[yesaswi92]

Hi,

I'm using Star CCM+ for simulating the filling of particles in a vertical pipe. The pipe is blocked at the bottom end, and I'm using injectors to add particles into the pipe from the top. The models I'm using are,
- Implicit Unsteady
- Lagrangian Phase
- Discrete Element Methods (DEM)
- Hertz-Mindilin Contact Model

After the particles get filled in the pipe, some of them start to penetrate out of the pipe wall and go outside the domain. The velocity magnitudes of the particles are also varying erratically.
Where did I go wrong? What could have possibly made the particles to go completely out of the domain? Is it the particle properties or something with the simulation models?

I did the same simulation with a shorter pipe and lesser number of particles. This never happened with that simulation. I'm guessing something is going wrong with the large number of particles.
Can someone please help me with this??

- Yesaswi

[me3840]

Use the coupled solver and a high CFL to make sure that your timesteps are converging well.

[yesaswi92]

Thank you.

I used smaller time steps and it worked.

[Ayush Jain]

Originally Posted by yesaswi92

Hi,

I'm using Star CCM+ for simulating the filling of particles in a vertical pipe. The pipe is blocked at the bottom end, and I'm using injectors to add particles into the pipe from the top. The models I'm using are,
- Implicit Unsteady
- Lagrangian Phase
- Discrete Element Methods (DEM)
- Hertz-Mindilin Contact Model

After the particles get filled in the pipe, some of them start to penetrate out of the pipe wall and go outside the domain. The velocity magnitudes of the particles are also varying erratically.
Where did I go wrong? What could have possibly made the particles to go completely out of the domain? Is it the particle properties or something with the simulation models?

I did the same simulation with a shorter pipe and lesser number of particles. This never happened with that simulation. I'm guessing something is going wrong with the large number of particles.
Can someone please help me with this??

- Yesaswi

Hi Yesaswi!!

How to decide the time step for lagrangian models, as when the unsteady lagrangian solver starts I am not able to know whether the solution is converging within the substeps.

What is active parcels??
How it calculates the progress percentage??
how maximum substeps are decided?