Hello,
I want to simulate a particle impact damper under hrizontal dynamic load to check its dynamic performance in terms of energy dissipation. The damper is configured as a vertical hollow cylinder partially filled with spherical beads. While trying to model the system in MFiX, I could not find options to apply the boundary conditions which could simultae its dynamic behavior aginst horizontal vibration. I am not sure if I am missing something. Is it possible to simulate in MFix-DEM such particle systems under the mentioned type of loading? Thanks.
1 Like
So are you talking about modeling this?
https://www.mwrf.com/technologies/nasa-transfers-damping-technology-topline
MFiX does not have moving boundaries out of the box. Are you modeling pure granular flow (no fluid)? What has to move dynamically? The cylinder?
Thanks for the reply @onlyjus. Yes, my model uses the same PID concept and it will be pure granular flow. The cylinder partially filled with granular material will be excited by a dynamic load acting in horizontal direction (shown in attached figure).
The granular material will act as damper to dissipate the vibration energy.
If your horizontal excitation can be cast as a force, then you could hack it into gravity. it’s not a frame indifferent equivalent of actually moving the container, but it’s the most straightforward if you can live with it. other options are to model the container with particles and force the motion of the “container particles”, Otis Walton has done this for a tumbler setup somewhat recently (@benyahia might have his udf? I can’t seem to find it in my email) . Finally, it’s possible to move the STL, but it’s difficult and expensive and as far as I know only @jeff.Dietiker knows how to do this.
Thanks @wfullmer. Yes, the horizontal excitation is a force but it is, being a dynamic force, a fuction of time and changes both its magnitude and direction with time so I think it cannot be modeled as gravity. I will try the other option (of modeling the container with particles) you suggested and see if it works.
we’ve hacked in something like this
! Rotating gravity vector pliu
GRAV(1) = - GRAVITY * DSIN(ANG_FREQ*MAX(0.0d0, time - 1.0d0)) !wdf
GRAV(2) = - GRAVITY * DCOS(ANG_FREQ*MAX(0.0d0, time - 1.0d0)) !wdf
!GRAV(3) = GRAVITY_Z
in des_time_march.f before the call to before to calc_force_dem to fake a rotating tumbler.
1 Like
Do you have any papers or references for the work by Otis Walton or Jeff Dietiker that you mentioned? I’ve been using the fluctuating gravity “hack” for some time, but I will need to model the periodic motion of my container using particles (or move the STL) for the next geometry case I have in mind.
I know Fluent has remeshing capabilities, so I could in theory export the periodic mesh position data for my vibrating container from Fluent and use it as an input to MFIX, but I’ve never done this before. Maybe that’s not the best method?
actually Jeff released some UDF examples using particles as moving boundaries in the latest release. I know @sathishsanjeevi worked on this, perhaps he could point you in the right direction.
I still think the gravity “hack” is best suited for your application, since moving STL would require considerably “larger” domain due to the displacement. Do you see any unphysical behavior with the gravity hack?
William, I hadn’t downloaded the latest release yet so I will check that out, thank you!
Sathish, the issue for me with the gravity hack is that my new baseplate geometry is going to have protrusions (similar to a sawtooth grate) and I have a 2D sinusoidal excitation - one component perpendicular to the grate’s primary axis and another parallel. I’m worried that using 2D fluctuating gravity for this is not going to capture the near-wall behavior correctly (i.e. particles contacting the grate protrusions should experience a higher “fluctuating gravity” force than particles between grate elements), but I haven’t tried it yet so maybe this won’t be a problem.
Likewise, moving STL file is difficult than fluctuating gravity. I don’t think you would be facing any issues as far as the relative frame of reference is appropriately considered, while using fluctuating gravity.