Change a tread recipe and see how rolling resistance, wet grip, wear, hardness, heat and cost are expected to move, compared with a conventional scooter tread.
Surrogate model · synthetic data
What this is. A surrogate model calibrated on published trends and supplier data. The numbers are synthetic, not Eurogrip lab data. The coefficients are meant to be retrained on in-house DMA (tan δ sweeps), DIN abrasion and Goodrich flexometer results before any decision is made on them.
Recipe
phr = parts per 100 rubber
Polymer total100 phr
NR share (balance)–
Total filler–
Silica / filler–
Silane dose–
Plain oil (total − resin)–
Predicted properties
better than ICEworse than ICEICE baseline
Magic triangle
index, ICE = 100
Current recipeICE baseline
Rolling-resistance index–tan δ 60 °C ÷ ICE tan δ 60 °C × 100. Lower is better.
Estimated e-scooter range gain–Estimate only: 0.5 × (100 − RR index) × 0.35. Assumes rolling resistance is about 35 % of e-2W energy at city speed, tyre RR scales with tread tan δ 60 °C, and the tread drives about half of tyre RR.
Pareto explorer
minimise tan δ 60 · maximise tan δ 0 · minimise abrasion
120
Press “Explore 3,000 recipes” to sample the recipe space.
Feasible, dominatedPareto-optimalICE baselineCurrent recipeFeasible = cost ≤ cap and hardness 58–68 Shore A. Infeasible samples are not drawn.
Top 5 Pareto recipes by composite score
Suggest next 5 lab trials
Bayesian-optimisation style
Acquisition = predicted composite score + κ × distance to the nearest tested recipe. Higher κ explores unfamiliar recipes; lower κ exploits the best predictions. In production this is a Gaussian-process surrogate retrained on real DMA, abrasion and flexometer data after every batch, with κ applied to its predictive uncertainty.
0.30
Tested recipes
About the model
trend assumptions and sources by type
Silica with silane replacing carbon black lowers tan δ 60 °C by roughly 20–30 % at equal hardness.Supplier technical bulletins (Evonik / Solvay type silica and silane data); Rubber Chemistry and Technology literature on “green tyre” compounds.
Silane coupling efficiency rises with dose and saturates near 8–9 % of silica mass. Below about 6 % the filler network stays strong (Payne effect), which raises hysteresis, hardness and wear.Silane supplier dosing guides; published strain-sweep (Payne) studies.
Higher vinyl content in S-SBR raises Tg, which lifts tan δ 0 °C strongly and tan δ 60 °C mildly.Polymer supplier S-SBR grade sheets; DMA studies of SBR microstructure.
High-cis BR improves abrasion resistance and lowers tan δ 0 °C (wet grip). NR lowers cost and tan δ 60 °C slightly but weakens wet grip.Tyre compounding handbooks; polymer blend studies.
Traction resin raises tan δ 0 °C and costs some rolling resistance and wear. Plain process oil softens the compound and lowers Tg.Resin supplier application notes (hydrocarbon / terpene-phenolic type).
Sulfur raises crosslink density, which raises hardness and lowers tan δ. Total filler raises hardness and heat build-up.Standard vulcanisation and reinforcement texts.
Heat build-up follows tan δ 60 °C and filler loading. Cost uses indicative Indian bulk prices per kg, normalised so the ICE baseline is 100 (about ₹165/kg).Flexometer correlation studies; public raw-material price ranges, rounded.
Every equation is written as the ICE baseline value plus deviations from the ICE recipe. The baseline recipe therefore reproduces the baseline outputs exactly. All coefficients live in one MODEL object in the page source so they can be refitted to lab data.