Assessing Urban Pavement Performance Under Mixed Traffic Loads: A Field Study

Thesis Abstract

Urban pavements in rapidly growing cities are subjected to increasingly mixed traffic loads arising from private vehicles, commercial fleets, informal transit, and evolving capture of non-motorized modes. This study addresses the gap in empirical evidence on how mixed traffic patterns influence pavement performance, serviceability, and maintenance needs in dense urban environments where traffic composition varies diurnally and seasonally. The aim is to quantify the relationship between traffic mix, loading characteristics, and pavement distress progression to inform more resilient design, maintenance planning, and policy implications. The specific objectives are (i) to characterize traffic load spectra—including axle loads, vehicle class distribution, and traffic flow rates—across representative urban corridors; (ii) to monitor pavement condition indicators (rutting, cracking, potholing, roughness, and layer delamination) over a 24-month period and relate deterioration rates to traffic variables; (iii) to develop and validate a predictive deterioration model linking mixed traffic loads to distress progression using regression and survival analysis; (iv) to assess the effectiveness of current maintenance strategies under varying traffic compositions through cost–benefit and life-cycle analyses; and (v) to provide actionable recommendations for design standards, maintenance thresholds, and policy interventions for urban pavements exposed to heterogeneous traffic. The methodology employs a longitudinal, field-based research design integrating both quantitative and quantitative-empirical elements. The population comprises urban road segments within a mid-sized metropolitan area characterized by diverse land uses and a spectrum of vehicle types. A stratified random sample of 40 road segments, each 1 km in length, is selected to capture high and low traffic variability, different pavement structures, and standard edge conditions. Traffic data are collected through embedded sensors and automatic traffic recorders to yield hourly and daily profile data, including axle-load spectra, vehicle class distribution, and equivalent single-axle-load (ESAL) accumulation. Pavement condition is assessed every six months using internationally recognized methods International Roughness Index (IRI), rut depth measurements, Falling Weight Deflectometer (FWD) for structural response, and surface distress surveys following ASTM D6433. Additional data include climate variables (precipitation, temperature) and moisture conditions, captured via on-site weather stations. Instruments include calibrated pavement distress surveys, calibrated FWD devices, dynamic cone penetrometers for in-situ modulus estimation, and GPS-enabled surveying equipment for geospatial accuracy. Instrument validity and reliability are established through pilot testing and inter-rater reliability checks (Cohen’s kappa for qualitative observations; intraclass correlation coefficients for repeated measures). Data analysis proceeds in three integrated streams. First, descriptive statistics summarize traffic compositions, loading magnitudes, and condition indices. Second, multivariate regression and generalized additive models quantify the relationship between traffic mix variables and pavement distress progression, controlling for climate and material properties. Third, time-to-deterioration analysis via Cox proportional hazards models estimates the influence of traffic variables on the onset of critical distress thresholds. A structural equation model is applied to test hypothesized pathways between traffic loads, structural responses (FWD-derived moduli), and surface distress. Model validation employs cross-validation techniques and information criteria (AIC/BIC). Economic evaluation uses life-cycle cost analysis and incremental cost-effectiveness analysis to gauge maintenance strategies under different traffic regimes. Expected findings include (i) a robust quantification of how heavy axle loads and high commercial vehicle presence accelerate rutting and fatigue cracking relative to passenger-car-dominated corridors; (ii) identification of threshold traffic compositions beyond which maintenance costs rise disproportionately; (iii) validated predictive models capable of forecasting distress progression under observed traffic scenarios; and (iv) evidence-based recommendations for revised material specifications, overlay strategies, and timing of preventive maintenance tailored to traffic mix. The study contributes to knowledge by bridging empirical field measurements with advanced predictive modelling to understand mixed-traffic pavement performance, enhancing the socio-technical basis for urban pavements design and management. It offers practical guidance for engineers and policymakers, including targeted maintenance windows, revised ESAL-based planning under heterogeneous traffic, and cost-effective adaptation strategies for cities experiencing rapid modal shifts. The main conclusion is that pavement performance under mixed traffic is governed not solely by total load but by the composition and temporal distribution of loads, necessitating adaptive design and maintenance frameworks. Recommendations emphasize updating pavement design codes to incorporate traffic mix indices, implementing targeted preventive maintenance schedules aligned with peak-load periods, and expanding sensor networks to monitor evolving traffic profiles for proactive lifecycle management.