Chetana Rao

NDT Technology for Quality Assurance of HMA Pavement Construction

This report contains the findings of research performed to investigate the application of nondestructive testing (NDT) technologies in the quality assurance of hot mix asphalt (HMA) pavement construction. The report contains the results and analyses of the research performed and presents several key products, notably a recommended manual of practice with guidelines for implementing selected NDT technologies in an agencys routine quality assurance (QA) program for HMA pavement construction and detailed test methods for the recommended NDT technologies. Thus, the report will be of immediate interest to construction and materials engineers in state highway agencies and the private sector.

Calibration of Nonnuclear Density Gauge Data for Accurate In-Place Density Prediction

Hot-mix asphalt (HMA) density is an important acceptance quality characteristic, which involves in situ tests for quality control and assurance (QC/QA). Highway agencies have conventionally used nuclear density gauges or core samples for mat density. More recently, alternate nondestructive testing methods have been considered to replace current test methods. Nonnuclear density gauges offer rapid testing while eliminating safety risks and costs associated with radioactive license. Although agencies have evaluated them, they are not implemented in acceptance testing so far. Results are presented from a field evaluation of three nonnuclear density gauges—PaveTracker, PQI 300, and PQI 301—conducted on Wisconsin Department of Transportation (WisDOT) paving projects. The main goal was to evaluate the performance and effectiveness of non-nuclear gauges for use in QC/QA activities by WisDOT. The study involved field tests at 16 project sites and included 21 mix designs and a variety of mix design and pavement design parameters, such as aggregate type, nominal maximum aggregate size, layer thickness, design traffic, and base type. Density measurements were recorded at 30 test points at each site with one nuclear gauge and three nonnuclear gauges. Although the mean standard deviation values of the nonnuclear gauge data were less than those of the nuclear gauge measurements, a consistent bias was observed between the two data sets. This bias was adjusted by using a calibration factor to yield density predictions statistically the same as the nuclear gauge measurements. It is recommended that a calibration factor determined from 10 points by using a slope function be implemented for agency use. Further, daily calibration for each mix design is recommended when the project involves multiple paving days.

Enhancements to Punchout Prediction Model in Mechanistic–Empirical Pavement Design Guide Procedure

The punchout prediction model for continuously reinforced concrete pavements (CRCP) from the Mechanistic–Empirical Pavement Design Guide, along with improvements to the procedure made since the software program was released in 2004 as a product of NCHRP Project 1-37A, is presented. The punchout prediction procedure is based on mechanistic principles for estimating top-down fatigue damage in the CRCP and on an empirical punchout prediction model calibrated to field data. The mechanistic structural evaluation includes models for estimating crack width, crack spacing, load transfer efficiency across transverse cracks, and fatigue damage accumulation based on Miners hypothesis. The punchout prediction model was calibrated to data on field distress development obtained from CRCP sections nationwide. Most of the field sections were part of the long-term pavement performance (LTPP) experiment, and most of the materials, traffic, climate, and construction data were obtained from the LTPP database. Since 2004, NCHRP has made two main technical enhancements to the design software. The first addressed comments from an independent panel that performed a formal review under Project 1-40A. Second, a systematic error was discovered in the coefficient of thermal expansion test procedure for portland cement concrete, and appropriate corrections were made in the LTPP database. The punchout model was recalibrated to account for this error. The developed models show that the procedure makes reasonable crack spacing, crack width, and punchout predictions that match field observations. The models resulted in similar CRCP thickness designs before and after the coefficient of thermal expansion corrections were made, if the corresponding model calibration coefficients were used.