Low pressure two-phase flow

[mdickau]

I was wondering if there is a good correlation to use for low pressure/vacuum two-phase flow? I am designing for Annular or Mist flow and want to make sure I properly account for the line losses when sizing the blower.

Also I was wondering if there is an equation or rule of thumb for finding the threashold water and air velocities for Annular or Mist flow in the vertical and horizontal positions. I have seen several graphs but I am not sure which to use.
Thanks for your help,
Michael

 

[BigInch]

2 phase flow analysis pressure drop calcs use several correlations, depending on flow as predicted by the gross liquid flow to gas (or vapor) flow ratios; VDI - for Horizontal and Vertical flow, Duckler for No Slip and Constant Slip cases, Beggs and Brill, and the Lockhart Martinelli correlations.

 

[zdas04]

One thing to keep in mind is that the multi-phase correlations were all developed at pretty high pressure (above 300 psig) and none of them extrapolate very well to low pressure or vacuum service. A really good FEA model of a multi-phase vacuum pipeline with be within 100% of measured values. The correlations aren't even that good.

The thing that every correlation of FEA has to deal with is that there is no such thing as "steady-state" multi-phase flow. Any given flow regieme will last for mili-seconds and then morph to another flow regieme. Every one of them has a different total energy conversion (and conversion mechinism) so local pressure drops are all over the map.

If you have enough flow energy to allow the use of a flow modifying device like VortexFlow's SX tool, then you can separate the flow into two more-or-less single-phase streams and get predictable results. Other than that you could do better with an Magic 8-ball than the fancy arithetic.


David Simpson, PE
MuleShoe Engineering

http://www.muleshoe-eng.com

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

The harder I work, the luckier I seem

 

[panduru]

@zdas04:

"...be within 100% of measured values..." ?

 

[zdas04]

If the actual dP is 200 psid, then the FEA programs may call it 100 or 400 psid. It merely takes a bit of poetical licence to call 200 a 100% increase over 100 and 400 is a 100% increase over 200.

Alright, I was a bit bombastic, but the actual results from this stuff are terrible and people write PhD theses on their "new" multi-phase correlation and only in the furthest appendix (after all the sexy 3D graphs and pictures) do they own up to how terrible their models matched measured data. This is a field where I've watched research very closely and done a bit myself and I can't see where the results are improving over the 13 years since I got my masters.

David

 

[MikeHalloran]

I think he means something like plus or minus a factor of two. Approximately. Which is watchmaker precision in mixed flow.

I've measured Diesel exhaust/ water flows in yacht exhausts, and actually witnessed some of the weird stuff that appears in graphs in Govier/Aziz for annular mist flow. Specifically, a relatively sudden increase in pressure drop with a small incremental increase in water flow, which happens at around 0.6 to 0.8 pct by volume water. The (wet dp/dry dp) goes up from something like 1.2 to 4, or 5.

I have found no published data that is directly applicable, but the phenomenon is well known in certain marine trades, where it makes the difference between getting an engine warranty, or not, takes a lot of people by surprise, and is at the root of some incredibly bad pseudo- science.

That's for gas that's mostly air, at ~800F and ~100mph in a pipe from 6" to 20" diameter, with seawater at 100..130F, at gas pressures up to ~80"wc.

The Govier/Aziz graphs are derived from experiments with natural gas/water in a 1" pipe, but they are as predictive as anything else.

I.e., use all the graphs you can find, and be prepared to make major adjustments when you get a real data point.



Mike Halloran
Pembroke Pines, FL, USA

 

[mdickau]

Thanks for all your help.
Michael

 

[Latexman]

mdickau,

Calculate dP using all the correlations mentioned, pick the largest dP of the lot, double this value, and move on, if it is reasonable and affordable.

Good luck,
Latexman