Learning Forum

[heb141]

Hi I am trying to simulating particle drop out in different lengths and arrangments of Bev-a-line tube. The model should be quite simple as it is just particulate air flow down a tube, and the particulates have diameters ranging from 0.5 to 5 microns. I am just using a surface injection from the inlet, and using initial conditions of 28 l/min. I have set the boundary conditions as the wall of the pipe to 'trap' and then inlet and outlet to 'escape'.

When I run the simulation, it tells me that all the particles have been trapped (within the first 70mm straight of the arrangement) and none have made it down the length of the tube. I changed the wall to reflect and it seemed that the particles didn't make it down the pipe, and were just stuck bouncing of the sides! I know this is not the case from real-life testing and I'm not sure what I am doing wrong!

Any help or tips on changing the model or setup or anything I would really appreciate as I am very new to DPM!! 😊

[Rob]

Look at what the DPM model is, and what the assumptions are. What sort of volume fraction are you looking at?

[heb141]

I am trying to simulate air flow through the tube, and the air has dirty air particles in it and we are testing particle counters to see if they can detect the dirty air so we don't know for sure the volume of particulates in the air, I just wanted to simulate particles with the density of air, but have a range of diameters to see if there were any problem areas in the arrangement of were particles were being lost! 

[Katya G]

Hi,

if the particles are not detected at the outlet, the Maximal Number of Steps in the Discrete Phase Model panel needs probably to be increased. It is imporant if the particle tracking is unsteady or steady.

Is the fluid flow turbulent? If yes, it is important to predict particle dispersion due to turbulence considering the Turbulent Dispersion as well.

Regards,

Katya

[heb141]

The flow into the the tubes is turbulent and steady, and from my research the tracking should be steady as well. I have turbulent dispersion on as well. When I have been selecting the wall boundary condition, I have been using the wall of the fluid so can't define any material properties of what the inside of the tube would be, do you think this is having an effect as well?

 

[jcooper]

What is the velocity in the tube? 0.5-5 micron particles are very small and should follow the flow in the tube.  If you plot the pathlines from inlet to outlet, and view them, the trajectories the pathlines show is exactly what the particles should do as well.  If the pathlines are very short, it may help to establish the tube flow first and then introduce the particles. If you have set a gravity vector, make sure it is pointing in the right direction.

Very heavy particle loading may just clog up the tube. Loading that is more than 10% by volume should be handled with multiphase modelling. A multiphase analysis is better for high particle loads because dpm cannot account for any interaction between particles.

 

[heb141]

The velocity in the tube is around 14 m/s (inner diameter of the tube is 6.5mm, and flow is 28 l/min). Would you say plotting the trajectory of the pathlines is better or more accurate at predicting particle behaviour in this case?

The main thing I would like to know is a percentage of the number of particles that enter the tube to the ones that are leaving, so looking at the pathlines is good at seeing how the particles flow, but doesn't give me an idea of the particle recovery rate. I hope this is all making sense!

[jcooper]

If the particles are sticking to walls, the pathlines will differ from the particle tracks.  I would use the particle tracks to visualize where the particles are going. The particles nearest to the walls will likely stick and the rest will keep going. 

Looking at the pathlines will tell you if the fluid is capable of carrying the particles out of the tube.  If this is a simple tube with sticky walls, then converging the fluid flow before injecting the particles may help them get further down the tube before sticking.