Nakamura 1991 Basalt Sphere Impact Fragmentation Validation - GASPHiA Test Case

plloningye@gmail.com (Keqi Ye) — Tue, 26 May 2026 15:20:00 +0800

Nakamura 1991 Basalt Sphere Impact Fragmentation Validation

Overview

This case reproduces the basalt sphere impact fragmentation experiment of Nakamura & Fujiwara (1991). The core comparison quantity is the cumulative fragment mass distribution:

Given a fragment mass threshold M_f, how many fragments have mass greater than it?

This type of statistic is sensitive to brittle damage, fracture propagation, fragment identification, and post-processing algorithms, making it suitable as a validation standard for whether the fragmentation chain is usable.

Key data from this run:

Initial total particle count: 524065
Particles retained after damage threshold filtering (pre-FoF): 157751
Total connected fragments identified: 1594
Effective fragments (particle count > 1): 509
Maximum fragment mass: ~0.262 of the target sphere mass

Validation Objective

Damage model
Particles form reasonable fragment clusters after fracture
Post-processing algorithm accurately identifies fragments
Final fragment mass distribution matches experimental or literature trends

Physical Setup and References

This case is based on:

Nakamura & Fujiwara (1991) basalt sphere impact fragmentation experiment

Model setup:

Target: Basalt sphere
Projectile: Lucite
Primary validation quantity: Cumulative fragment mass distribution

Key geometric and physical quantities set in the input script:

Basalt target sphere radius: 3 cm
Target density: 2700 kg/m^3
Projectile density: 1180 kg/m^3
Projectile initial velocity: 3200 m/s

How to Run

Test directory:

1

GASPHiA-Tests/Nakamura1991_BasaltSphereImpact

The full workflow consists of four steps:

1
2
3
4


python input/input.py
./compile.sh
CUDA_VISIBLE_DEVICES=0 ./GASPHIA -i impact.ini
python post/fast_fof.py -input output_no_bals/nakamura1991_00020.h5 -outputF post/fragments.txt

Actual Run Data

Key metrics extracted from this run and post-processing:

Metric	Value
Initial total particle count	`524065`
Particles participating in FoF after damage filtering	`157751`
Total FoF-identified fragments	`1594`
Effective fragments (particle count > 1)	`509`
Maximum fragment normalized mass	`2.6208e-01`
Minimum fragment normalized mass	`1.6694e-06`
Target sphere total mass	`3.0536e-01 kg`

Currently this case uses the von Mises yield model. Interested readers may re-run it with the Lundborg yield model; the difference between the two is approximately 1%.