questions related DEM examples in the libarary

[elhamramyar]

I have some difficulties (questions) in DEM simulation in MARS, I kindly request you to help me for finding the following questions' answers.
1-
From files that you added in examples as DEM simulation, I found that we should use prism command with specific rows in X, Y and Z direction, or something like that for particle generation step, if we want to have a container full of particles how we can estimate the required prism row’s numbers for creating particles that will be compatible with our container capacity (volume)?

I mean in both approach, either with using Macro-Particles or without using Macro-particles (using spherical particles), we don’t know the exact position of particles after their rest in generation step, it means we don’t have their total exact volume, therefore, how can we select proper number for X, Y and Z prism rows?

If we cannot estimate the exact numbers of required particles (or prism rows or etc.) I think, in the generation step, for being sure that our container will be full of particles and we won't have any void, we should generate huge number of particles more than our container volume,do you know how can we delete extra particles in generation step (when we let particles move by gravity) before our main analysis?

2-
In the folders related to DEM examples, you put gen.mrs, run.mrs, hourglass.mrs, Sand.mrs and so on. In the sand.mrs (this file is in both DEM examples), we have X, Y and Z for all particles’ center (with their different radius) obtained at the end of generation step (after letting particles settle by gravity). Then, you properly used sand.mrs (Node list) for the main run in run.mrs.
How did you extract the second position of the particles’ center at the end of generation step? I mean, how did you create Sand.mrs for using in the main analysis?

Thank you for all you help.

Elham

[zhouxinwei]

1. We use a simulation (the generation step) to get packed particles so it is general (instead of generating particles for containers of specific shapes, which needs to be imlemented). As documented in the manual, you can choose from several schemes to generate the particles (prism is one of them). You can use the radius of the particles, the volume of the container to estimate how many particles you need (or add more to be sure they will fill the container fully). Once get the number of particles, you can decide how many rows in X, Y, Z according to the position of the container. When there are more particles than filling the container, those particles can be filter out using the scheme discussed next.

2. To get the particles (sand.mrs), do the following:
- run the generation step in interactive mode (without -B option), monitor and plot the particles to see where they are
- once the particles settled (fill up the container), enter interactive mode (see end)
- follow screen instructions, go to the node list of interest, and dump its informaiton to a file (essentially sand.mrs, may need some formating)

In the main run, do the following

Code: Select all

NodeList Sand { // the whole particle list
  EditNodeList {
    Select ... // choose the particles inside the container
  }
  Make NodeList NodesInsideContainer
}


Then use NodesInsideContainer as the particle for the main simulation.

------------------------------------
How to enter interactive mode:
- while MARS is running, open another terminal window, go to the working directory, run "echo i>stop", which will generate a file called stop with a single character i in it
- MARS will enter interactive mode when it finds stop with i in it under working directory

------------------------------------
How to dump coordinates of a node list in interactive mode:
- once in interactive mode, enter "L", it will print all the lists
- find the index of the node list of interst, e.g. j
- enter "l j"
- enter "W" to dump selected nodes