GeoXPM output tutorial

Hi @madelestin,

GeoXPM outputs are currently visualized using ParaView. You can find tutorials online on how to work with point data in ParaView.

The readily available outputs have strain values, not deformation. For quantities like deformation, you can use ParaView filters to compute them from particle positions. Similarly, time rates can be derived by back-calculating from the time series.

@tienlk2004 Thank you, i will try to search more on it.

@tienlk2004 hi, i would like to ask what is the unit on "time of simulation"

@madelestin The time of simulation in the output log is second

@tienlk2004 Hi, Hi, excuse me, I’d like to ask something. I want to create a slope in GeoXPM. I followed the tutorial like the one for excavation, where the slope is created using PadDesign so that it already has a shape and mesh. However, when I run it, why doesn’t the result appear in ParaView?

when I did the meshing before running, there is still the soil. 

@madelestin, I can see overlapping both in the geometry you plotted and in the initial particle configuration (soil.jpg). Some soil particles appear to be trapped within the boundaries, and some boundary particles seem to be replaced by soil particles at the left corner.

It’s best to ensure that the object you plot does not have any overlapping. Also, check that your meshing order is correct.

Make sure to check the initial particle configuration in ParaView before running the simulation as well.

@tienlk2004 ahh I see, so if i put the boundary around 500 mm does that mean that I should deduct the sample size by 500mm x 2 for left and right?

@Madelestin, it depends. If you use Face meshing mode and mesh the boundaries to the outside normal directions (e.g., 0, 1, 2), then you won’t have overlapping. However, if you mesh the boundary objects using Solid or Full mode, there will be overlapping based on how you plotted the soil object. In those cases, you can simply move the soil object away from the lower left corner.

@tienlk2004 I see, thank you. I will try as per suggested

@tienlk2004 Hi, sorry I would like to ask again

As for the output, it gives different time stamp as in stages. whereas in geoxpm in the stages, the time of simulation was stated in seconds. How about in the output? because the time steps are different, I tried to put 1 hour (3600 s) and the output is only 17 time

@madelestin both the stage and the time of output are in iterations. If you output every 200 steps then you have 17 frames for 3600 steps. To get the real time you have to multiply the iterations with timestep (dt)

@tienlk2004 so to be confirmed, if i put the iteration as 3600 steps, so the output is 17 frames each representing 1 second? or 200 second?

Because i would like to know the impact of time for the model

@madelestin As I mentioned, it's based on iterations, not seconds. If you set the output interval to 200 and run for 3600 steps, you’ll get 17 frames, each spaced 200 iterations apart.
With a timestep of dt = 0.001s, each frame represents a 0.2-second interval. So by frame 17, the simulation time is 3.6 seconds (Paraview counts frame from 0)

You can get the timestep dt from the output log