous flow problems such as melts, polymer solutions etc.

>The solution at the stagnation point is important to be find (I think so), what about the solution for other points along the cylinder surface, is it as important as at the stagnation point?

I think solution is important over the complete surface of the cylinder to analyse the detailed characteristics because one can not estimate the local and macroscopic properties like Drag and heat transfer coefficients, etc. without getting the solution over the cylinder surface.

Its not possible to solve the full N-S equation analytically and so the limiting cases such as boundary layer approximation (solution upto the point of separation) and creeping flow (no separation) solution are available in the literature.

best regards.... Ramp

 All times are GMT -4. The time now is 10:41.