RECOMMENDED
PARETO™ K
The knee point — best overall balance
- Full melt2.80 s
- Cost / m³ ice$2,390
- Hazard index0.087
Beats every certified fluid on every axis. $3.26/gal vs $6.41 for Type I PG.
PARETO MATERIALS · INTELLIGENT FLUID DESIGN
An AI engine that designs next-generation aircraft de-icing fluids from first principles — optimized simultaneously for melt speed, cost, and human & environmental safety.
Averages across the optimized Pareto set vs. the global-standard SAE Type I propylene-glycol fluid.
Why it matters
Today's certified de-icing fluids are 1950s chemistry: binary water–glycol blends tuned for one thing at a time. Pareto Materials' engine searches a >1019-formulation chemical space and returns fluids that win on all three axes at once.
Full ice ablation in 2.8 s under a −20 °C jet — vs 8.4 s for Type I PG. Faster melt means shorter holdover, faster turnarounds, fewer cancelled departures.
$2,390 vs $14,489 for Type I PG. The fluid is ~2× cheaper per gallon and needs ~3× less volume — the savings compound.
Composite hazard index 0.087 vs 0.46 — combining oral toxicity (LD50) and biodegradation load (BOD5). Less glycol in the watershed, by design.
Inside the engine
Materials discovery used to mean years of lab screening. Pareto Materials replaces it with a closed computational loop: high-fidelity CFD trains a physics-informed neural network; the network drives a phase-change melt engine; an evolutionary optimizer breeds thousands of candidate formulations against all three objectives — in hours, not years.
168 jet-impingement simulations spanning Re 500–100,000, validated against published experiments to within 3%.
COMSOL · k-ω SSTA PINN surrogate for convective heat transfer — anchored to scaling laws so it extrapolates where pure ML fails (R² = 0.94 beyond training range).
Nu(Re, Pr, H/d, r/d)2-D enthalpy model with solute-driven freezing-point depression, jet wash-out, and temperature-dependent mixture properties.
tfull-melt · φmeltNSGA-II breeds 17-component "cocktails" + jet temperature over generations, against melt speed, cost, and hazard simultaneously.
3 objectives · 18 variablesNot one fluid — an entire menu of non-dominated formulations, tunable to your feedstocks, temperatures, and effluent limits.
your spec → your fluidDiscovered by the engine — a closed-form heat-transfer law extracted from the trained network:
Nu = 0.0245 · Re0.968 · Prjet0.434 · (Prjet/Prwall)0.073 · (H/d)0.006 · (r/d + 1)−1.015
The physics, live
This is the actual phase-change model running in your browser: a warm fluid jet impinging on a −20 °C ice slab. Step up through the three model generations and watch what the missing physics is worth.
Model A — pure thermal melting: heat conducts in, ice melts at 0 °C. This is what glycol-free water would do.
De-icer molecules diffusing into the melt depress the local freezing point toward the eutectic (−46.5 °C) — the front advances faster than heat alone allows. Most CFD studies skip this entirely.
Impinging flow continuously replaces spent, diluted liquid with fresh, hot, concentrated fluid — sustaining both the thermal and solutal attack on the ice.
Viscosity, conductivity, and heat capacity of every candidate mixture are evaluated locally, cell by cell, from Pareto Materials' validated thermodynamic models — so the optimizer never designs on fiction.
The product
The engine doesn't return one compromise — it returns the entire non-dominated frontier. Four representative formulations from the PARETO™ family, all evaluated at −20 °C under identical jet conditions:
The knee point — best overall balance
Beats every certified fluid on every axis. $3.26/gal vs $6.41 for Type I PG.
Velocity — maximum melt speed
For hubs where every second of pad time is revenue.
Cost — minimum spend per m³
90% cheaper per m³ of ice removed than Type I PG — and still 3.3× faster.
Green — minimum hazard footprint
87% lower toxicity & environmental load than Type I PG — for the most sensitive watersheds.
All formulations are multicomponent blends from 17 candidate chemistries, snapped to production-pump resolution. Compositions available under NDA.
Proven on de-icing
Head-to-head against the three SAE industry standards, at identical −20 °C jet conditions, on a like-for-like raw-material basis (2024–26 N. American bulk prices).
Your numbers
Feed in your own fleet and weather profile. The model combines fluid price and melt efficiency into cost per m³ of ice removed — the number your winter actually bills you for.
| Flights de-iced per year | — |
| Ice removed per year | — |
| Baseline annual fluid cost | — |
| PARETO™ annual fluid cost | — |
Cost per m³ of ice removed reflects fluid price and melt efficiency at −20 °C; defaults reproduce the validated internal benchmark. Adjust to your airport's weather exposure and fleet mix.
Engage the engine
Put your product envelope through the engine. We deliver a PARETO™ set of optimized candidate formulations — tuned to your operating temperatures, feedstock constraints (PG-only, bio-content targets), and effluent limits — benchmarked head-to-head against your current product.
Start the conversationContact
Operating temperatures, feedstock constraints, effluent limits — describe your envelope and we'll come back with what the engine can do for it.
We typically respond within two business days. Prefer email?
contact@paretomaterials.com