Your First HBM Simulation with Open Tools

A student learning project by Yash Kumar Vari. Inspired by students’ questions on LinkedIn and conversations at IRCOBI 2026 about entering crash simulation and biomechanics.

Learn how a VIVA+ 50F head-drop simulation at 3.5 m/s was adapted for OpenRadioss, how to inspect results, and how to question a force curve. Start with the included data: you do not need a solver license or HPC access to reproduce the figures.

Start here

1. Read: open the guide through the website link in this repository’s About section. Alternatively download the repository using Code → Download ZIP, extract it and open index.html locally.
2. Watch: smooth five-loop animation, 23.125 seconds in slow motion.
3. Analyze: follow the commands below. Four figures and result tables are included.
4. Ask: use this repository’s Discussions → Q&A for learning questions; use Issues for broken instructions and errors.

Reproduce the analysis

Open Windows Command Prompt in this extracted folder:

py -m venv .venv
.venv\Scripts\python.exe -m pip install -r analysis\requirements.txt
.venv\Scripts\python.exe analysis\analyze.py

Linux/macOS:

python3 -m venv .venv
.venv/bin/python -m pip install -r analysis/requirements.txt
.venv/bin/python analysis/analyze.py

Outputs go to figures/ and analysis/. Expected filtered support-force peak: approximately 3.913 kN. The analysis has been executed locally from this layout; the Windows command route is standard Python syntax but has not been independently run on a Windows machine. The raw VTK sequence is not included; the finished MP4 is included.

What the case teaches

• Model inputs, units, constraints and contact.
• Starter and Engine roles, logs and result checks.
• Why support reaction differs from head-pad contact force.
• Component-wise CFC1000 filtering and careful time-axis interpretation.
• How to document discrepancies and plan a controlled investigation.

Current result and limits

The final filtered support force is 3.913 kN, versus the published experimental average of 4.601 kN (14.95% lower). It peaks 1.635 ms later on the reference plotting convention. Published LS-DYNA curves peak at 4.785 and 4.825 kN. This adaptation has not reproduced the published experimental response. Material translation, contact and exact source-version equivalence require investigation.

The Linux/HPC head-drop run is demonstrated. A complete portable solver installer, the Windows head-run route and independent student reproduction remain unfinished. The guide labels these boundaries rather than promising a turnkey simulation kit. Methods · Input edits · Source notices.

Files

Location	Purpose
index.html, assets/	Student guide and learning illustrations
media/	Five-loop final animation
figures/	Four PNG/PDF result figures
analysis/	Script, raw and processed CSVs, requirements, checksums and upstream filtering dependency
inputs/	Record of input edits; not a portable installer
docs/	Methods and student exercises

Credits

Original HBM: VIVA+ team. Published setup and reference curves: VIVA+ validation contributors, using the Yoganandan head-drop experiments. Solver: OpenRadioss community. Force filtering: Dynasaur / VSI TU Graz. Yash Kumar Vari’s contribution is the educational adaptation, execution, troubleshooting, analysis and explanation. This is an independent educational project; no endorsement by Audi, VIVA+ or solver providers is implied.

See VIVA+ citation instructions and the reference case. This repository is a learning pilot, not an injury-prediction or safety-qualification tool.