Learning Forum

[makarand.srinivas24]

I am conducting a CFD analysis of the Electrostatic spray coating process in ANSYS Fluent, and I am currently facing an error of floating-point exception (image 1) after 8.5 hours of runtime. As there are many parameters to tweak, I'm getting confused about what to change now.

The following are the details of geometry and boundary conditions:

Geometry (image 2): domain of 0.3 x 0.3 x 0.4 m. The distance from the spray nozzle tip to the substrate (plate to be coated) is 0.25m.

Models used:

Multiphase > Dense discrete phase model > Phases: 2 - Air (Primary phase), Paint (secondary phase).

Turbulence model > SST k - omega

DPM > interaction with continuous pahse>ON

Injections> ring cone type (inert particle)> velocity=18m/s, mass flow rate=0.0132kg/s, particle mean dia=65 microns distribution pattern: Rosin ramler

Potential > ON

Boundary conditions (image 3);

(top face of enclouser)Downdraft vel (vel inlet): 0.3 m/s

Shaping hole inlet (vel inlet): 10 m/s

Assisting hole inlet (vel inlet): 8 m/s

Atomising air inlet (vel inlet): 3 m/s

Side walls (4 sides): pressure outlet

Substrate (bottom face of enclouser): reflect DPM (all other walls are set to escape for DPM) and voltage of -60,000 V

Material properties: image 4

Methods: image 5

Number of Timesteps: 1000

Timestep size: 0.00005 sec

Iterations/ timestep: 20

The error is coming after 175 timesteps. Please help me with this analysis.

[Rob]

The error is point at an instability in the solution. That could be linked to mesh resolution/quality or time step size. However, the other clue is the potential residual: what's causing it to bounce like that?

You're using DDPM so are interacting with both the DPM and Eulerian Multiphase model: do you actually need DDPM?

[makarand.srinivas24]

I think I would need DPM, because I'm simulating a spray. Also, mesh resolution is enough as the smallest feature is of 0.5mm, and I've given a face size of 0.2mm with 25 layers of inflation around the nozzle body. Also, I've refined the mesh in the area of flow of spray. Below I've attached an image.