FSI including flow induced vibrations in porous models
 

Hi

I was wondering if anyone here has any experience of modelling flow induced vibrations and the corresponding FSI in filtration equipments in the pipelines. Few of the contributing factors that I see can be:

1) Flow pulsation due to compressor in the vicinity in the pipeline
2) Vortex shedding at strainer pores (perforated sheets)
3) Acoustic characteristics due to incoming flow and the inherent design of the equipment
4) Transient phenomena because of the intricate cavities/recirculation etc in the equipment.

While incorporating the relevant contributing factors from the ones listed above will be difficult enough as it is, the situation will be more convoluted when there are porous interfaces/regions in the equipment.



Thanks
OJ

On the same topic, I have found an interesting paper:

Y Khulief, M. A. Habib, S. A. M. Said, R Ben-Mansour, H. M. Badr, M Anis, and A.E. Khalifa, Vortex shedding-free strainers, Proceedings of the Institution of Mechanical Engineers, Part E: Journal of Process Mechanical Engineering, Volume 226, Number 2, 2012, Pages 91-104.

But I wonder if anyone has any experience of similar simulations. I was wondering, if all these 4 cases are isolated and if the results are obtained separately to find the peaking frequency in each case would give me anything...

Also, given that strainer will be represented as a porous zone/interface, it will practically smear off the perturbations. In this case, would it be reasonable to trust the transient instabilities obtained by FSI on a full model, with contributions from #1, #3 and #4 taken into account?

Any thoughts around this will give me something...

Thanks!
OJ

I think you could simulate a small piece of strainer (which, as I assume, consist of many identical parts) in order to observe the Amplitude of such perturbations. Then you will be able to judge whether these values are significant or not. I suppose not. Such perturbations have a very small scale and, in fact, they smooth the flow. Large scale vortexes (which cause significant pressure oscillations) don't go through them.

Is this flow multiphase? If so, why don't you take into account multiphase instabilities? Are you sure that the flow could be considered as homogeneous?

Andrey,

Thanks for the reply, I knew there wouldn't be many!
Have a look at the schematic of the strainer assembly.

http://www.cfd-online.com/Forums/att...l_strainer.jpg

The typical sizes (of diameter) range upto 72" and the conical strainer is made of perforated sheet.

By small portion do you mean the small periodic section of the perforated sheet? The perturbations for such a small section of strainer will have too small lengthscales but I am not sure if it will/won't have significant effect on strainer. Besides, if we only consider the effect of #1, #3, #4 mentioned in the first post and decide to simulate the full filter with strainer modelled as the porous region, neglecting the effect of perforated sheet perturbation (we have to, we can't model all the holes!), we will have the frequency spectrum. Would the peaking frequency and the magnitudes of forces on strainer that be representative of reality?

BTW, The fluids we consider are single phase, and have too low mach nos. so they can be assumed incompressible.

Thanks
OJ

Yes I meant small periodic section.
I've made a fast look through the article which you mentioned. The Amplitude which represented there looks pretty significant. However the frequency is high, so the question is - would the strainer response to such fluctuations? and would that response be significant?
You could also try to simulate structural behaviour of a strainer under high frequency pressure fluctuations. (not FSI, just transient structural)

The example given in the article is just representative. In another set of operating conditions/size/compressor frequency etc, you may have different results. So, I wouldn't generalise. That said, the question you posed, is the same I am pondering over. Would the effect of perforated sheet be too significant and is its omission is fair?

Approach of "only structural" simulation will be far from the actual situation, and clients are actually insisting on flow induced phenomena, because this simulates their operating conditions where they see failures.

Thanks
OJ

I didn't mean that structural approach solves your case. If it were performed you could probably assess the significance of the effect.
You can simulate several cases with different frequencys and amplitudes of oscillations (based on articles or your expectations/span of a particular interest/intuition) and see what deformation during the time occurs in the structure. It won't necessarily give you the answer but it may help a lot.
And don't abandon idea of periodic simulation of the strainer piece.