meltalice · RHEOX
Forward simulation engine · v0.2

Hierarchical branched
polymer architectures.
Real-time rheological
prediction.

G′, G″, η* — sub-second.
No statistical approximation. No truncated trees.

Validated · DOW LDPE L150R · RepTate dataset · 160°C
G′ / G″ — simulated spectrum
ω [rad/s] 10⁻³ 10⁻¹ 10¹ 10³ 10⁵ G′ storage modulus G″ loss modulus
η₀ computed · O(n) · <1s ✓ factorisation valid
scroll ↓

Existing hierarchical solvers
are accurate. But slow.

On heavily branched architectures — LDPE, LCBI polyolefins, asymmetric combs — established solvers require minutes to hours per architecture. Systematic topology exploration in process time is not feasible.

Engineers compensate with heuristics, empirical parameters, and trial-and-error. The cost is not the computation time. The cost is the synthesis campaigns that should never have happened.

O(2ⁿ)
classical hierarchical complexity
on branched topologies
O(n)
RHEOX · algebraic factorisation
of tube model relaxation
< 1s
per architecture · full G′ G″ η*
frequency sweep

Forward simulation.
No approximation.

01

Algebraic factorisation

The viscoelastic response of a branched topology factors through its segment multiset — a structural result in the theory of operadic morphisms applied to decorated tree algebras. The complexity reduction is a consequence of this algebraic structure, not a numerical shortcut.

02

Tube model physics — intact

Arm retraction, constraint release, contour length fluctuations — the physical mechanisms are preserved. RHEOX is not a regression model or a neural network. It is a reformulation of the same physics at lower computational cost.

03

API — your logic on top

The engine is accessible via API. No imposed workflow. Your team defines how the computation integrates into your existing characterisation or simulation pipeline. Access is by request — data stays on your side.

04

Formal treatment in preparation

The mathematical framework underlying the factorisation property is being formalised for academic publication. The engine is the industrial implementation of that result.

One system validated.
Rigorously.

We do not claim broad applicability before we have demonstrated it. The current validated perimeter is stated precisely below.

System Dataset Temperature Perimeter Status
DOW LDPE L150R RepTate reference 160°C G′, G″, η* — full frequency sweep — terminal zone + rubbery plateau ✓ validated
Controlled star / comb (anionic) Known exact topology — pending blind partner ⟳ sought
Industrial LCBI grade Blind prediction vs GPC-MALS reference ⟳ sought

The strongest validation comes from architectures we have never seen. This is why we are looking for blind partners.

Blind validation partners.
Critical collaborators.

🔬

Pilot partners

Companies willing to run a blind validation on 3–5 of their own grades. You provide the rheological data and an independent structural reference. We provide the prediction without seeing your reference. You evaluate independently.

industrial · confidential · no commitment
🎓

Academic collaborators

Particularly on bimodal distributions and hyperbranched extensions of the algebraic framework. Controlled architectures with known topology — synthesised by anionic or ROMP routes — are the most valuable test cases.

academic · co-publication possible
💀

Critical feedback

Especially from sceptics. If you believe the approach is physically inconsistent, algorithmically flawed, or industrially irrelevant — we want to hear it. The strongest validation comes from the hardest questions.

scientific · direct · welcomed

No forms. No commitment. No online access without prior exchange.

contact@meltalice.com

API access
€ 1 /call
First 5 calls free — by request
Request access