Hitting a bullet with a bullet - an analysis of colliding bullets using experimental data and explicit simulation methods
Data files
Jul 13, 2026 version files 3.42 GB
-
README.md
8.98 KB
-
RSOS_251800_Dataset_Sandlin.zip
3.42 GB
Abstract
A method to simultaneously fire and achieve a nose-to-nose collision of two bullets was created and compared to simulated results. This research focused on symmetric Taylor impact test techniques applied to 45 caliber Hornady 45177 full metal jacketed munitions. An explicit LS-Dyna model using Finite Element Analysis (FEA) and Smoothed Particle Hydrodynamics (SPH) explicit models of a dual bullet impact were created to determine which simulation approach is more accurate at replicating the realistic impact event.
The experimental collision was observed using high speed cameras to measure impact angles, velocities, and spatial fragment distribution. A Johnson-Cook constitutive material model and failure criteria were used to simulate the bullet materials. This manuscript describes the methods used to approximate and adjust alloy material properties to reduce model correlation uncertainty if the Johnson-Cook values are not available for a selected material.
This paper includes an array of techniques gathered from across the field and discusses how they are employed to study the internal stresses, deformation, petaling, and internal fragmentation in an assembly under impact. Johnson Cook parameters from known alloys were not sufficient to reproduce the symmetric Taylor impact experiments. Specifically, damage parameters, erosion criteria and the hardening exponent (n) had to be modified to reproduce the observations from the experiment.
Dataset citation:
Sandlin D, Cassibry J. 2026 Hitting a bullet with a bullet - an analysis of colliding bullets using experimental data and explicit simulation methods. [Dataset].
Dryad Digital Repository. https://doi.org/10.5061/dryad.1vhhmgr72
General Information
Title of Dataset:
Data supporting "Hitting a bullet with a bullet: an analysis of colliding bullets using experimental data and explicit simulation methods"
Author/Corresponding contact:
Destin Wilson Sandlin, Department of Mechanical and Aerospace Engineering,
The University of Alabama in Huntsville College of Engineering, Huntsville, AL, U.S.A.
Email: destin.sandlin@uah.edu. ORCID: 0009-0009-5781-9555
Co-author:
Jason Cassibry, Department of Mechanical and Aerospace Engineering,
The University of Alabama in Huntsville College of Engineering, Huntsville, AL, U.S.A.
Related publication:
Sandlin DW, Cassibry J. 2026
"Hitting a bullet with a bullet: an analysis of colliding bullets using experimental data and explicit simulation methods."
R. Soc. Open Sci. 13: 251800. https://doi.org/10.1098/rsos.251800 (Received 30 September 2025; Accepted 6 February 2026)
Published in: Royal Society Open Science https://royalsocietypublishing.org/rsos
Simulation Context
These simulations model a nose-to-nose collision of two identical Hornady 45177 full metal jacket (FMJ) bullets
(gilding metal jacket, 95% Cu / 5% Zn; lead core, 97% Pb / 3% Sb),
each travelling at 215 m/s, using the Johnson–Cook constitutive material model in Ansys LS-Dyna.
Simulations run to approximately 85 microseconds of simulated impact time.
Full methodology, material parameters, and results interpretation are described in the referenced paper
The ultra slow motion video of the real bullet collision may be seen on the Smarter Every Day YouTube channel here: https://www.youtube.com/watch?v=tcQVrD7RnNI
Description of the Dataset
This dataset contains the LS-Dyna simulation input files and corresponding rendered visualizations
for six ballistic impact simulations comparing two projectile-modeling methods
- Finite Element Analysis (FEA) and Smoothed Particle Hydrodynamics (SPH) -
with and without an element-erosion (failure) algorithm enabled, plus
SPH runs using Finite Element Model Updating (FEMU) adjusted material properties.
As listed in the paper's Data Accessibility section, the six simulations are:
- FEA without erosion
- FEA with erosion
- SPH without erosion
- SPH with erosion
- SPH with updated (FEMU) material properties without erosion
- SPH with updated (FEMU) material properties and erosion
Software Needed
Three types of files are included:
- LS-Dyna ".k" and ".key" files
These are known as "Run cards" and can be used to recreate the results of these simulations. Run cards have been provided for each simulation.
To recreate the simulations, Ansys LS-Dyna software is required.
The full LS-Dyna simulation output files (d3plot binaries, etc.) are hundreds of gigabytes each and are not included in this dataset.
