Floorless closed-loop microdosing for rail friction: a physics-only digital twin on Kuwait city’s blue corridor

Purpose This study aims to present a lightweight digital twin for top-of-rail (TOR) friction management that relies only on physics and closed-loop control; no artificial “friction floor.” Design/methodology/approach The twin couples along-track film transport (diffusion; diurnal evaporation, curvature-weighted axle consumption, optional rain wash-off) with a monotone, saturating µ(h) map and a local microdosing controller that throttles actuation using a deadband and a footprint-aware safety clamp. The authors first validate on a synthetic loop, then port the workflow to a real corridor by digitizing the proposed Kuwait City Blue Line. This study represents a design-stage digital twin for a planned metro line, providing a transferable framework that can be calibrated and validated on existing operational networks. Findings A configuration with K = 56 micro-dispensers at about 350 m spacing over the top 25% curvature and a ±80 m spray footprint, with per-device cap qmax = 8 × 10–5 m/axle and daytime multiplier 3×, maintains µ = 0.20 everywhere on curves while achieving 6.1% time-averaged and 10.2% final-hour coverage in the target band µ ∈ [0.22, 0.32] under hot/dry summer forcing. The system maintains safety margins while concentrating friction control where it matters most (on D sharp curves) and respects lubricant consumption constraints under challenging environmental conditions. The floorless closed-loop architecture adaptively responds to corridor geometry and time-varying forcing without requiring a friction floor. Social implications The digital twin makes rail travel safer and more reliable by keeping wheel-rail friction steady in rough weather. Braking becomes more predictable, which means fewer sudden stops for passengers and fewer delays across the network. That kind of reliability gets more people onto trains and off roads, cutting both congestion and emissions. And because the model is built on physics rather than massive datasets, engineers can actually understand and trust its outputs, and teams can run it locally without needing to collect mountains of data first. Originality/value The result shows that safety can be enforced by control alone, avoiding policy floors that mask the physics. All scripts, figures and CSV outputs are mapped in an open, reproducible pipeline from route digitization to closed-loop evaluation.