Generating very small size voronoi blocks UDEC7

I am attempting to generate very small size Voronoi polygonal blocks for a rock sample. The mean grain size of the rock is 0.0001m. I have been able to build voronoi blocks of 0.002m (figure given below). However, when I further reduce the size to 0.001m (10 times higher than 0.0001m), it gets stuck and does not move further. Then I shut down the program and restart. The code is given below

block tolerance corner-round-length 0.00001
block create polygon 0.0 0.0 0.0 0.110 0.050 0.110 0.050 0.0
block cut split 0.00 0.005 0.050 0.005
block cut split 0.00 0.105 0.050 0.105
block joint-region id 25 0.0 0.005 0.0 0.105 0.050 0.105 0.050 0.005
block joint-region id 30 0.0 0.0 0.0 0.005 0.050 0.005 0.050 0.00
block joint-region id 35 0.0 0.105 0.0 0.110 0.050 0.110 0.050 0.105
block cut voronoi edge 0.002 random-seed 100 iterations 100 round 1.0e-5 range pos-x 0.00 0.050 pos-y 0.005 0.105

Do I need to use smaller round length or more number of random seeds and iterations? Thank you.

PS: The reason for using very small size is because the mean grain size of the rock is very small (0.1mm). If their is workaround where I can use a larger size voronoi blocks, then please suggest.

When trying to calibrate a numerical simulation to laboratory behaviour, you may find that you do not need to exactly match the Voronoi block size to the grain size of the laboratory specimen, but instead also need to consider computational efficiency. The Voronoi size only needs to be small enough (relative to the sample size) to allow representative progressive failure mechanisms to occur.

I would suggest undertaking a sensitivity analysis, gradually changing your Voronoi block size and seeing the failure patterns that occur, and ensure that your chosen Voronoi block size is not controlling the failure patterns that form.


I could not agree more. I would only add that the respective joint parameters (esp. stiffnesses) should be altered accordingly with the varying Voronoi-sizes in order to reproduce a similar macroscopic response. For example, with few Voronois the effective bulk modulus of the block assembly is dominated by the bulk modulus of the grains (since you have comparatively small amount of joints), while joint stiffness dominates at higher Voronoi count.

From my experience, this does not only affect the elastic response, but the failure patterns as well. The stiffness estimation formulae in the manual are typically a good starting point.


Thank you very much. It makes lot of sense.

Thanks a lot. Yes, I will calibrate my model parameters according to Voronoi block size.