4 replies to this topic

[JTL]

Hi,

 

I'm dealing with a simulation code for dynamic pressure-flow calculation of different fluids. The solution assumes the fluid as homogeneous mixture. The problem is when calculating fluid properties (density, viscosity, thermal conductavity, heat capacity, ..) for a mixture including both liquid and gas (of a single substance such as methane, or different substances such as water and air).

 

So, after calculating the pure phase properties, how to derive the mixture properties? Especially viscosity and thermal conductavity should probably be calculated other than mass-weighted average? Are there any general approach available?

 

 

 

[PingPong]

There are no mixture properties, not even the mixture density can be determined accurately because vapor and liquid flow at different velocities, depending on the actual liquid holdup in the pipe.

 

If your software, or simulation code, cannot handle input of vapor and liquid separately then it is safe to assume that that software is unsuitable for the purpose you have in mind.

 

I understand that that is not what you want to hear, but it will save you a lot of time to wake-up now instead of next month.

[JTL]

Thanks for your straight answer. It's next month already and still in front of this question

 

It's right that the mixture properties are basically not defined, especially Cp is not but also other ones are not reflecting real world like you described.

 

But, let's suppose we do approve the assumption of homogeneous mixture because mostly the targeted studies concern 1-phase cases (or have such good mixing). But occasionally we may have 2-phase flow, so some rules for mixing the properties must be used to provide some results anyway. Any working solutions known?

[DanW]

Maybe this helps: https://github.com/D...pertyPackage.vb

 

Take a look at the code from line 927 to line 1070.

[JTL]

Thanks a lot DanW,

that was really interesting to look at, when I finally did it, sorry for the long delay.

 

I was happy to see similar approach for most properties that used myself, and interesting to see one differing: thermal conductavity. It is calculated based on molar fractions of the phases in the code. Is there any argumenting for that choise?

 

I was unhappy to find nothing for viscosity, any ideas for averaging that property for a Liq+Gas mixture?