[AWalmsley11]

Good afternoon,

I've been using this site for a while for the resource section as well as checking the forums but thought I'd finally sign up and get a bit of guidance with fluent.

I've been using it for a while but been having problems with a particular simulation. It's a simulation of pollutant dispersion but from a small source in a large area. Let's go through dimensions first.

The source is a 0.037m circle and this vents from the ground in a central position. The domain is a horizontally extruded sketch with dimensions 18 hori by 7.7332 vert but with an arc of radius 10m from the centre of the ground line. This is extruded 40m. At one end there is a pressure outlet of 18x7.7332.

The circle is on top of a cylinder 0.1m high

The models employed are species transport, K-e Realisable with buoyancy as well as transient.

I have tried to set the characteristics of the velocity inlet and timing so that it would act as though the flow stops when the cylinder is empty of the chosen fluid. For this to be realistic the flow is 0.1m/s. This however means the Re is 150 and the canister empties in a second. The inlet flow is 100% the pollutant fluid whereas ambient is air.

When I have tried this set up in a smaller domain with non transient flow, the correct physics are observed although the inlet velocity here is 5m/s. However with this, there is a slight change in concentration right above the outlet towards but nowhere near 100% pollutant and no correct jet physics are observed.

Peclet numbers indicate the flow is convection dominant and so inlet diffusion is disabled.

All the converging is fine, it's literally just the jet flow with regards to species and velocity characteristcs. If i initialise the simulation with pollutant as 1 then some velocity characteristics are observed at the bottom of the simulation but anything above this region (about 1m) is made up of pollutant where as the moving area is air.

Can anyone provide any help?

Inlet (shown from under)
http://i.imgur.com/H5jsGR1.png

Underside
http://i.imgur.com/GzGKorP.png

Outlet (square section)
http://i.imgur.com/FCCht3f.png

Edit; URFs not changed due to converging (courant number is 75) and methods are pressure; standard, density; 2nd order, momentum; 2nd order, both turbulence; 1st order, pollutant and energy; 2nd order
Time step, 0.2s 20 time steps, time step iteration limit; 20


RESULT EDIT;
So here's the residuals, velocity and concentration plots from using all first order eqs;
http://i.imgur.com/jF6Xa1U.png
http://i.imgur.com/FibFfU4.png
http://i.imgur.com/NZ7aCoD.png

And the same from the setup from the first edit;
http://i.imgur.com/Ndvf6Bi.png
http://i.imgur.com/oNG3Ak4.png
http://i.imgur.com/SArrfHO.png

As you can see, the velocity is fine here (with the pollutant is initialised as 1) but those species plots are way off.

Sorry for the multi edit, just trying to provide as complete a picture as I can. (anyone spot the irony of this sentence?)

Post thought;

Having had a look around may I be better off using a UDS? The thought had crossed my mind and I have recently acquired a nice tome full of information. The other option was multiphase.

I have had a look through the forum and I've struggled to find something to answer the question. I'm curious as to whether domain size to inlet size ratio is having an effect but as I've shown the velocity seems good so unless said ratio is only affecting the species it doesn't add up.

I'd be really grateful for any answers.