Hello Community
I am facing a problem where I am not able to provide periodic boundary at STL boundaries which are inside the larger simulation box. In the CFD side, I am providing the cyclic bc at inlet and outlet of stl geometry. However, In DEM side, I cannot find any option where I can provide periodic bc at inlet and outlet STL boundaries. As per my understanding, in the liggghts scripts, we define boundary type of simulation box, not of STL boundaries. I looked in the portal but could not find the answer. I wonder, if I can do it anyway?
If I would have had stl boundaries at the same location as that of boundaries of simulation box, I could avaid this and particle will be recirculated from outlet back to inlet like flow is recirculated. But in my case I cant do it because I am injecting particles from inlet using instert/stream command and in order to generate and inject the particles, i need some space (extrude length) behind the inlet, Therefore, my simulation box is bigger than the cfd-domain and this results particle going out from outlet and getting injected back from simulation box boundary rather than STL boundary.
Any suggestion to achieve that?
Best Regards
Atul
mschramm | Wed, 04/13/2022 - 18:49
questions
Correct that there is no way to set up custom periodic interactions with stl files.
My question to use is, does it matter where you insert particles if you are looking at periodic flow?
For me, I would generate a steady-state flow on the cfd side. Then use this steady-state solution as the initial condition of my cfd-dem simulation and set the atoms to have the mean velocity of the flow at insertion. I would then let this evolve over time.
I could be wrong but from the information provided, this would be my guess.
atul2018 | Thu, 04/14/2022 - 10:44
particle at certain mass flow rate from Inlet
Hello
Thanks for your answer!!
As per the experiment I am attempting to simulate, I have to inject particle from inlet at certain mass flow rate (say X kg/min). This can only be possible with insert/stream command that requires that I leave some space behind the inlet STL boundary in order to generate that ammount of particles. Insert/pack command is not suitable for this purpose as I want to insert particles continuously from inlet at certain mass flow rate. So I guess, the location of insertion is important.
About the option you suggested: I am not sure what exactly you mean? Are you suggesting not to use cyclic bc in CFD and DEM sides, rather do steady state CFD calculation first, the use it as initial condition for cfd-dem simulation. But in this case also i need to inject particles from inlet and again that would requires some space behind the space to generate it. when you say "set the atoms to have the mean velocity of the flow at insertion", did you mean that I wont be requiring space behind the inlet to generate it?
I hope I provided enough information and looking forward to hear from you.
Best Regards
Atul