[varunrajendra]

Hi,

I am looking to model a fluid (diesel) spray into ambient air.

The main results i'm looking for droplet size distribution, spray penetration.

I'm using 2d axisymmetric model. Boundary Conditions (in GAMBIT) i'm using are as follows:

• Axis
• Wall (side, bottom, and top)
• Velocity Inlet ( the top line is made up of two segments, the smaller segment is the velocity inlet, the larger one the wall)

Now I set up Fluent by the following steps

• Select Axisymmetric in Solver
• Select k-epsilon model (realizable) under Viscous Model
• Under Species & Transportation I select "Species Transport", Mixture properties "diesel-air" with everything else unchanged.
• Now I set up DPM. Enable "interaction with continous phase", enable "unsteady particle tracking", Number of Continous Phase Iterations per DPM Iteration set to 2.
  • under Physical Models i enable "DRoplet Collision" and "Droplet Breakup Model". Under Break up constant I set "Breakup Parcel to 50".
  • Under Tracking Tab I set Drag Law to "dynamic-drag"
  • Particle time step set to 0.0001

At this point i get the message saying data required for Standard State Enthalpy, and data required for Reference Temperature.

• Now I go to boundary conditions. Set Inlet Velocity BC. Velocity Magnitude set to 100m/s
• Injections: I create an injection. "Single". Particle Type "Droplet". Material "Diesel Liquid". Flow Rate to be 1.4e-4 kg/s. Start Time 0, Stop Time 0.5s.
  • Under Turbulent Dispersion I enable "Discrete Random Walk Model" and "Random Eddy Life"
• Under Solve-> Controls -> Solution I leave default settings and select OK
• Now I select Initialize. Here out of All Zones/Inlet Velocity/Walls I chose all zones.

I initially got a warning saying non positive volumes exist

• Now I Iterate (200).

I get the following error - floating point error: invalid number


I know i've messed up. I've tried the Evapourating spray tutorial, and refered some other to come up with the procedure I followed. Basically I dont need to create evapourating sprays. Jus simple liquid spray.


An guidance would be appreciated.

[Allan Walsh]

In my experience, the non-positive volume should be more than a warning because Fluent can't run when this exisits. You can either go back to the mesh in Gambit or sometimes you can manipulate the mesh in Fluent (smooth, convert, etc.) to achieve what you want. But, you have to get rid of negative volumes before proceeding.
Just curious, did you export the mesh from Gambit as a 2-D grid?

[varunrajendra]

Yes I exported it as 2D in the mesh save as window.

my geometry is very simple in gambit -

I've got a rectangle. the top edge is basically two segments (one very small 0.5 to act as velocity inlet and the from the end of the 0.5 segment the rest as wall)

[varunrajendra]

I've been trying to sort out the non-positive volume issue... and now i dont get that error, i've prepared the mesh again in GAMBIT and now tried to generate the injection spray simulation.
I've given a lot of tries still get the same "floating point error"


I'm abondoning 2D and gonna try 3D....

[varunrajendra]

Update:

I ditched the 2D model and started woking on 3D.... managed to sort out the issues and the can iterate now....

The spray cant be seen... i can see a point (single particle) when i plot the particle tracks... but when i plot velocity vectors i see quite a few vectors...

any help would be appreciated....