# Turbulent mixing of two different gases, how?

#1

I want to model turbulent mixing of two different gases within a flowing pipe, say nitrogen @ 5 m/s from a branch pipe into hydrogen flowing at 1 m/s through the main pipe. I can’t seem to do this using the Multiphase model, as I can’t work out how to assign the two gases to a single volume, or start with hydrogen on the single volume and then introduce the nitrogen at a nominated velocity inlet. How do I do this?

#2

Hi @dsimon!

You can achieve that by using the so called passive scalar transport option. Please note that no chemical reactions are taken into account there. Feel free to check out the public projects library searching for “passive scalar transport” and you will find some good projects.

If you have any questions, feel free to reach out to us and we will jump back in.

Jousef

#3

Hi Jousef,

My simulation is intended to model a piping junction, where the large main run pipe contains natural gas (methane) at 0.2 barg and 0.05m/s flow. A nitrogen gas stream @ 4 m/s flows into the methane gas stream from an attached small bore piping branch.
There is no chemical reaction however I expect the incoming nitrogen will impact the methane flow, is this still sufficiently “passive” to utilize the passive scalar transport option?.
My goal is to better understand how the nitrogen displaces and reduces the methane in the main run pipe, does the incoming jet of nitrogen cause local turbulence and re-circulation proximal to its entry point, whats the optimum nitrogen flow to minimise turbulence and maximise the nitrogen residence time and concentration near the entry point?

Regards, Dean

#4

Theres a mock-up project on my Workbench called “Flare Hot Tap”, to give you an idea of the geometry I’m looking at.

So far I’ve run a number of cases using nitrogen for whole domain, altering the the temperature of the incoming streams as a proxy for the concentration

#5

I personally haven’t seen a multiphase simulation of two gases (because then it’s not multiphase ) but I gave it a try to see if it were possible to run. Can you @dsimon please inspect the animation under this simulation run and let me know if this flow is plausible? I’m not 100% sure if this setup is valid
I used a very coarse mesh just to test it out. Surface tension value was lowered as much as I could to try and get the gases to mix.

#6

Hey nice work!. I think this looks like something I can work.

How did you allocate the different materials/phases to their velocity inlets?

#7

So each material is associated with a phase:

The initial condition is given by this parameter. You can also make subdomains:

And the inlets are being linked to the phase fraction like this (Value 0 = 100% phase 0; Value 1 = 100% phase 1):

.

You should probably also play a little bit with the boundary conditions, especially the pressure outlet, to find out which one is the more appropriate. For this sim I used a plain pressure outlet = 0

Edit: it would be highly appreciated if you update us on your findings when the project is finished (i.e if this modelation proved to be good or not), I’m quite interested in it