Cinematic view
A fictional horizon makes the event legible in profile: fireball expansion, reflected light, ground dust, rising hot gas, cloud development, debris, atmospheric drift, and restrained camera response.
Field notes · Educational guide
What the lab visualizes
This interactive explosion simulator turns broad physical ideas—rapid energy release, expanding pressure fronts, buoyancy, drag, turbulence, and wind—into a cinematic browser experiment. It is meant to make phases visible, not to predict a real event.
01 / Method
How it works
Each event follows a deterministic timeline. A short flash gives way to a growing fireball or hot gas volume, an expanding shock-front visualization, surface interaction, a rising column, cloud formation, and eventual dissipation. Presets turn phases on or off and change their timing, energy, debris, dust, cloud, lighting, and wind response.
In Cinematic mode, every preset uses an event-specific simplified GPU fluid profile for velocity, temperature, smoke, incandescent material, dust, pressure projection, and vorticity. GPU-advected tracers provide small-scale detail, while an analytical Canvas layer supplies the environment and family-specific early flash, shock, trail, vent, ground, or electrical effects. Effects overview intentionally remains an abstract analytical view, and Canvas remains available as a compatibility fallback. The system is designed for real-time artistic visualization rather than engineering prediction, and every exposed value remains normalized.
02 / Observation
Two simulation modes
Effects overview
A top-down scientific-instrument view separates normalized fireball, thermal, strong-wave, light-wave, and particulate extents. The circles are qualitative and intentionally use no map, real location, or casualty layer.
03 / Event library
Why the presets behave differently
Conventional blasts and industrial fireballs
Compact events are brief, localized, and debris-heavy. Larger conventional presets push the visible shock front farther. The industrial and fuel-air-style visual archetypes emphasize slower rolling fire, smoke, and atmospheric illumination; they provide no information about fuels, mixtures, construction, or triggering.
Meteors and volcanic events
An airburst forms in the atmosphere without a crater, while a ground impact emphasizes ejecta, surface displacement, and a persistent dust plume. The volcanic preset replaces the detonation flash with a sustained eruptive column and falling ash-like particulate.
Nuclear-scale visualizations
The nuclear-scale presets use broad public visual archetypes: intense flash, rapid fireball growth, shock-front propagation, column rise, and mushroom-cloud formation. A ground burst carries more surface dust than an airburst. These are approximate visual categories only—not yield calculators, damage models, or weapon simulations.
Historical visual-reference presets
Four presets reference broad, publicly known historical test-scale categories—early fission, early airburst, thermonuclear surface, and the largest atmospheric test. They differ in flash duration, fireball growth, stem and cap proportion, dust loading, camera framing, and timeline length. Each is a stylized visual scale category shown over fictional terrain; none contains device details, locations, targeting, damage, or casualty information.
Nuclear Airburst — Research Model
The flagship airburst combines an analytical early shock with a deterministic WebGL2 velocity, temperature, and density field for the later fireball, rising column, and cloud cap. Its browser solver adapts published computer-graphics methods for smoke, fire, turbulence, and volume rendering on normalized fields; it is research-inspired, simplified, and non-predictive.
04 / Concepts
Shock fronts, clouds, and surface interaction
What is a shockwave?
A rapid energy release compresses nearby material and sends a pressure disturbance outward. The lab represents that motion as an expanding, fading front with a simplified radial falloff.
Why do clouds rise?
Hot gases are less dense than cooler surrounding air, so buoyancy drives them upward. In the research model, pressure projection, entrainment, vorticity confinement, and paired cap vortices create the rolling column and outer return flow.
Why does the ground matter?
Surface interaction reflects light and pressure and can lift generic dust or ejecta. Ground-coupled presets therefore look denser and dirtier than comparable events placed above the surface.
What changes cloud shape?
Event duration, buoyant rise, density, coherent vortices, cooling, altitude, and wind all influence the silhouette. The lab exaggerates some of these relationships so they remain easy to see.
05 / Boundaries
Approximation and safety limits
Educational visualization only. All effects, distances, timings, and scales are simplified approximations and must not be used for safety, engineering, emergency planning, targeting, or real-world predictions.
- No explosive construction, materials, quantities, ratios, detonators, triggers, or weapon-design information.
- No maps, addresses, coordinates, real targets, population data, infrastructure overlays, or casualty estimates.
- No building-level damage, optimized burst heights, military accuracy, or recommendations for increasing harm.
- Only fictional, generic, and abstract environments are shown.
06 / FAQ
Frequently asked questions
Is this an accurate blast calculator?
No. The lab is a creative-coding and educational visualization. Values are normalized, rounded, and intentionally avoid the false precision required for real planning or prediction.
Does it use real maps or targets?
No. Every environment is abstract, fictional, or generic. There is no location search, coordinate input, map tile, population layer, casualty estimate, or targeting function.
Can I replay the same event?
Yes. Particle properties and procedural variations are generated from the visible random seed. Replaying with the same seed and controls produces a substantially similar sequence.
How does MP4 export work?
The lab first checks whether your browser can record a genuine MP4. If it cannot, the optional FFmpeg WebAssembly fallback converts a temporary browser recording locally. No video is uploaded. If both routes fail, the original WebM remains available as a correctly labeled recovery download.
Why might mobile exports be limited?
Video encoding uses substantial memory and power. Recent phones default to 720p and shorter durations; older devices may need lower detail density or a desktop browser.