[andy_]

Originally Posted by prapanj

In a previous post, you had mentioned increase in mean flow. So when you say, the increase in velocity due to blockage, does it mean increase in mean-flow?

If you have a constant mass flow (rho*velocity*area) and you reduce the area with a blockage (i.e. your cylinder) then the velocity has to increase. You symmetry condition (and the walls of wind tunnels) prevents the flow moving sideways out of the solution region which is what it would do in a real free flow over a significant number of body sizes.

Originally Posted by prapanj

By symmetry, I would not only prevent flow going out, I would also redirect some flow that has phi outward. Right? So you mean I could also use a wall condition here. This would then simulate a proper wind tunnel.

Why would you want to simulate a wind tunnel? There aren't any walls around your chimney. Your symmetry condition is an artificial one but a wall is even worse because it will not only set the normal velocity component to zero but the tangential ones as well.

Originally Posted by prapanj

In Openfoam, I now tried using inletOutlet velocity condition for the boundaries, which imposes a zero gradient when flux is outbound and a inlet flow when flux is inbound. But this has resulted in very high velocity regions in the domain.

You would seem to be trying to use a modelling package without a grasp of the relevant modelling assumptions it makes. Treating it like a black box is going to waste an awful lot of your time and when you get CFD results out for cases where you have no supporting information from other sources you will not be in a position to know whether to believe them or not making them of little practical value.

I am curious about why you and a significant numbers of other posters here seem to be in this position.

Originally Posted by prapanj

I will try 2D to play with the boundary conditions.

In the early stages of learning about CFD, I would suggest that lots of quick 2D simulations with a range of parametric changes is going to teach you a great deal more than a few, probably failed, 3D simulations.

Originally Posted by prapanj

My reynold's number is always going to stay ~10^7.

The reason for asking is because this puts you in a Reynolds number range where the physics of the flow does not change substantially enabling simpler approaches to turbulence modelling to be reliable.

Originally Posted by prapanj

I am using LES only for a proper vortex shedding, because as a next stage of this project, I will be placing another cylinder in the wake to study the relative load on the second cylinder, and this will be performed for different angles of inflow. Hence using LES. With K epsilon and k omega SST I was not able to get a good vortex street.

Why do you think the real flow has a "good vortex street"?

How much time on a reasonable sized parallel computer do you have to support a set of adequately resolved LES simulations? If you do not have this resource, and it would be odd if you did given that you are clearly not resourced in other ways, I would suggest dropping two of the dimensions and using 2D steady state at least initially.

[prapanj]

Thank you for the clarifications Andy.

Unfortunately the forums are confusing sometimes and once one gets to a forum, he is bombarded with ideas, mostly trivial.

In fact I didn't think there would be a vortex street at this reynold's number, otherwise why would the properties (like Cd) reach a plateau after 10^7. The idea that there will be a good vortex shedding at Re 10^7 was from another member. I couldn't find the truth behind it, because I couldn't read it in any of the books available with me. A few papers i could find were in the Re of 10^5 - 10^6. And almost all papers/books said that properties like Cd would reach a plateau after 10^7. But I couldn't find relavant work published related to this.

The package I use is opensource. The only way I could understand it is by reading the code and talking to people(not good documentation available). It takes iterations to get me somewhere, I am aware of it and would take it really serious when I try to publish some work. And the inletOutlet condition behaves in the way I mentioned before. I was not sure what physical situation I could aptly apply it.

A lot of people are indeed in my position. And I believe this is owing to the vastness of CFD. I had worked on high speed flows, written Euler equation solvers. But when it comes to incompressible slow flows, I am baffled. I hope, with inputs from learned people like you, I would be able to do quality simulations in the future.

You had been really helpful, I really appreciate your advices. I would act accordingly and post my results here.

Thank you

Prapanj.

[hansel]

If you run your cylinder right up to the symmetryPlane aren't you essentially simulating an infinite cylinder, not one 28 m tall? Also is a 1m grid size fine enough for this problem?

[prapanj]

Hi Hansel,

The problem is that of an infinite cylinder essential. I had a span of 28 m because I wanted to have sufficient 3 dimensionality in the domain. Because at a later stage I would be simulation a few other cylinders in the same domain that work as a stack, and study the interference effect.

The grid size is not 1m, the y plus is between 0.4 and 50, with an average of around 10.

Thanks,
Prapanj