Notifications
Clear all
General
6
Posts
2
Users
0
Reactions
28
Views
Topic starter
10/11/2025 10:26 am
Hello everyone,
I’m trying to simulate a 2D tunnel excavation following the 3D Excavation Tutorial as a reference, but I’m adapting it for a 2D case.
Here’s what I did:
-
I created a cube to represent the soil domain.
-
Following Step 2(c) of the tutorial, I created a cylinder inside the cube and used the Slice apart function to separate the soil and the tunnel zone.
-
Then, I added a FillBox ensuring that it covers the entire cylinder, with the FillPoint exactly at the center, where the 2D simulation domain passes.
-
In the GeoXPM tab, I enabled Autofill only for the soil domain, and disabled it for the cylinder (which represents the tunnel to be excavated later), as shown in the tutorial.
However, when I generate the particles, only the particles along the boundary of the cylinder are generated, and the interior remains empty.
I tried activating Autofill for the cylinder as well, and in that case, the particles are generated correctly, but when I run the simulation, the timestep becomes infinite and the simulation fails.
So, I’m not sure what’s going wrong.
Any help or suggestions would be greatly appreciated. I’m just trying to adapt the tutorial for a simple 2D tunnel excavation case.
Thanks in advance!
Topic starter
10/11/2025 10:28 am
10/11/2025 1:44 pm
@joao-vieira The error "unpack requires a buffer of 4570792032" meaning your array size is too big. Check the particle distance and make sure you dont have it too small which will result in huge number of particles and out of memory on your system
For infinite timestep, check that you’ve assigned soil models in your simulation (For the fillbox as well). Without soil models, there’s no information available to calculate the timestep.
A quick tip for your 2D case: keep the 2D cutting section right at the front surface. That way, once it’s cut through the faces, particles will be generated automatically, eliminating the need for fillbox or autofill.
Topic starter
13/11/2025 3:20 am
Thank you for your reply and for the tips.
The particle generation is now working correctly, that part is solved.
However, the issue with the infinite timestep still persists.
I double-checked the particle spacing, and it seems fine, the number of particles is reasonable.
I also made sure that the soil models are properly assigned, including the FillBox, as you suggested.
I’ll attach some screenshots of the material properties and soil assignments so you can check if something looks wrong.
Maybe I’m still missing a detail related to the model definition or the assignment order.
Thanks again for your help!