Computational food physics

Butter Lab BL–05

Butter temperature 7.0°C core 7.0 °C
Transition enthalpy 0.0% crystal network intact
Mass balance 100.0% — g modeled
Wetted footprint —cm² no measurable spread
Thermal history estimating transition time
00:00.0

Phone-safe GPU mode is active. If an in-app browser stops updating, open this page in Safari.

Arrows: orbit · Shift + arrows: heater · +/−: zoom · Space: run/pause

What is being solved

Matter first, pixels second.

BL–05 evolves shared-face butter thickness and eight vertical enthalpy layers on the GPU. Heat arrives through implicit contact, forced convection, long-wave radiation, and partitioned solar absorption. A literature-informed solid-fat prior controls the crystal network, yield stress, and viscosity; the puddle moves through a conservative, positivity-preserving shallow flux.

ρ ∂H/∂t = ∇·(k∇T) + q″    ·    ∂h/∂t = −∇·q

This is a physically grounded educational approximation calibrated from published butter and milk-fat data—not a certified food-process model. The adaptive flow step reports when its positivity limiter is active; capillary contact-angle dynamics remain outside this release. Composition, crystal history, water fraction, and surface condition vary between real butter samples.

Open the source and calibration ledger →

BL–05 GPU laboratory

Preparing the material state

Requesting a high-performance WebGPU adapter…