The user can input these files into a licensed copy of LS-Dyna and reproduce the full simulation output d3 plots.
Warning: These simulations are processor intense, and depending on your hardware they may take days to run. - PNG Image files of simulation renders can be viewed with any standard image viewer.
- MP4 Video files of simulation renders can be viewed with Any standard video/image viewer
These files were generated using Ansys EnSight but do not require Ensight view them.
Pressure values are shown in Pascals, Pa unless specifically labelled as
Folder Structure
README.md and all folders sit under /RSOS_251800_Dataset_Sandlin.zip/
| Folder | Method | Erosion Enabled? | Description |
|---|---|---|---|
| 01_FEA_No_Erosion | FEA | No | Finite Element Analysis without element erosion |
| 02_FEA_with_Erosion | FEA | Yes | Finite Element Analysis with element erosion enabled |
| 03_SPH_No_Erosion | SPH | No | Smoothed Particle Hydrodynamics without erosion |
| 04_SPH_with_Erosion | SPH | Yes | Smoothed Particle Hydrodynamics with erosion enabled |
| 05_SPH_FEMU | SPH + FEMU | No | SPH with Finite Element Model Updating (FEMU) |
| 06_SPH_FEMU_with_Erosion | SPH + FEMU | Yes | SPH with FEMU-updated material properties plus erosion |
For 05 The lead core's Johnson–Cook hardening exponent n adjusted from 1 to 0.392
For 06 (EFFEPS = 2, VOLEPS = 1.5) applied to the copper jacket
Each of the six folders is organized identically, into two subfolders:
Subfolder 1. Simulation_[name]/ - video and image renders of that simulation (see Video/Image Naming Convention below)
Subfolder 2. LS-Dyna Run Card_[name]/ - the ".k"/".key" input deck for that simulation
File Naming Convention
All attempts were made to make everythign self explanatory. Video and Image filenames are long and descriptive.
Video and image filenames are built from the simulation name plus a combination of descriptive keywords.
The files across all six folders follow the pattern:
[FOLDER NUMBER 01-06][FEA/SPH status][Erosion status]_[Keyword Descriptions]
| Keyword | Meaning |
|---|---|
| ISO | Isometric camera view |
| SIDE | Side-on camera view |
| REAR | Rear/back camera view |
| NOSE | View centered on the projectile nose/impact face |
| No Perspective | Orthographic (non-perspective) camera projection |
| Von Mises | Von Mises stress contour visualization (legend colorscale in Pascals, Pa) |
| DATA / GRAPH | Video includes an overlaid data plot (e.g., stress vs. time) alongside the render |
| Copper / Lead / Jackets | Visualization isolating a specific projectile material component |
| MAX | View highlighting/tracking the location of maximum stress |
| 3 Views | Multiple camera angles shown simultaneously in one video |
| TIGHT / CLOSE | Closer-zoomed camera framing |
| Split | Split-screen comparison view. One bullet shows the lead core and the other does not |
| Mesh lines | FEA mesh overlay visible on the model |
| Erosion / No Erosion | Whether the element-erosion (failure/deletion) algorithm was active for that run |
| Blue | Alternate (blue) stress colormap |
NOTES
Folders 03 SPH and 04 SPH with Erosion additionally contain a nested subfolder of numbered PNG still frames (e.g., 03 SPH_015SHBlue/, 04 SPH with Erosion With Erosion Blue/) sequential screenshots taken at fixed simulation timesteps.
The filenames in these 2 nested subfolders follow the pattern:
[Run Number/Name][Time in microseconds][Erosion status]
For example, 03 SPH_008 _SPH Blue No Erosion.png is from the 03 SPH run, captured at 8 microseconds, with erosion off.
Frames are numbered sequentially every 2 microseconds (000, 002, 004, ... 036), giving a still-frame time series of the impact event from 0 to 36 microseconds.
Known issue: A rendering error exists in the interactive Max-Stress probe for Simulation 6 (06_SPH FEMU with Erosion)
The Maximum stress probe location becomes stuck partway through the run as seen in files:
06_SPH FEMU_WITH Erosion GRAPH DATA_Probe stuck at 6 microseconds.mp4
06_SPH FEMU_With Erosion_MAX_STRESS_LOCATION_2 Probe stuck after 10 microseconds.mp4
This is a visualization artifact only and does not reflect the underlying simulation data.
