Validation of heavy vehicle loading responses and temperature predictions in flexible pavements using field data
Time: Fri 2022-12-16 13.00
Location: U1, Brinellvägen 26, Stockholm
Video link: https://kth-se.zoom.us/j/64328878182
Subject area: Civil and Architectural Engineering, Building Materials
Doctoral student: Denis Saliko , Byggnadsmaterial
Opponent: Professor Inge Hoff, NTNU Norges teknisk-naturvetenskapliga universitet
Supervisor: Adjungerad professor Sigurdur Erlingsson, Byggnadsmaterial; Doctor Abubeker Ahmed, VTI; Docent Denis Jelagin, Hållfasthetslära (Avd.), Väg- och banteknik, Byggnadsmaterial, Vägtekniks laboratorium
It is well established that both traffic-related loading and environmental conditions influence the structural behaviour of pavements. Pavement design methods aim to consider the effect of traffic loading and environmental variables on pavement structure, foresee their changes during the lifetime of the pavement and predict the resulting distresses and pavement life. Newer models are required to further advance the development of pavement design methods. Validations using reliable and representative data are required prior to incorporating these models in pavement design methods.
The impact of environmental factors such as temperature, moisture content and freeze-thaw cycles on pavement behaviour have been examined in this doctoral thesis. Furthermore, the impact of increased loading by new long heavy vehicles on low-volume roads subjected to large variations of the environmental conditions has been investigated. The findings presented in this thesis have been based on field data collected on roads that have been and currently are in day-to-day operation. The collected data on mechanical response, temperature, moisture and frost was used to validate models on mechanical behaviour and thermal behaviour, as well as the effect of their interaction in pavement structures.
The models developed and validated in this study are aimed to be integrated into a new mechanistic-empirical pavement design framework that is currently under development in Sweden. The work done for this thesis is presented hereby in the form of 5 papers and a short summary part. This thesis is a continuation of a licentiate thesis previously published at KTH Royal Institute of Technology. Part of the material published in the licentiate thesis has been included in this doctoral thesis.
In paper 1, air temperature data recorded over a span of 10 years from 44 meteorological stations and temperature data from built-in sensors in 49 pavement structures located in different locations throughout Sweden were analysed. The data was used to statistically correlate the freezing index, calculated using the mean daily air temperature and the frost penetration depth in the cross-section of the pavement. Comparisons were made for the results obtained for various climatic zones in the country. The output of the paper is a country-specific empirical chart obtained through exponential interpolation and nonlinear prediction limits that indicates a range of expected frost penetration depth based on historical air temperature data.
Paper 2 presents a study in which the structural response of a test section was evaluated using built-in sensors. The instrumentation consisted of asphalt strain gauges (ASG) recording the tensile strain in the bottom of the asphalt layer, strain measuring units (εMU) recording the vertical strain in the granular layers, and soil pressure cells (SPC) recording the vertical stresses in the granular layers. Falling weight deflectometer (FWD) measurements were performed on the structure to backcalculate the stiffness of the layers from the measured surface deflections. The aim of the study was to evaluate the structural response of the structure under loading by three long heavy vehicles (LHV) weighing ~64 tonnes, ~68 tonnes, and ~74 tonnes and compare the resulting estimated accumulated damage from each vehicle. The main finding from the paper was that the damage caused to pavements by long heavy vehicles was slightly larger than the damage caused by shorter vehicles with fewer axles but higher axial loading and tyre pressure.
Paper 3 focuses on the effect of environmental factors and their variation on the structural behaviour of a thin pavement structure. Loading by long heavy trucks was applied to a test section at four different measurement campaigns performed at different seasons over one year. The variation of temperature and moisture in the structure was monitored continuously for the entire duration for which the study was performed. Thermocouples embedded in the asphalt layer and a frost rod placed in the granular layers were used to monitor the temperature variation in the structure. The moisture variation was monitored using time-domain reflectometer (TDR) probes. The correlation of the changes in temperature and moisture to the changes in mechanical stiffness of the layers was investigated.
The instrumentation used to monitor the mechanical response, temperature variation, and moisture variation in the structure was found to be reliable for collecting data over the entire duration of the study. The main finding of the study is that it is possible to model the mechanical behaviour of thin pavement structures using multilayer elastic theory (MLET) calculations modelling, using linear-elastic material models if the stiffness of the asphalt layer is adjusted based on temperature and the stiffness of the granular layers is adjusted based on moisture levels.
In Paper 4, the same response testing procedure as in Paper 3 was performed for a second pavement structure with a thicker asphalt layer. Data from response testing results for 2 pavement structures on 4 different dates, with a focus on the spring thaw period, were considered in the paper. Three different strategies for material modelling were used to investigate the mechanical response of the pavement structures. The layers were initially modelled using linear material parameters and the response results were compared both to calculations in which a viscoelastic model was used for the asphalt layer and to calculations in which a nonlinear K-Theta model was used for the granular layers.
Comparisons were made between the calculated response using each modelling strategy and the measured response values. It was found that the viscoelastic and nonlinear models provided only marginal improvements in the range of 1%-4% in predicting the mechanical response of the structures. Based on the results, it was concluded that the linear elastic model was sufficiently accurate in capturing the mechanical behaviour of both pavement structures, including at the critical locations.
Paper 5 presents the development and validation of a one-dimensional finite control volume (FCV) model capable of predicting temperature in pavements. The model is intended to be implemented into a new mechanistic-empirical pavement design framework currently under development in Sweden. The model uses easily obtainable meteorological data for air temperature, solar radiation, and wind speed for the three main modes of heat transfer, namely conduction, convection and radiation.
To validate the model and estimate its accuracy, comparisons were made between the measured temperature and the calculated temperature values, using the FCV model. Comparisons were made for the pavement surface temperature, the temperature within the asphalt layer, and the temperature in the granular layers for 4 pavements located in different climatic zones in Sweden. In general, good agreement was found between the measured and calculated temperature values. Points for future improvements include better consideration of the surface properties, including the latent heat transfer in the calculations, and coupling the model to a moisture transfer model.