Indian urban traffic is the most heterogeneous flow on earth. Motorcycles, auto-rickshaws, cars, buses, e-rickshaws, and hand-pulled carts all share one notional lane. Each behaves like a fluid of fundamentally different viscosity. Nobody has modelled this properly.
Current models — LWR, Nagel-Schreckenberg, IDM — treat the stream as one species. The result is design guidance that fails Delhi's reality: junctions where the bottleneck appears to be a stop line, but the real cause is exit geometry hundreds of metres downstream.
Our hypothesis: a multi-phase, multi-viscosity formulation — coupled to a junction-scale Reynolds number — can predict the moment of transition from laminar to turbulent flow. Fix the exit geometry, and you fix the jam.
The jam front propagates upstream from a poorly shaped exit — exactly as fluid mechanics predicts for compressible flow through a sudden contraction.
Each vehicle class on a Delhi road behaves like a fluid with a distinct viscosity — set by physical footprint, gap acceptance, speed variance, and lateral clearance. The first task of MVTFD is to derive a defensible μeff for each class.
In classical fluid mechanics, the Reynolds number separates smooth flow from chaos. We propose that mixed traffic exhibits the same regime structure — that there exists a critical Re above which a junction collapses from orderly merging into self-sustaining jam propagation.
The central research question is empirical: does this threshold hold for Delhi? And if it does — at what value, and how does it shift with composition?
Thresholds are hypothetical at this stage. Empirical calibration against Delhi field data is the year-one objective.
No existing framework combines multi-class heterogeneity with fluid-mechanics physics at junction scale. That gap is what this programme exists to close.
Structured review of all major macroscopic and fluid-analogy models. Original paper, core assumptions, vehicle types it was designed for, prior application to Indian traffic, and the precise gap relative to MVTFD.
Document the simulation tools state PWDs and traffic planners actually use. CRRI reports, NITI Aayog transport documents, IRC junction design guidelines, MoRTH standards — and the assumptions baked into each.
Survey of multiphase fluid models from chemical engineering and biomedical research — Bingham plastic, Herschel-Bulkley, Carreau, blood-flow analogues. Identify which is the best structural fit for the vehicle-mix problem.
Identify 4–6 candidate Delhi junctions for field data collection. Each documented with vehicle mix, geometry sketch, known congestion pattern, and accessibility for camera placement. Two sites selected after joint field visit.
From published data on vehicle dimensions, typical speeds, and gap-following behaviour: a first-pass derivation of effective viscosity for each Indian vehicle class. Starting from gap acceptance, speed variance, and lateral clearance.
Compute the modified Reynolds number across morning and evening peak windows at the two selected sites. Test whether the proposed threshold reliably predicts the onset of jam propagation.
